Usability and Internationalization

of Information Technology
 Human Factors and Ergonomics
Gavriel Salvendy, Series Editor

Aykin, N. (Ed.): Usability and Internationalization of Information Technology.

Hendrick, H., and Kleiner, B. (Eds.): Macroergonomics: Theory, Methods and
Applications.
Hollnagel, E. (Ed.): Handbook of Cognitive Task Design.
Jacko, J.A., and Sears, A. (Eds.): The Human-Computer Interaction Handbook:
Fundamentals, Evolving Technologies and Emerging Applications.
Meister, D. (Au.): Conceptual Foundations of Human Factors Measurement.
Meister, D., and Enderwick, T. (Eds.): Human Factors in System Design, Development,
and Testing.
Stanney, K. M. (Ed.): Handbook of Virtual Environments: Design, Implementation, and
Applications.
Stephanidis, C. (Ed.): User Interfaces for All: Concepts, Methods, and Tools.
Ye, Nong (Ed.): The Handbook of Data Mining.

Also in This Series

HCI INTERNATIONAL 1999 Proceedings 2 Volume Set

• Bullinger, H.-J., and Ziegler, J. (Eds.): Human-Computer Interaction: Ergonomics
and User Interfaces.
• Bullinger, H.-J., and Ziegler, J. (Eds.): Human-Computer Interaction:
Communication, Cooperation, and Application Design.
HCI INTERNATIONAL 2001 Proceedings 3 Volume Set
• Smith, M.J., Salvendy, G., Harris, D., and Koubek, R.J. (Eds.): Usability
Evaluation and Interface Design: Cognitive Engineering, Intelligent Agents and Virtual
Reality.
• Smith, M.J., and Salvendy, G. (Eds.): Systems, Social and Internationalization
Design Aspects of Human-Computer Interaction.
• Stephanidis, C. (Ed.): Universal Access in HCI: Towards an Information Society for All.
HCI INTERNATIONAL 2003 Proceedings 4 Volume Set
• Jacko, J.A. and Stephanidis, C. (Eds.): Human-Computer Interaction: Theory and
Practice (Part I).
• Stephanidis, C. and Jacko, J.A. (Eds.): Human-Computer Interaction: Theory and
Practice (Part II).
• Harris, D., Duffy, V., Smith, M. and Stephanidis, C. (Eds.): Human-Centered
Computing: Cognitive, Social and Ergonomic Aspects.
• Stephanidis, C. (Ed.): Universal Access in HCI: Inclusive Design in the Information
Society.
Usability and Internationalization
of Information Technology

Edited by

Nuray Aykin
Siemens Corporate Research, Princeton, NJ

LAWRENCE ERLBAUM ASSOCIATES, PUBLISHERS

2005 Mahwah, New Jersey London
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2005 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20130731

International Standard Book Number-13: 978-1-4106-1109-3 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. Reason-
able efforts have been made to publish reliable data and information, but the author and publisher
cannot assume responsibility for the validity of all materials or the consequences of their use. The
authors and publishers have attempted to trace the copyright holders of all material reproduced in
this publication and apologize to copyright holders if permission to publish in this form has not
been obtained. If any copyright material has not been acknowledged please write and let us know so
we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.
copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc.
(CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organiza-
tion that provides licenses and registration for a variety of users. For organizations that have been
granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents in Brief

Series Foreword—Gavriel Salvendy xiii

Foreword to Book—Jakob Nielsen xv

Preface xix

PART 1: INTRODUCTION
AND CULTURAL CONSIDERATIONS

1. Overview: Where to Start and What to Consider 3

Nuray Aykin

2. Practical Issues and Guidelines 21

for International Information Display
Nuray Aykin, Allen E. Milewski

3. User Interface Design and Culture 51

Aaron Marcus

4. Synthesizing the Literature on Cultural Values 79

Emilie W. Gould

5. Managing Multicultural Content in the Global 123

Enterprise
Jorden Woods

v
vi CONTENTS IN BRIEF

PART 2: DESIGN ISSUES AND USABILITY ENGINEERING

6. Graphics: The Not Quite Universal Language 157

William Horton

7. “Sunday in Shanghai, Monday in Madrid?!”: 189

Key Issues and Decisions in Planning International
User Studies
Susan M. Dray, David A. Siegel

8. Cost-Justifying Usability Engineering 213

for Cross-Cultural User Interface Design
Deborah J. Mayhew, Randolph G. Bias

PART 3: CASE STUDIES

9. Cross-Cultural Design for Children in a Cyber Setting 253

Jorian Clarke

10. Intercultural Human–Machine Systems: Empirical 277

Study of User Requirements in Mainland China
Kerstin Röse

11. Travel Planning on the Web: A Cross-Cultural 313

Case Study
Helmut Degen, Kem-Laurin Lubin, Sonja Pedell, Zheng Ji

Appendix 345

Author Index 351

Subject Index 357

Contents

Series Foreword—Gavriel Salvendy xiii

Foreword to Book—Jakob Nielsen xv

Preface xix

PART 1: INTRODUCTION
AND CULTURAL CONSIDERATIONS

1. Overview: Where to Start and What to Consider 3

Nuray Aykin
How and where to start? 3
Why internationalization and localization? 4
Terminology 4
The Globalization process 5
Globalization teams: A need for change
in the organization 7
Internationalization and localization team roles 8
Legal issues 15
Summary 19
References 19

2. Practical Issues and Guidelines 21

for International Information Display
Nuray Aykin, Allen E. Milewski
International information display 21
Graphics and icons 22

vii
viii CONTENTS

Language: Rendering and translating 23

Data object formatting 34
Color 46
Layout 49
Summary 49
References 49

3. User Interface Design and Culture 51

Aaron Marcus
Introduction 51
Mapping dimensions to components 58
Future study of culture in relation to user interface
design 64
Conclusions 67
Acknowledgments 68
References 68

4. Synthesizing the Literature on Cultural Values 79

Emilie W. Gould
Introduction 79
The advantages of intercultural communication
theories for HCI design 80
Edward T. Hall: Chronemics, proxemics,
and context 83
Edward C. Stewart and Milton J. Bennett:
Subjective and objective culture 86
John C. Condon and Fathi S. Yousef: Extending
the value orientations approach 89
Geert Hofstede: The durability of the
“Software of the Mind” 91
Harry C. Triandis: Adding culture
to social psychology 95
William B. Gudykunst: Building a unified theory
of communication 98
Fons Trompenaars: Culture hits the best-seller list 103
David Victor: Capturing intercultural
communication in a mnemonic 111
Summary 114
Future directions 115
Conclusion 117
References 117
CONTENTS ix

5. Managing Multicultural Content in the Global 123

Enterprise
Jorden Woods
Introduction and the ideal enterprise 123
Challenges to successful globalization 127
Best practices 131
The details of implementing best practices 136
Workflow and process automation 146
Linguistic technologies 151
Translation service providers 153
Overview and conclusions 154

PART 2: DESIGN ISSUES AND USABILITY ENGINEERING

6. Graphics: The Not Quite Universal Language 157

William Horton
Introduction 157
Why use graphics if they are not universal? 158
Select an approach 158
First, globalize 160
Second, localize 178
Summary 184
References 186

7. “Sunday in Shanghai, Monday in Madrid?!”: 189

Key Issues and Decisions in Planning International
User Studies
Susan M. Dray, David A. Siegel
Introduction 189
Selecting where to go 191
Arranging for research resources in-country 196
Modifying a recruiting strategy and screener
to reflect local cultural issues 200
Preparing or adapting the protocol and/or test
plan 204
Arranging for translation 206
Training local facilitators and preparing
facilities 208
Multicountry studies: serial or parallel? 211
Conclusions 212
References 212
x CONTENTS

8. Cost-Justifying Usability Engineering 213

for Cross-Cultural User Interface Design
Deborah J. Mayhew, Randolph G. Bias
Introduction 213
Framework for cost justification 220
Sample cost-benefit analyses for cross-cultural
usability engineering projects 223
Summary and conclusions 245
Acknowledgments 248
References 248

PART 3: CASE STUDIES

9. Cross-Cultural Design for Children in a Cyber Setting 253

Jorian Clarke
Introduction 253
Positive indicators of children’s usage
of new media 255
Issues surrounding cyber design 259
Case Study: KidsCom.com development
experiences 264
References 276

10. Intercultural Human–Machine Systems: Empirical 277

Study of User Requirements in Mainland China
Kerstin Röse
Introduction 277
Approach to intercultural human–machine system
design 278
INTOPS-2: Intercultural human–machine system
design for the Chinese market 282
Selected results of the investigation in China 292
Conclusions on human–machine system design
for the Chinese market 307
Acknowledgments 310
References 310

11. Travel Planning on the Web: A Cross-Cultural 313

Case Study
Helmut Degen, Kem-Laurin Lubin, Sonja Pedell, Ji Zheng
Introduction 313
Applied methods 317
CONTENTS xi

Results 320
Interpretation of cross-country results 336
Summary and conclusions 342
Acknowledgments 343
References 343

Appendix 345

Author Index 351

Subject Index 357

Series Foreword

With the rapid introduction of highly sophisticated computers, (tele)com-

munication, service, and manufacturing systems, a major shift has occurred
in the way people use technology and work with it. The objective of this
book series on Human Factors and Ergonomics is to provide researchers
and practitioners a platform where important issues related to these
changes can be discussed, and methods and recommendations can be pre-
sented for ensuring that emerging technologies provide increased produc-
tivity, quality, satisfaction, safety, and health in the new workplace and in
the Information Society.
The present volume is published at a very opportune time when industry
and commerce is internationalized and products and services are marketed
across national and cultural boundaries. This book provides a theoretical
foundation and practical examples and guidelines for designing information
technology for different cultures, languages, and economic standing and the
methods to be utilized to ensure their usability across cultural boundaries.
The book’s 11 chapters are authored by some of the foremost practitioners
from the United States, Europe, and Asia. The book starts with two chapters on
guidelines and practices for internationalization and localization, and includes
four chapters on cultural consideration and guidelines in the design of infor-
mation technology; two chapters on usability evaluation methodology and
cost-benefit analysis for cross-cultural design; and three chapters on case stud-
ies of cross-cultural design. The book’s 368 references, 62 figures, and 45 ta-
bles provide useful modes for further in-depth study of the subject.
The book’s editor is a leading international figure in the subject area.
The book should be of special value to people designing and evaluating in-
formation technology products and services for international use and for
researchers in the subject area.

—Gavriel Salvendy
Series Editor

xiii
Foreword

You are probably wondering about how to better serve the international
users of your product, Web site, or intranet. Why else would you have
bought this book? :-) Simply being aware that there is an issue is a great
first step. A lot of designers and project managers have no clue that people
in other countries are different and need special attention in a user inter-
face design project.
I have three simple messages:

1. International usability is not small potatoes; it has a major impact

on your success.
2. Truly serving international users well is a deep design problem that
requires serious work.
3. Despite message #2, you should not give up if you only have a small
budget to devote to international usability.

First, what’s the magnitude of the problem? In one project,1 we tested

20 American e-commerce sites with both American and European users.
The users’ ability to successfully shop on the sites was 61% on the average
for American users and only 47% for European users. In other words, these
sites would be able to increase their overseas sales by almost 33% if only they
could deliver the same usability to international customers as to domestic
customers. Averaged across several studies we have conducted, measured
usability was 46% higher for domestic users than for international users.
Second, why is international usability a deep design problem that re-
lates to more than simply translation? Because a user interface is acted on,
not simply read. In one of our projects, we tested the Web site for a big Euro-
pean bank in a variety of countries, including the United States and several
European countries. The Web site was available in several languages, in-
1
For more information on this project and the detailed guidelines that resulted from it,
please see www.nngroup.com/reports/ecommerce.

xv
xvi FOREWORD

cluding English, and the translation was of excellent quality (contrary to the
sloppy translation jobs we often see). Still, the American users had great dif-
ficulties using the site, because of a variety of cultural differences between
the United States and Europe. For example, the American users could not
use the feature for choosing their language because it was presented in a
way that made sense to Europeans but not to Americans who are used to al-
ways seeing everything in a single language. The English version of the site
was great in terms of language, but the user interface elements didn’t work
for American users.
In another example, we tested a customer service site for an American
technology product in Korea. Even though the interface was in Korean, the
Korean users had difficulty navigating the site because they had difficulty
understanding the information architecture (i.e., the way the information
was structured).
In a user interface, it’s not enough to understand the words. You have to
understand the features and you have to be able to navigate the information
space. Web sites and intranets are particularly vulnerable to difficulties in-
troduced by poor international usability: One wrong click and you are in
the wrong area of the site, but you may never discover that.
The basic tenet of usability is that good design depends on the users and
their tasks. People in different countries will often approach similar tasks in
different ways because of cultural differences. Furthermore, the users are
certainly different. When a user interface is being used in a different coun-
try, it’s really a different user interface, whether or not it has been trans-
lated. With different users and different tasks come different usability
requirements for the design. That’s why international usability doesn’t
come simply from wishing for it. Catering to international users requires in-
ternational usability studies: You need to modify your design to handle
those usability problems that have been discovered by testing users in each
of your important target countries.
Third, isn’t this impossible? I just said to “test users in each important
target country,” but that would cost a fortune for any company that tries to
target more than two or three countries. Yes, truly ideal international us-
ability will indeed be expensive. There are hundreds of countries in the
world, and each will have its own considerations. Unfortunately it’s also sev-
eral times as expensive to conduct international user testing as it is to test
with local users in your own country.
Despite these facts, don’t despair. It is possible to conduct most usability
activities domestically, where user research is cheap, and supplement with a
smaller number of smaller studies in a few countries, sampling the main re-
gions of the world. At a minimum, make sure that you have test data from
the Big Three: United States, Japan, and Germany, all of which are very dif-
ferent. This strategy will not be ideal, and I don’t think it will be considered
FOREWORD xvii

acceptable 20 years from now, but for now it works. You can increase usabil-
ity substantially for your most important customers in the most important
countries through a fairly small effort. They will not get the same quality as
your domestic customers, but they will get much better quality than if you
hadn’t done any international testing. Even customers in smaller countries
where you do not test will still benefit somewhat from your increased aware-
ness of the needs of users outside your own country.
Any international usability efforts are better than none. If you only have a
small budget for international usability, you may not be able to close the en-
tire 46% quality gap between the domestic user experience and the interna-
tional one. But if you can just cut the problem in half, you will now be
serving your international customers 23% better than you did in the past.
That’s surely worth a small investment, considering the market potential of
the rest of the world.

—Jakob Nielsen
www.useit.com
Preface

On entering a country, ask what is forbidden;

on entering a village, ask what are the customs;
on entering a private house, ask what should not be mentioned.
—Chinese Proverb

If this book were published about 10 or even 5 years ago, I would be spend-
ing a lot of time explaining why internationalization is so important. And I
would need to provide all those statistics about the trends on how compa-
nies are going global, how e-business is shaping up global commerce, and
how the number of multilingual sites is increasing rapidly. It is not that I do
not believe in providing statistics to show the readers the importance of
globalization in design. There is no need now to provide all those statistics
to convince people about the importance of being part of the global market
trends. It is happening now and it is everywhere. Not designing with lan-
guage and cultural differences in mind is no longer an option that compa-
nies can afford. The second reason is that the statistics reflect the moment
they are referred to and by the time this book is published, any statistics
shown here will be almost obsolete given the fast trends in the Internet in-
dustry. However, there is one clear fact that remains the same: Only 8% to
10% of the world’s population speaks English as their primary language.
Even in the United States, there are Spanish and Japanese TV stations that
address the needs of the high percentage of non-English-speaking ethnic
groups. In cases where two countries speak the same language, a design for
one may not work well in the other. Designers often face questions such as,
“My customers read English, so why should I localize (or should we spell it
localise) my product?” The answer is easy. Just superficially looking at the
differences in the language and time/date formatting, we can clearly see is-
sues that would lead to confusion. For example, the United Kingdom uses

xix
xx PREFACE

the 24-hour clock, whereas the United States uses a 12-hour clock; centre in
the United Kingdom is spelled center in U.S. English; and the toilet in the
United Kingdom means bathroom in the United States.
Today, more and more web sites are providing content in multiple lan-
guages, and more and more products are being designed with cultural dif-
ferences in mind. The concept of cross-cultural design, however, has not yet
become a priority issue on the practitioners’ and educators’ agenda. There
is a lot of work that still needs to be done to make internationalization a
standard part of the design process that lead to successful product launches
in different parts of the world.
I remember a full page advertisement that the Japan Automobile Associa-
tion placed in the New York Times in 1995. They listed reasons why American
cars were not selling in Japan. The list included a number of bloopers that
were not thought through clearly by the U.S. automobile manufacturers: (a)
the size of the cars was not practical, (b) the steering wheel was on the wrong
side, (c) seat sizes and adjustments did not fit the average Japanese, (d) en-
gines weren’t designed to run on Japan’s lower octane fuel and thus per-
formed poorly, (e) the letters R, D, and L on the shift had no meaning for
Japanese drivers, and (f) repairs required hard-to-find English-based tools.
Everyone has his or her favorite examples related to translations. In
some cases, bloopers even occur when a company uses its native language.
For example, in 1997, Reebok named their new sneaker line Incubus, which
means “an evil spirit supposed to descend upon and have sexual inter-
course with women as they sleep.” When the gaffe was disclosed, they had to
trash all packaging and advertising. The moral is: We need to do our home-
work very carefully, even when we think we know our language well.
Another great blooper is the use of flags to represent languages on
Web sites. Flags represent countries, not languages. In some cases, a lan-
guage is shared by many countries. For example, English, although
there are differences, is the language in Great Britain, the United States,
Australia, and New Zealand. In other cases, some countries have multi-
ple languages. For example, the official languages in Belgium are Dutch,
French, and German. There are also cultural and political issues regard-
ing how a nation treats its flag, and we should be very careful about how
we use flags in our designs.
In using examples, however, we need to be careful about not to fall into
the urban legends. We see examples of bloopers related to pitfalls in doing
business in other countries. One of the most widely told stories is about the
Chevy Nova. According to sources, the car did not sell well in Span-
ish-speaking countries because its name translates as “doesn’t go” in Span-
ish. This, however, is actually an urban legend. The car did sell well in
Spanish-speaking countries. “Nova” in Spanish means “star” as in English.
PREFACE xxi

The phrase “doesn’t go” is no va in Spanish and has nothing to do with the
word Nova. There is even a government-owned oil company in Mexico,
Pemex, that sells gasoline under the name of Nova.
Although many designers and developers do tackle various issues on in-
ternationalization, the main questions remain the same: What does it take to
design products and services, including software applications and Internet
sites, for multiple markets? How much will it cost? Can we justify the cost?
As the Internet has moved toward ubiquity, the need for international-
ization is apparent. Initially in the United States, designers felt no immi-
nent need to make sites available in multiple languages. The rationalization
was that “the world knows English.” Perhaps this is no longer true. Why else
would Microsoft spend millions of dollars to release Microsoft Office XP in
more than 40 languages?
While several books have been written since the late 1980s on interna-
tionalization and localization, this remains an ever-expanding field and
there is no book that covers it all. Some emphasize the developers’ and soft-
ware engineers’ perspective, some emphasize management’s perspective,
and some emphasize the designers’ perspective.
Our goal, as contributors to this book, is to provide designers with con-
crete tools and processes—design practices, guidelines, case studies, and
lessons learned. Further, we hope to provide you with extensive resources
to aid you in your work. This book does not attempt to cover all design as-
pects of internationalization and localization, and it is definitely not in-
tended to be a software internationalization reference book. We list those
books, along with the design-related ones, in the appendix section at the
end. As designers, we need to be intimately familiar with the software inter-
nationalization terminology and processes, but we should also emphasize
the cultural studies that really help us understand cultures in a much more
detailed level than just learning to choose the right colors from tables show-
ing color–culture pairs. And it is not just using the right data formatting
(e.g., date and time) provided in software tools either. We all believe that
this is a vast and wonderful area to explore, and we invite our readers to
share their experiences and learnings with others. The more we learn, the
more we will make a difference in design practices.
This book is organized into three sections. Part I includes four chapters.
Chapter 1, by Aykin, approaches the topic from a management perspective,
defining roles and responsibilities. Chapter 2, by Aykin and Milewski, cov-
ers the practical issues and guidelines regarding international information
displays, what we call the tip of the iceberg of internationalization and local-
ization. Chapters 3, 4, and 5 go deeper into the lower layers of the iceberg
and talk about the cultural dimensions and the impact of culture on design.
Chapter 3, by Marcus, looks in depth at Hofstede’s study on cultural dimen-
xxii PREFACE

sions, while chapter 4, by Gould, describes the studies looking at and theo-
ries regarding cultural differences. In chapter 5, Woods describes how to
manage multilingual content in the global enterprise setting, as well as
methodologies for managing the internationalization and localization of
web sites.
Part II focuses on some specific aspects of design, cost justification, and
international usability evaluations. In chapter 6, Horton provides a vast ar-
ray of information on global icon design and includes lots of examples and
do’s and don’ts that are very useful for designers. Dray and Siegel put all
their experiences, along with their knowledge regarding international user
studies, into chapter 7. They provide not only a good perspective on con-
ducting international user studies, but also a lot of practical tips for the
practitioner. In chapter 8, Mayhew and Bias apply their expertise in
cost-justifying usability to cross-cultural user interface design, which costs
much more than designing for a single target market.
Part III covers three case studies on international user interface design.
In chapter 9, Clarke describes the differences in web site design for children
in different countries. Chapter 10, by Röse, looks into various design issues
for China. In chapter 11, Degen, Pedell, Lubin, and Zheng discuss design-
ing a travel reservations site for three different countries: Germany, the
United States, and China.
At the end of this book is a list of books and magazines on the topic of inter-
nationalization in user interface design, project management, and software
design and development. Other valuable resources, leading journals, and
nonprofit organizations that deal with internationalization are also listed.
I regret not being able to include more topics on internationalization
and localization from the design perspective. However, I hope this book
will become one of those dog-eared, often used volumes on your bookshelf
and remind you that the world is an amazing place where we learn some-
thing new every day. As the Dalai Lama’s Instructions for Life says: “Once a
year, go someplace you’ve never been before.”
And I hope this book will show that the internationalization and localiza-
tion do not merely mean translation of content, or redesigning and devel-
oping an entire new product every time a new market is identified, or just
expecting your local offices to handle the adaptation of your product to
their markets. It is much more than that, and we hope we have touched
many aspects of it.

ACKNOWLEDGMENTS

I would like to thank Gavriel Salvendy for encouraging me to edit a book on

this topic and providing in-depth comments on content. My thanks also go
to LEA, especially Anne Duffy, Kristin Duch, and Sarah Wahlert for pub-
PREFACE xxiii

lishing this book and working hard with us to make it happen. I have been
grateful for the suggestions and insights I received from my long-time
friend, Diane Z. Lehder, who also shared my early years in evangelizing in-
ternationalization at AT&T.
Many special thanks go to my husband, Al, and my son Bora for giving
me the time and for understanding the sacrifices they made while support-
ing me in this effort.
 PART 1

INTRODUCTION
AND CULTURAL CONSIDERATIONS
 Chapter 

Overview: Where to Start

and What to Consider

Nuray Aykin
Siemens Corporate Research, Princeton, NJ

HOW AND WHERE TO START?

In this book, the contributors discuss many aspects of internationalization

and localization and provide valuable insights regarding cultural differ-
ences, design practices, globalization, content management, and so on.
This chapter is intended to give the reader an overall firm basis for under-
standing why we need to localize our products to target cultures, and how to
begin doing so. In addition to summarizing key points made by all the au-
thors throughout this book, it also addresses the roles and responsibilities
within an organization dealing with internationalization and localization,
and provides an action checklist along with practical “how to” guidelines.
The content provided in this chapter is by no means complete. The
globalization of a product has many facets and it is almost impossible to
cover them all here. However, this chapter, and also this book, is a good
starting point when your company decides to move into global markets
with the goals of making products suitable, usable, and preferred by the
users in the target locale.

3
4 AYKIN

WHY INTERNATIONALIZATION AND LOCALIZATION?

The world is a diverse place. The environment in which software applica-

tions are designed often is not. It takes a Herculean effort to keep a design
team focused on what is required to successfully design products for global
markets. But, the change is here, the challenge is real, and we need to face
it. The facts about differences in languages and cultures are spread
throughout many resources. Consider these facts:

• Only 8% to 10% of the world’s population of 6.1 billion speaks Eng-

lish as its primary language.
• In some cities in the United States, large populations speak a lan-
guage other than English as their native language: 78% in Miami,
49% in Los Angeles, 45% in San Francisco, and 42% in New York
City (25% of subway users in New York City do not speak any Eng-
lish at all). Eighty percent of European-based corporate sites are
multilingual (Dunlop, 1999). Globally, there are 580 million peo-
ple with Internet access as compared to 563 million in the last quar-
ter of 2002: In the United States, 29% of the population has
Internet access. This compares to 23% in Europe, 13% in Asia-Pa-
cific, and 2% in Latin America (CIO, 2003).
• U.S.-based Palm has a 68% share of the Latin American market
(CIO, 2002a).
• The United States is responsible for shipping 38.8% of personal
computers worldwide, followed by Europe with 25%, and Asia-Pa-
cific with 12%. The total personal computers shipped worldwide
reached 1 billion in 2002 (CIO, 2002b).

Whether to expand to global markets and localize your product for target
locales is your company’s decision, and it should be based on business needs,
user needs, and technology considerations. But if the decision is to go global,
your company will need to define new roles and processes if it is to carry out
the internationalization process successfully. In this chapter, I touch on these
issues. Additionally, I cover some legal issues you need to consider when ex-
panding into global markets, especially in designing global Web sites.

TERMINOLOGY

As defined by the Localization Industry Standards Association (LISA) in The

Localization Industry Primer (LISA, 2001), here are definitions of the basic
terminology used throughout this book.
Globalization: The general process of worldwide economic, political,
technological, and social integration. More specifically, the process of mak-
1. OVERVIEW 5

ing all the necessary technical, managerial, personnel, marketing, and

other enterprise decisions necessary to facilitate localization. Globalization
addresses the business issues associated with taking a product global. It in-
volves integrating localization throughout a company including product
design and development, as well as marketing, sales, and support in the
world market.
Internationalization (I18N): The process of ensuring, at a technical/design
level, that a product can be easily localized. Internationalization is thus part
of globalization. Internationalization helps define the core content and
processes so that they can be easily modified for localization to specific mar-
kets. To abbreviate the lengthy word Internationalization, the term I18N was
introduced in the early 1990s. The I and N represent the first and last let-
ters of the word Internationalization, and 18 represents the number of letters
between I and N.
Localization (L10N): The process of modifying products or services to ac-
commodate differences in distinct markets. The term L10N was derived in
the same way as I18N. Localization makes products or services usable in,
and therefore acceptable by, target cultures.
There are two other terms that are related to the field and that are en-
countered often:
Translation: The linguistic component of localization. It is choosing the
appropriate text in a target language to convey the proper meaning from a
source language text.
Locale: That part of the user’s environment that is dependent on lan-
guage, country/region, and cultural conventions. For example, Argentina,
Spain, and Mexico represent three different locales even though their com-
mon language is Spanish.

THE GLOBALIZATION PROCESS

Because globalization involves “the process of making all the necessary

technical, managerial, personnel, marketing, and other enterprise deci-
sions necessary to facilitate localization” (LISA, 2001, p. 34), it is important
to understand where to start.
The first question a company must address is whether it makes sense to
expand into global markets. If the answer is yes, then the company must de-
fine global business strategy, which includes identifying target markets, and
how to enter them, and restructuring the organization to facilitate global-
ization efforts. These initial steps are beyond the scope of this book. There
are excellent books and references available on this topic; see, for example,
O’Hara-Devereaux and Johansen (1994), Rhinesmith (1993), and
DePalma (2002). Instead, this book concentrates on the internationaliza-
tion and localization parts of this process, which include (a) ensuring that
6 AYKIN

your product and web site are designed and developed free of any cul-
ture-specific attributes, so they can be easily localized (Internationaliza-
tion), and (b) making the necessary resource files and design changes so
your products and web site are culturally and technically suited for the tar-
get culture (Localization).
The internationalization and localization processes follow the same basic
steps. First, define the core product or web site that will be localized for the
target markets. This involves defining the software architecture, common
software, and user interface elements that will constitute the core, regard-
less of the target locale. The resource files that define the target locale’s spe-
cific elements, as well as the common global user interface design that
accommodates different languages, text expansion, navigation structures,
and different content within a corporate defined look and feel (without sac-
rificing brand) are all considered in this step. If there is no existing product,
you can define the core product to be free of any cultural elements. It will
then be easier to localize the components for different cultures.
For example, assume date and time will be displayed on a web site. If
the software developers use separable resource variables for date and
time rather than the fixed U.S. date and time in the code, it will be easier
to retrieve the correct date and time representation from the resource
files for a particular target locale and place them in the software when
needed, even during runtime. If the user interface designers can leave
enough space on the screen to accommodate different lengths and for-
mats in displaying date and time, the insertion can be accomplished
without changing the layout design. In most cases, however, the prod-
ucts are designed for one country at the beginning, and the need to glob-
alize comes later. In this case, the product must be reengineered, which
includes conducting a code inspection to make the software internation-
alized. For example, if the date and time code is fixed in the core prod-
uct, it would be replaced by resource variables so that resource files for a
particular target locale could be inserted to provide the correct date and
time on the screen.
Second, localize the internationalized core product for different target
locales. Using the core product and core user interface design guidelines,
software designers and developers can add the locale-specific formats and
content to the design. In this way, the usage of common elements is opti-
mized. Internationalizing the product first facilitates the localization of the
product and makes it more likely that the final product will meet your cus-
tomers’ needs and expectations.
Creating successfully localized products requires a tightly integrated
corporate team, and successful integration is dependent on clearly de-
fined roles, responsibilities, and tasks. These are discussed in the remain-
der of this chapter.
1. OVERVIEW 7

GLOBALIZATION TEAMS: A NEED FOR CHANGE

IN THE ORGANIZATION

If your company is moving into global markets and your products need lo-
calization, then it is time to start building the corporate infrastructure to
carry out the necessary tasks. When a company goes global, all units within
the company need to understand the meaning and scope of going global
and the impact on their own roles and responsibilities. It is definitely not
merely having your product manuals or your web site content translated
into a target language, without any changes. While this is a common strat-
egy, it really is not the way to meet the needs and expectations of customers
“over there” (a place we sometimes know little about). Starting from the
market analysis and continuing until the localized versions of the product
are out the door there is a lot of work to do. This work does not end with
maintenance and the definition of future releases; it continues as long as we
operate in outside markets.
In this section, I talk about the roles in the organization and the changes
needed to accommodate the globalization process. The following teams in
an organization carry out specific roles in the internationalization and lo-
calization processes. Of course, it is possible that some of these teams can be
combined under one umbrella. However, the roles and the work that needs
to be done should not be sacrificed to save time and money. If it is not done
right at the beginning, getting it right later will be far more costly.

• Management Team: Without buy-in from the management team

and without a clear understanding of what it takes to international-
ize and localize a product, it is very difficult to get the synergy and
enthusiasm along with a commitment from the other teams. The
management team includes all top-level managers, including
country managers if you have a global presence. It is very important
to identify a champion who is also accountable for the company’s
globalization efforts. This champion will ensure the globalization
process is overseen properly and will serve as the globalization ad-
vocate within the company.
• Marketing Team: The marketing department can easily play the
lead role in globalization efforts. They understand and select target
countries, define business needs, conduct research/studies in the
target cultures to understand needs and preferences, perform
branding analyses, validate globalization decisions by defining the
return on investment (ROI) from moving into global targets, help
the sales team to form distribution channels or partnerships, and
work with the legal department to understand the target country’s
standards, legal issues, customs, and business conduct.
8 AYKIN

• Implementation (Internationalization/Localization) Team: This is

the core team that will work on day-to-day activities as well as the
overall project management. While actual team size depends on
the complexity and the magnitude of work involved, each imple-
mentation team should include:
• Project managers—who can manage the entire process of inter-
nationalization and localization.
• User interface designers—who can design user interfaces, and
know how to study the needs and preferences in the target cul-
tures. Their work requires an understanding of languages spo-
ken, cultural preferences, interests in the types of products and
services you offer, use of colors and icons, work practices, ethics
and relationships, daily life, learning styles, religion, gender
roles, purchasing habits, political influences, customs, and so
forth. This is the group that plays a crucial role in defining what
the final product looks and feels like when it is finally shipped to
the target country.
• Software engineers and developers—who can (a) define the soft-
ware architecture based on the target culture’s technological ad-
vances or limitations, interoperability with other systems, and
infrastructure; and (b) develop software that is internationalized
and ready to be localized to different target markets. This group
must understand how to internationalize a product from the soft-
ware design perspective.
• Technical writers and translators—who can (a) write documents
with the international audience in mind, so that the text is easily
translated into different languages; and (b) translate documents
and customize them for the target languages. Translation is a
major expense in the localization process. Creating the right
documents up front can save time and money in the translation
process. Following guidelines on writing text for translation (see
Aykin & Milewski, chap. 2, this volume, for more details), identi-
fying clearly what is translated and what is not, and creating a
translation process are key elements in translation.

The following section provides high-level guidelines for each of these

teams to carry out a smooth internationalization and localization process.

INTERNATIONALIZATION AND LOCALIZATION TEAM ROLES

The three teams (management, marketing, and internationalization/lo-

calization implementation) should be able to define strategies and pro-
1. OVERVIEW 9

cesses together to ensure a smooth flow of information among teams. In

this section, I do not cover the roles of management and marketing teams
but focus instead on the implementation team, beginning with the key
role of the project manager. The project manager is the glue for the entire
process and is responsible for getting the localized product out to the tar-
get culture on time.

Role of the Project Manager

As the person managing the internationalization and localization process,

the project manager has to understand that localization is not just the trans-
lation of the documentation and the user interface content. He or she has to
help the team understand the extent of internationalization and localiza-
tion and be a strong advocate for this work. The project manager is respon-
sible for the following:

• A Well-Defined Strategy. Working closely with the management

and marketing teams, the project manager must define the prod-
uct’s globalization strategy and localization process. Without a solid
strategy, it is not possible to decide whether to go global or not,
what level of globalization to aim for, what the target markets are,
time to deliver, and what changes are required in the organization
structure.
• A Statement of Work. This includes localization methodology, lo-
calization vendor information if work is to be outsourced, and a
high-level project schedule which includes milestones and budget.
The localization vendor can be responsible only for translation, or
can include software localization. Adding languages can affect the
processes in terms of scaling and re-engineering costs.
• Vendor Management. In most cases, the companies use outside ven-
dors to handle translation. In this case, it is absolutely critical to choose
the right vendor, as this work is a key component of the international-
ization process. There is a need to have a clear project structure and
guidelines for the translators to avoid “garbage in—garbage out.” For
a project manager, this responsibility includes:
• Evaluating and qualifying vendors.
• Establishing corporate pricing structures for translation and lo-
calization.
• Developing long-term partnerships with vendors.
• Setting clear guidelines on intellectual property and other tools
or guides that are generated for the project.
• Communication. Without good communication channels, it is dif-
ficult to coordinate activities, reach milestones, and keep budget
10 AYKIN

and project schedules under control. The communication should

be established to ensure a continuous interaction with the entire
team. A few suggestions include:
• Scheduled meetings. Nothing beats the face-to-face project meet-
ings where project issues are discussed, goals are communicated,
and proactive measures are taken to ensure milestones are met. De-
pending on the geographic locations of the team members, face-
to-face meetings can be replaced with video- or teleconferencing.
However, the project manager still needs to go to the other loca-
tions to meet with the team members face-to-face. This increases
the team’s cohesiveness and productivity.
• Online. Build a strong communication infrastructure for teams
to share work. A good way to enhance communication and
share resources is via company intranets, separate shared links
where all documentation and resources can be posted, shared,
and discussed.
• Content Management. Manage international corporate content
centrally, and leverage the expertise of locals to generate lo-
cale-specific content. Communicate with the locale experts re-
garding new content, changes on the site, branding requirements,
user interface guidelines, and so forth. What works for localiza-
tion of software applications often isn’t good enough for web con-
tent. Simply translating the words from English is not a good fit
for pre-sales and marketing content typically found on web pages.
Overall content needs to be reworked for different markets and
must “feel right” for a given language. It should be conversational
with appropriate style and local expressions. The portions of the
content that change when targeted to a specific locale need to be
identified. This is achieved by understanding the users of that lo-
cale and specifying what works and what the users need. If not
done, the customers from the target culture can easily detect con-
tent that has been “just translated” rather than created specifically
for them.
• A Common Glossary. Developing a central glossary management
system prevents using multiple words for the same thing. This fa-
cilitates translation, reduces translation cost, and improves the
quality of translation. In developing web sites for different target
countries, the project manager can choose to make regional teams
responsible for maintaining and updating the sites, given that the
content and the layout are reviewed by the central office for con-
formation to company standards including complying with the
common glossary.
1. OVERVIEW 11

• Localization Kit. This includes all pertinent information that needs

to be communicated to the localization vendor:
• Product overview
• Target audience
• Contact information for the localization managers, engineers,
developers, and user interface designers
• Project description
• Project scope and metrics
• Detailed project schedule
• Specific localization instructions (source and target languages,
translation instructions [what and what not to translate], test-
ing instructions [browsers, operating systems], formatting
guidelines, etc.)
• Tools to be used by the localization vendor
• Directories and files required to build localized versions, and
testing tools.
• A Repository of User Interface and Software Elements. It is ex-
tremely valuable if there is a central repository that includes com-
pany branding tools and guides, all internationalized and localized
graphics and icons, reusable code, templates, et cetera.
• An International Quality Assurance Methodology. To ensure de-
livery of quality products, quality assurance should be carried out
throughout the product development and deployment process.
This includes defining metrics for quality and methods to carry
out quality assurance, including in-country reviews and testing.
• Maintenance of Internationalization and Localization Guidelines.
These guidelines are created with input from the internationaliza-
tion and localization team. They can include international and
country-specific user interface design guidelines and software in-
ternationalization guidelines.
• A Post- Localization Issues List. This list may include deploying and
maintaining updates, analyzing user feedback, and refining pro-
cesses. For example, managing updates in the content can be a
nightmare if a structure is not placed around it. This involves iden-
tifying changes, translating only the changes, and incorporating
them back into the content.
• Legal Issues. Working with marketing and legal teams to identify
any legal limitations, international and national laws, tariffs and
customs, tax laws, and export/import structure is crucial to ensur-
ing smooth entry into the target markets. There is a separate sec-
tion at the end of this chapter that provides high-level guidelines
regarding legal issues in internationalization.
12 AYKIN

Role of the User Interface Designer

A solid background in user interface design is essential here as in any design

work, and gathering user requirements before anything else is done must
be a top priority. It does not matter what our target culture is, a strong re-
quirements-gathering process can help prevent future problems when go-
ing global. We have heard embarrassing anecdotes related to the
internationalization efforts of companies. But, we definitely would not want
them to happen to us. And more important, we don’t want to lose sales be-
cause we did not meet the customer needs. Dealing effectively with interna-
tionalization and localization process and design issues demands more
than just a solid background in user interface design. There are, unfortu-
nately, still user interface designers who do not understand why we are lo-
calizing products. These designers may even believe that people from other
countries are used to the “American” way and that they like to be “Ameri-
canized.” This may be true in part: We see kids wearing Hard Rock Café
T-shirts in a rural town in Tibet, and we see kids playing Pokeman games
(from Japan) in Greece. But what we don’t see is kids playing American foot-
ball in Tibet and people waiting excitingly for the baseball season to start in
Greece. It is true that the boundaries are getting grayer every day, but cul-
tural differences do exist and are a key part of usability that designers need
to confront. The topic of designing for different cultures is becoming more
and more visible at conferences and in publications. Learning from others
and reading about cultural differences is a good start. The real understand-
ing comes from going “over there” and observing and collecting data. If
that is too expensive, working with in-country people to collect data for you
and educate you about their culture regarding design issues will also work.
Here are a few items to consider while designing for different cultures:

• Design with internationalization in mind - If your company has

plans to expand its customer base globally, it is important to con-
sider design characteristics that would be acceptable for most coun-
tries. It starts from branding your products to developing icons that
are acceptable and understandable worldwide. The more design
decisions that are truly global, the fewer will have to be localized for
individual markets.
• Icon design: A simple well-known example of a global icon is
“mail/e-mail.” Instead of having to accommodate different mail-
boxes/slots around the world, one common icon, “an envelope,”
solves this design problem. And you do not need to localize it for
different cultures. Other examples of commonly accepted and
understood icons include books, cameras, movie projectors,
clocks, eye glasses, microphones, planes, trains, scissors, no
1. OVERVIEW 13

smoking signs, and so on. Horton (chap. 6, this volume) de-

scribes the rules on icon design in detail.
• Religious references: Unless your site is related directly to reli-
gion, it is best to avoid religious references. This way, you do not
need to worry about who visits your site and who may be of-
fended by it. It is also important to respect the beliefs and prac-
tices of different religions. For example, showing uncovered
body parts is not appropriate in Islamic countries.
• Culture-specific examples: Referring to culture-specific heroes,
symbols, puns, and idioms either may not translate well into
other cultures or may be misinterpreted—and both may cause
problems. For example, referral to heroes in content is very cul-
ture-specific. Presidents Jefferson and Lincoln may be the great
heroes in American culture, but cannot replace Mandela in
South Africa. Benjamin Franklin is better known as a scientist to
the rest of the world, as compared to the United States, where he
is better known as one of its first political figures. Similarly, yel-
low school buses are the transportation vehicles for children go-
ing to school in the United States. Other countries have different
ways for children to get to school, including walking.
• Cultural preferences: This can include everything from learn-
ing styles to color choices. Whereas some cultures may prefer
to read the manual before operating a device, others may pre-
fer to talk to sales representatives, and still others may prefer
to go with trial and error.
• Data formatting: Choosing the right formatting for the target
cultures is the start of any design for international markets.
Many times, there may not be any common formatting for all
target cultures, and a localized version of the data becomes nec-
essary. In some cases, such as address formats that differ widely
from country to country, the designer can still create a generic
address format to accommodate most of the world address for-
mats. Aykin and Milewski (chap. 2, this volume) cover many of
the data-formatting issues.
• Perform usability evaluations with cultural differences in mind-
Early testing of localized versions of the product is very important
because testing in one country does not guarantee product usability
in other countries. Usability testing uncovers design problems that
may need to be corrected in the early stages of development. Susan
Dray and David Siegel’s chapter on international user studies
(chap. 7, this volume) covers a vast amount of information about
conducting global usability studies in the trenches of the globe.
Anyone who has worked on any global project can tell you stories
14 AYKIN

about how difficult usability studies can be. One thing is clear: You
need to accept that any event can go wrong in your studies. The
more prepared you are, the less chance you will have for surprises
that cause costly delays in the study.

The main principle in global usability testing is to go to the target locale

and run the requirements gathering before the design and testing, and per-
form usability evaluation during and after the design. Of course, this “best
case” scenario is more expensive in the short term, but the return on your
investment will be tremendous.
Here are some guidelines on global usability testing. (See Dray & Siegel,
chap. 7, this volume, for more details.)

• Leverage in-country resources.

• Obtain early and ongoing feedback using a variety of sources of
input (customer visits, expert evaluation, user evaluation of pro-
totypes, user trials, etc.).
• Use well-defined processes and modify them when necessary.
• Translate and adapt materials as needed.
• Develop generous schedules; consider local holidays.
• Expect the unexpected (equipment specifications/failure, re-
cruiting, etc.).
• Interpret behavior within the context of the target culture.
• Run the test with local users and in their native language.

Role of the Software Engineer and Developer

Software engineers and developers are the key people who will be imple-
menting all the internationalization concepts and design issues into the
product. They are indeed lucky if management calls for a product that must
be designed and developed from the beginning with internationalization in
mind. Unfortunately, in most cases, the product is an existing one, de-
signed initially with a single country and culture in mind, one with little or
no flexibility in the software. Before translating anything, before changing
code, and so forth, the software needs to be assessed as to whether it can
handle internationalization or not, and whether it can be reengineered to
be internationalized. It is important that the software be sufficiently flexible
to handle all localization changes for multiple locales. A truly localized
product should include:

• One set of source code.

• A single bug-tracking system.
1. OVERVIEW 15

• Isolated localizable resources, including dialogs, macro language,

status bars, messages, menus, prompts, toolbars, and sounds.
• Localizable elements, such as time, date, currency, address, names,
and so on, that are not hard-coded and should support different
character sets.
• Buffers large enough to handle translated text in cases where the
text expands.
• Character parsing that is not limited to Latin script.
• Code clearly marked showing what must and must not be translated.
• Content that is presentable in users’ language and character set.
• User input that can be received in users’ native language and char-
acter set.

Role of the Technical Writer and Translator

Technical writers create content that may end up being translated to many
languages. In chapter 2 of this volume, Milewski and I cover the general
rules on how to write for translation. Besides these general rules, it is very
important to pay extra attention to:

• Common glossary/dictionary: For an organization, especially going

global, it is very important to create a standard and consistent ter-
minology that is shared by all parties involved in generating con-
tent and user interface elements. Placing this standard and
consistent terminology in a common glossary is the best way to
share among the team members.
• Spell checker: Regardless of the language, a spell checker should be
used consistently throughout content generation to ensure correct
spelling and use of accents in the target language.
• Culture-specific language: Eliminate all culture-specific meta-
phors, acronyms, abbreviations, humor, puns, idioms, slang,
sports, and other culture-specific analogies from the content to
make the content truly internationalized.

A well-written, concise, and culture-free source document serves as a good

starting point for translation. If there is a need to include any culture-specific
referrals, clear communication among all team members will ensure correct
translation and interpretation of these culture-specific examples.

LEGAL ISSUES

Introducing a product into a new country or locale requires more than

adapting it to the country’s cultural and linguistic requirements. It includes
16 AYKIN

dealing with a variety of legal issues, tax laws, duty fees, value-added taxes,
advertisement rules, and warranty policies. Most of these issues are best
handled by experts in international and in-country laws and standards.
Microsoft (2002) describes their introduction of Windows in mainland
China as a good case study for businesses expanding into global markets.
Apparently, Microsoft was not aware of the political issues surrounding de-
veloping the Simplified Chinese edition of Windows offshore and then im-
porting into the People’s Republic of China. Microsoft had to change its
strategy after angering the Chinese government and then had to work very
closely with the Chinese government and local developers to ensure the
success of Windows XP.
Sullivan (2003) reports that the Google search engine has 3.3 billions of
text documents indexed as of September 2, 2003. This figure does not in-
clude Google discussion posts, image and multimedia files. These docu-
ments, discussion posts, and images are from all around the world. NUA
(2002) reports that the estimated number of online users in September
2002 was 605.6 million. Not all online users come from a single region or
country. Out of this total, there are 190.91 million in Europe, 187.24 mil-
lion in Asia/Pacific, 182.67 million in Canada and the United States, 33.35
million in Latin America, 6.31 million in Africa, and 5.12 million in the
Middle East. Considering these statistics, legal issues become more compli-
cated. Given that there are more than 3.3 billion documents posted from all
around the world and 605 million users accessing these documents from all
around the world, whose laws will apply?
Ellis (2001) and Aykin (2001) give examples of how cultural differences
can have legal implications if one is not careful.
The Internet is full of stories about who is suing whom for damages,
copyrights, privacy issues, and so forth. For example, France sued Yahoo
for selling Nazi memorabilia on their auction site, and it was not Yahoo’s
French site that was sued, but its U.S. site. Although the server is located in
the United States, the French people could access the site and see the Nazi
memorabilia being auctioned there (Aykin, 2001). Hongladarom (2000)
lists several guidelines regarding messages on the Internet in Thailand.
The guidelines state that messages critical of the King and his royal family
are absolutely prohibited, and suggest not posting any messages that con-
tain foul language, sexually explicit content, insults, criticism of any reli-
gion, or personal data of any kind.
Finally, the governments of Italy, England, and Scotland have seized
computer equipment of anarchist groups that advocate antigovernment vi-
olence via the Internet (Schönberger & Foster, 1997).
Currently, there is no single solution, and there is no one legal entity to
resolve the legal issues that arise due to the Internet. However, we see some
actions now being taken by e-commerce companies, governments, interna-
1. OVERVIEW 17

tional organizations, and regulatory agencies to bring resolutions to the

conflicts raised by the Internet. The following organizations and regula-
tions deal with issues of this type.

The International Organisation for Economic Co-Operation and Devel-

opment (OECD) (adopted OECD Guidelines for Consumer Protection
in the Context of Electronic Commerce (December 1999) for consumer
protection in online commerce.)
Global Information Infrastructure Commission (GIIC).
American Bar Association.
Hague Convention on Jurisdiction & Foreign Judgments in Civil and
Commercial Matters (HCCH).
E.U. Data Protection Directive.
European Council of Ministers, Brussels 1 regulation.
Safe Harbor Privacy Principles signed by the United States Commerce
Department and the E.U. Commission.

These are just a few organizations that are trying to deal with the interna-
tional issues related to global use of the Internet.
Sinrod (2000) and Ellis (2001) provide a few recommendations on how
to protect an entity and its content on the Internet from the laws of various
countries around the world:

• Domain Names: Under the Anti-Cybersquatting Consumer Protec-

tion Act (1999), rightful owners of trademarks or personal names
may now bring an action against anyone who, with a bad intent to
profit, registers, traffics in, or uses a domain name that is identical
or confusingly similar to a mark that was distinctive or famous when
the domain name was registered. However, this does not prevent
cases of having same domain names in other countries. An example
includes a battle over the use of “Bud” or “Budweiser” between St.
Louis-based Anheuser-Busch and Czech brewer Budvar. The two
companies signed an agreement in 1939 that kept Budvar out of
the United States and Anheuser-Busch’s Budweiser from using that
name in many European countries. In 2000, Budejovicky Budvar
introduced their beer to the United States under the Czechvar
brand (Realbeer, 2001).
• Privacy: Since the explosion of the Internet, privacy laws have
gained a lot of attention due to the fact that people enter their
names and credit cards online, and that information about people
18 AYKIN

can be shared by many online providers. People need to trust the

confidentiality of the process and the security of their personal
data. If your web site collects personal information, there should be
a clear privacy policy regarding the handling of the information
that is collected. The European Union Data Protection Directive
(1995) requires companies that collect information from individu-
als in Europe to disclose how the information is used. It also re-
quires that individuals be allowed to edit or delete this information
(Ellis, 2001). Having a U.S.-based site does not exclude the site
owners from this directive. Additionally, the directive bars transfers
of personal information from E.U. countries to non-E.U. countries
if they do not meet the directive’s standards.
• Arbitration: A corporate site should include provision for arbitra-
tion. There are more than 100 countries that recognize arbitration
as a solution to international disputes.
• Terms of use and legal page: Including a Terms of Use/Legal page that
provides rules about using the site helps avoid and resolve disputes
and misunderstandings.
• Insurance: Businesses can purchase insurance to cover their liability
for claims resulting from online transactions across borders.
• Tax: If a company has a physical presence in a country that has a
value-added tax (VAT), the company’s online transactions, even
with customers outside that country’s borders, may also be subject
to that tax. If the company does not have a presence in the country
in which the customer is located, the online transaction may be sub-
ject to import taxes. There are a lot of resources to get information
on what taxes apply in which countries, including legal entities,
shippers, customs brokers, and freight forwarders.
• Physical location of assets: If a company has a physical presence in a
country, this office is considered to be legally responsible for any is-
sues regarding the company’s web site even if the web site server
may not be located in that country.
• Third-party content: Hosting third-party content or chat lines can
become a headache if the content is not monitored carefully. Any
offensive material published or advertised becomes the host’s li-
ability. For example, eBay was sued by the French government
for auctioning Nazi and other racist items to French citizens
(Morris, 2000).

There are also government regulations on use of languages in many

countries. For example, Quebec in Canada enforces that the language of
any published content should be first in French, and other languages can
only be introduced after the French version is implemented.
1. OVERVIEW 19

SUMMARY

As was mentioned at the beginning of this chapter, it is very hard to explain

all aspects of globalization in one book, let alone in one chapter. This chap-
ter is intended to give a real-world checklist of actions required to success-
fully globalize a product. In addition, it highlights a few key issues. In the
remaining chapters, experts in this field describe the issues in more detail
and provide information regarding how differences across cultures and
languages affect design.

REFERENCES

Aykin, N. (2001). Legal issues reconsidered: Do the country boundaries matter any-
more? In M. J. Smith & G. Salvendy (Eds.), Systems, social and internationalization
design aspects of human–computer interaction (Vol. 2, pp. 548–552). Mahwah, NJ:
Lawrence Erlbaum Associates.
CIO. (2002a). Metrics: PDA Shipments grow 13.8 percent in Latin America. October 1,
2002. Source: Gartner Dataquest. Retrieved January 22, 2004, from
http://www2.cio.com/metrics/2002/metric445.html?CATEGORY=11& NAME
=Globalization
CIO. (2002b). Metrics: One billion PCs shipped, July 2, 2002. source: Gartner
Dataquest. Retrieved January 22, 2004, from http://www2.cio.com/met-
rics/2002/metric397.html?CATEGORY=11&NAME=Globalization
CIO. (2003). Metrics: Internet population reaches 580 million (source: Niel-
sen/Netratings), February 27, 2003. Retrieved January 22, 2004, from
http://www2.cio.com/metrics/2003/metric508.html?CATEGORY=11&NAME=
Globalization
DePalma, D. D. (2002). Business without borders; A strategic guide to global marketing.
New York: Wiley.
Dunlop, B. (1999). Success in any language. March 9, 1999. Retrieved January 22,
2004, from http://www.clickz.com/int_comm/article.php/812791
EU Data Protection Directive. (1995). EU Directive 95/46/EC. The Data Protection Di-
rective. 24 October 1995. Retrieved January 22, 2004, from http://www.
dataprivacy.ie/6aii.htm
Ellis, R. L. (2001). Web site globalization and the law. IQPC Seventh International Con-
ference on Web Site Globalization, San Francisco, March 25–28, 2001.
Hongladarom, S. (2000). Negotiating the global and the local: How Thai culture co-opts
the Internet. Retrieved January 22, 2004. First Monday, http://www.first mon-
day.dk/issues/issue5_8/hongladarom/
Localisation Industry Standards Association. (2001). The localization industry primer.
Retrieved January 22, 2004, from http://www.lisa.org/products/primer.html
Microsoft (Dr. International Team). (2002). Developing international software.
Redmond: WA, Microsoft Press.
Morris, S. (2000). The importance of international laws for web publishers. June 2000. Re-
trieved January 22, 2004, from http://www.gigalaw.com/articles/2000/mor-
ris-2000-06.html
20 AYKIN

O’Hara-Devereaux, M., & Johansen, R. (1994). Global work: Bridging distance, culture
and time. San Francisco: Jossey-Bass.
Sullivan, D. (2003). Search engine sizes. September 2, 2003. Retrieved January 22,
2004, from http://searchenginewatch.com/reports/article.php/2156481
NUA. (2002). How many online? Retrieved January 22, 2004, from http://www.
nva.ie/surveys/how_many_online/
Schönberger, V. M., & Foster, T. E. (2001). A regulatory web: Free speech and the
global information infrastructure. Michigan Telecomm. Tech Law Review, 45,
University of Michigan Law School. Retrieved January 22, 2004, from
http://www.mttlr.org/volthree/foster_art/html
Sinrod, E. J. (2000). Upside counsel: Know the laws of cyberspace. Retrieved January 22,
2004, from http://listweb.bilkent.edu.tr/anti-spam/2000/nov/ att-0000/01-up-
side_counsel.htm
Realbeer. (2001). A Bud by any other name …. Retrieved January 22, 2004, from
http://www.realbeer.com/news/articles/news-001477.php
Rhinesmith, S. H. (1993). A manager’s guide to globalization: Six keys to success in a
changing world. Alexandria, VA: McGraw Hill.
 Chapter

Practical Issues and Guidelines

for International Information Display

Nuray Aykin
Siemens Corporate Research, Princeton, NJ
Allen E. Milewski
Monmouth University, West Long Branch, NJ

INTERNATIONAL INFORMATION DISPLAY

Internationalization and localization include the study of differences

among aspects of locales and how to deal with them. Differences relevant to
international design span a wide range. Some differences are abstract and
involve conceptualizations of culture and their implications for users’ reac-
tions to products and services. Others are more pragmatic and concrete.
There are many locales and many dimensions of differences among locales.
For the practitioner who is about to tackle an internationalization project, it
is daunting to consider the sheer size of the knowledge base required to do a
good and insightful job.
This chapter is a guide to a portion of that knowledge base and is in-
tended to be used by the practitioner as a reference during internationaliza-
tion projects. The chapter focuses on pragmatic guidelines for information
display. These are critical, especially in the context of web page design and
evaluation. In some design cases, clear guidelines are possible. In other
cases, the complexities of international design make straightforward guide-

21
22 AYKIN AND MILEWSKI

lines elusive. In these cases, we resort to providing useful resources and ex-
amples that demonstrate the kinds of problems that can be confronted.
We have found it useful, in tackling international projects, to keep in
mind the following list of topics and potential problem areas as the top layer
of internationalization in any cross-cultural design:

1. Graphics and icons.

2. Language: rendering and translating.
3. data Object formatting.
3. Color.
5. Layout.

This, of course, is a grossly incomplete list of concerns for international-

ization in general (e.g., content, etc.). But, these items do cover a significant
proportion of the typical internationalization problems found in designing
web pages (Aykin, 2003). The acronym formed by this list, GLOCL, also
provides a useful checklist of problem areas that is easy to remember by re-
calling that the area of internationalization is motivated by a constant ten-
sion between what is Global and what is Local.

GRAPHICS AND ICONS

William Horton (chap. 6, this volume) already covers an extensive array of

useful guidelines on graphics and icons that are motivated by cultural dif-
ferences. These detailed guidelines are not repeated here, but it is worth
listing several general rules in using graphics in design:

• Keep the design simple.

• Leave the text out of graphics, or use overlaying text on graphics so
that when translated, the icons need not be redone.
• Use universally recognized objects. Examples include using an en-
velope image for mail instead of a rural mailbox (which is specific to
the United States) and using an octagonal stop sign instead of a
hand to stop an action.
• Implement different icons for each culture where you cannot find
culture-free graphics and icons.
• Be cautious in using images that are potentially sensitive or poten-
tially incorrect. For example:
• Animals and People—Use of animals as symbols can be offensive
to some cultures. For example, the pig is considered a filthy ani-
mal in Muslim and Jewish cultures, but it is a symbol of savings in
the United States (see coolsavings.com web site).
2. INTERNATIONAL INFORMATION DISPLAY 23

• Maps—Maps must be drawn based on the latest information. We

are in a world of ever-changing borders. It is not acceptable, and
is considered as an insult, if a map is not drawn to scale or does
not show the current legal borders.
• Directions on how images are read—Not all languages are read
from left to right. The same direction in reading applies to dis-
playing images in sequence. In languages where the text is read
from right to left, as in Arabic and Hebrew, the images are read
from right to left as well.

LANGUAGE: RENDERING AND TRANSLATING

Language is one of the key elements of internationalization. It is probably

the element with the highest visibility. It affects everything from how char-
acters are displayed to how layouts need to accommodate potential expan-
sion during translation. One of the complexities here is that languages and
countries, while related, do not form a one-to-one relationship. First, many
countries can share the same language; for example, English is spoken in
some 70 countries. Second, some countries are multilingual, such as Bel-
gium (Dutch, French), Switzerland (Italian, French, German), and Canada
(English, French). Finally, we can see different dialects (such as Parisian,
Swiss, and Canadian French) and also different writing scripts (see below) in
the same language. The lack of a one-to-one relationship between coun-
tries, languages, and scripts means that designers must consider both lan-
guage and country as potential determinants of design. The multitude of
permutations of language, dialect, country, and script has implications for
at least two key aspects of an internationalization project: language render-
ing and translation.

Language-Rendering Issues

An Example: Japanese and Chinese

Some of the difficulties confronted in rendering languages for user inter-

face displays can be shown with examples of two locales: Japan and China.
Japan has one language, but several scripts. All of these Japanese scripts can
be used and intermixed in a single sentence.

• Kanji is the complete written Japanese language. It has 50,000 Chi-

nese-style characters and can be used for writing a majority of the
language (example below from webjapanese, 2001).
24 AYKIN AND MILEWSKI

• Kana—While Kanji is essentially an ideographic system, Kana are

alphabets of written syllables that historically derive from Kanji.
Since Kana is simpler than Kanji, they make reading and writing
easier. There are two Kana forms:
• Hiragana represents all native Japanese phonemes and sounds.
It is used for articles and endings (example below from Japanese
for Otakus, 2004).

• Katakana is used to represent foreign names and words other than

Chinese or Korean (example below from Japanese for Otakus, 2004).
2. INTERNATIONAL INFORMATION DISPLAY 25

• Romaji (Roman Characters) is a set of Roman or English characters.

They are used for nontranslated words and numbers. They include:

Since characters from all of these sets can be mixed when writing in Japa-
nese, an extremely large number of characters is needed to represent a typi-
cal web page in Japan. But although the large number of characters needed
to render Japanese web pages is problematic, it is still more manageable
than accommodating countries with multiple, regional languages and dia-
lects as well as scripts. Chinese, for example, has two main dialects:

• Mandarin, which is used primarily in Mainland China, and

• Cantonese, which is used primarily in Taiwan.

In •
addition to the dialects, there are two main scripts:

• Simplified Chinese, which is used in China and Singapore, and

• Traditional Chinese, which is used in Taiwan and Hong Kong.

Character Sets

In order to deal with the processing and rendering of multiple and large
scripts, several character sets have been established and standardized. A
character set is simply an encoding of some set of language characters into a
digital, numeric code. The basic character sets commonly used in software
today are as follows:

ASCII. The standard ASCII (American Standard Code for Informa-

tion Interchange) character set consists of 128 characters ranging from 0
through 127 with the first 32 consisting of nonprinting characters. The rest
is assigned to letters, numbers, punctuation marks, and some special char-
acters. The extended ASCII character set includes the standard ASCII plus
characters from 128 through 255 representing additional special, mathe-
matical, graphic, and foreign characters.
The ASCII character set supports only English, Swahili, and Hawaiian
languages.

The ISO-8859 Family of Character Sets. Latin-1 (ISO 8859-1) is the

default character set for the Web. However, it is sufficient only for the Euro-
26 AYKIN AND MILEWSKI

pean languages. It uses 8-bit encoding and permits a maximum of 256 char-
acters. The ISO 8859 Character Set includes language-specific groupings:

ISO 8859-1 [Latin 1] Western Europe (French, Italian, German,

Spanish, …)
ISO 8859-2 [Latin 2] Central/Eastern Europe (Hungarian, Polish, Ro-
manian, Croatian, …)
ISO 8859-3 [Latin 3] Northern Europe (Esperanto and Maltese)
ISO 8859-4 [Latin 4] Southern Europe (Estonian, Latvian, Lithuanian,
Greenlandic, and Lappish)
ISO 8859-5 [Cyrillic]
ISO 8859-6 [Arabic]
ISO 8859-7 [Greek]
ISO 8859-8 [Hebrew]
ISO 8859-9 [Latin 5] Turkish
ISO 8859-10 [Latin 6] Nordic (Removed Latvian from 8859-4 and
added Icelandic)
ISO 8859-11 [Thai]
ISO 8859-12 [unassigned]
ISO 8859-13 [Latin 7] Baltic Rim (Estonian, Latvian, Lithuanian)
ISO 8859-14 [Latin 8] Celtic (Gaelic and Welsh)
ISO 8859-15 [Latin 9] Latin 0 (Removed little used characters from
8859-1 and added Euro i)

Unicode (ISO 10646 BMP). Unicode is a large character set that in-
cludes most of the world’s languages. It uses 16-bit encoding and allows
65,000 characters. This form is called UCS-2 (Universal Character Set
2-bytes). Unicode is the first level of ISO 10646, and it is also called a BMP
(Basic Multilingual Plane) or Plane Zero. More information on Unicode can
be found at www.unicode.org.

UTF-8 (Universal Character Set Transformation Format). UTF-8 is

an addendum to ISO 10646 and provides compatibility with ASCII. The
ASCII characters are represented by 1 byte (8 bits) rather than 4 bytes (32
bits). UTF-8 format has the unique ability to preserve the full ASCII range.
2. INTERNATIONAL INFORMATION DISPLAY 27

Consequently, UTF-8 is compatible with file systems, parsers, and other

software that rely on US-ASCII values.

Fonts

Character sets are ordered collections of abstract references to charac-

ters and do not specify their exact visual appearance. On the other hand,
fonts are collections of glyphs of a specific style and appearance. A single
character set can be associated with, or used by, one or more fonts. Text
must be written in a font that supports the corresponding character code set
being used. Characters, especially non-Latin characters may require differ-
ent line heights, spacing between lines, spacing between characters, charac-
ter widths, and screen resolution. Some characters, such as those in
Japanese, may not require ascenders, descenders, or capitalization. Below
are guidelines related to fonts when dealing with different characters sets:

• Provide enough space between lines to accommodate changes in

line heights, or automatically adjust line heights depending on the
characters used.
• Provide enough interline spacing to ensure clear separation be-
tween lines and to accommodate underlining.
• Do not use overly ornate or decorative fonts because accents and
special characters can become illegible.
• Choose fonts that can accommodate target languages. For exam-
ple, Helvetica and Times can provide accents. But Helvetica CE or
Times CE should be used to support the characters in central Euro-
pean languages.
• Do not assume the availability of standard fonts.
• Be prepared to support mixed (multinational) formats.
• Test your fonts for different target languages with different browsers.
• Most non-Latin languages require proportional spacing. For exam-
ple, Arabic cannot be implemented on a monospaced display.

Text Direction

Not all languages read from left to right. There are three types of text
directionality: left-to-right, bidirectional, and vertical. The most common
is the left-to-right direction that we see in Latin, Cyrillic, Greek, Thai, and
Indic languages. Chinese, Japanese, and Korean use left-to-right and verti-
cal text. Hebrew and Arabic are bidirectional (Bi-Di) where Hebrew and
Arabic characters are displayed right to left, and numbers and Latin charac-
ters are displayed left to right. Arabic and Hebrew are typically read from
right to left; however, words using non-Arabic characters and Arabic num-
28 AYKIN AND MILEWSKI

bers are read from left to right. The following example is from Microsoft’s
Saudi Arabia Web site http://www.microsoft.com/middleeast/saudi/cam-
paigns/msdn.asp. The text shows how the bidirectional text is displayed.
Note that Arabic characters are right justified, whereas left-to-right direc-
tional languages, in non-Arabic text, are left justified.

Order may also be dependent on context. For example, consider the

string “123-4567” in an Arabic context. If it is a telephone number, the
reader reads it as: 123-4567. If the string is a subtraction, the reader reads it
as subtracting 123 from 4567.
When designing a page or screen layout, it is important to consider
whether different text direction may need to be accommodated in the final
product. The design should be flexible enough to support this functionality
of text direction.

Paper Size

Printing onto paper is another aspect of language rendering. Except in

North America, most countries use the paper sizes specified by ISO 216,
which is based on the metric system. Therefore, it is important to allow the
right amount of margin if documents are shared across the globe; otherwise,
printing could be a problem because printers do not always supply both U.S.
and ISO 216 types of paper. Although Japan has adopted the ISO 216 paper
sizes, it still has different standards for the sizes of folders, defined in the JIS P
0138-61 standard (Kuhn, 2002). Table 2.1 shows the most commonly used
paper and envelope sizes defined by the United States and ISO 216.

TABLE 2.1
Paper Sizes in the United States and ISO 216
United States ISO 216 In metric units In inches
A (letter) 8½ × 11 in. A4 210 × 297 mm 8.27 × 11.69 in.
Legal 8½ × 14 in.
B (ledger) 11 × 17 in. A3 297 × 420 mm 11.69 × 16.54 in.
Business Card 2 × 3½ A8 53 × 74 mm 2.07 × 2.91 in.
2. INTERNATIONAL INFORMATION DISPLAY 29

Hardware Concerns

A final set of language rendering concerns in internationalization has to

do with display hardware. There are several key points to consider when
you are dealing with text entry and display hardware from different regions
or countries.

• Systems must be able to understand and display the character set

used by the keyboard. Make sure all required characters can be en-
tered by testing your system with different locale keyboards.
• Different keyboard layouts may be used in different countries. For
example, many European countries use an AZERTY keyboard in-
stead of a QWERTY keyboard.
• Each country has specific power, signaling, labeling, and safety re-
quirements.
• Make sure the printer has the fonts to display your coded-charac-
ter set.
• Make sure the printer can handle different paper sizes (e.g., A4).
Or your page cannot be printed without losing text in the margins
due to different paper sizes.

Language Translation Issues

Language translation is an essential component of globalization and is one

of the most expensive. One way to keep translation costs down is to provide
translation tools. For example, Microsoft (2002) defines its localization kit
as a subset of tools, source files, and binary files that can be used to create a
localized edition of a program. The localization kit is generally given to
translators or third-party contractors. Having a localization kit helps maxi-
mize the accuracy of translation and minimize the support the translators
need from the team.
In regard to translation, it is important to:

1. Identify the content of strings in the code to be translated. This is

done by either including comments in resource files or providing
separate text files to the translators.
2. Provide clearly written text because translators will often have diffi-
culty translating ambiguous terminology.
3. Provide translation glossaries. A common glossary will improve
consistency across translators and across product lines.

A strong team of technical writers is important in generating content

ready for translation. The quality of translation is highly correlated with
30 AYKIN AND MILEWSKI

how well and clearly the original text is written. Having the content pre-
pared correctly from the beginning saves costs and time during the transla-
tion process, but it requires careful preparation. Even just making sure that
there are no unnecessary words used in writing would save money, given
that most translators charge per word. In her book, Hoft (1995) provides
detailed guidelines for technical writers who develop content for an inter-
national audience.

Abbreviations and Acronyms

Abbreviations and acronyms can be a headache for translators. They

should be included in the translator’s glossary so that the meaning does not
get lost during translation. They can be spelled out, translated, or left un-
changed if they are universal (e.g., ISO, WHO). The full explanations for
acronyms and abbreviations should be available to the user in a glossary.

Spelling

The same language used in different countries can have different spell-
ing rules. For example, neighbor in the United States is spelled as neighbour
in U.K. English. Similarly:

English (U.S.) English (U.K.)

internationalization internationalisation
program programme
color colour
center centre

It is important that the users feel the language is theirs and not an im-
ported version from another country. This means that content prepared for
one locale should be reviewed for appropriateness before use in another lo-
cale that speaks the same language.

Text Expansion

When translating from English to other languages, text can expand as

much as 30% to 200%. Hoft (1995) states that most translation companies
recommend that the layout should be designed to accommodate a 30% text
expansion rate. This 30% rate applies to text that is presented in para-
graphs (over 70 characters). If the text is less than 10 characters long (e.g.,
labels, field names, or icons), the expansion rate can go as high as 200%,
2. INTERNATIONAL INFORMATION DISPLAY 31

sometimes reaching 400%. Another common rule is that a double-byte

character occupies twice as much physical space as a single-byte character.
Some examples of text expansion are shown below:

News (English) <-> Haberler (Turkish)

Product Support (English) <-> Produktunterstützung (German)
Search (English) <-> Búsqueda (Spanish)
Undo (English) <-> Ongedaan (Dutch)

In order to accommodate possible text expansion due to translation, the

following guidelines are useful:

• Provide flexibility in the screen/page layout to accommodate differ-

ent size text and fonts.
• Leave room to accommodate text expansion. Hoft (1995) suggests
a few ways to do this:
• Use a larger type size for the source language and a smaller size
for the translated version. (Be certain, however, that readability
is not sacrificed.)
• Increase the width of margins on the source document when text
expansion occurs after translation.
• Consider kerning, leading, and changing the space between
paragraphs. This reduces the space between lines.
• Place labels above the fields when the screen real estate permits.
Placing labels above the entry fields provides flexibility in case the
length of the label expands after translation. Using wraparound la-
bels is another solution if the screen real estate is a problem.
• Use symbols and icons rather than text. When appropriate, using
graphical representation can reduce the cost of translation and re-
duce the number of changes to page layout.
• Use tabs to reduce the amount of information on a page.
• Use automatically expanding text objects.
• Consider alternative layout styles.
• Allocate enough memory to allow for text expansion.

Sorting

Languages also differ in the ways characters are sorted. The sorting rules
for different languages should be able to handle:

• Accents.
32 AYKIN AND MILEWSKI

• Combinations of characters (e.g., in Spanish, “cho” comes after “co”

because “ch” is treated as a separate character that comes after “c” in
the alphabet; also, in German the character ‘ß’ is sorted as “ss”).
• Upper- and lowercase differences (in many languages uppercase is
sorted after lowercase).
• Non-Latin Scripts (e.g., Hebrew and Arabic).
• Far-Eastern languages (Japanese, Chinese, Korean).

The ISO/IEC 14651:2000 standard titled “International String Order-

ing” provides universal default collation orders for multiple languages
(available online at http://www.iso.org/iso/en/cataloguedetailpage.catalog
uedetail?csnumber=250).

Writing Practices

The following guidelines, summarized from the GNOME Documenta-

tion Style Guide (2003) and Hoft (1995), concern writing practices that can
make the translation easier and more cost effective.

Sentence Structure.

• Use short sentences and substantial white space.

Instead of: If you would like to turn the computer off, you can select the Shut
Down item from the Start menu.
Use: To turn the computer off, choose Shut Down from the Start Menu.
• Do not use joined sentences (and/or/so)—use one thought per
sentence.
Instead of: This menu item allows you to set the page properties, so you can
define your own page formatting.
Use: This menu item allows you to set the page properties. You can define
your own page formatting.
• Use active prose.
Instead of: User privileges are set by the system administrator.
Use: System administrator sets the user privileges.
• Choose formal vocabulary.
Instead of: I will pack my bag.
Use: I will prepare my luggage.
• Avoid gerunds (-ing form).
Instead of: Clicking File opens the File menu.
2. INTERNATIONAL INFORMATION DISPLAY 33

Use: Click File to open the File menu.

• Avoid indefinite pronouns (e.g., this, it, that, those).
Instead of: The Spell Checker identifies misspelled words. It is activated by
selecting “Spell Checker” from the Tools menu.
Use: The Spell Checker identifies misspelled words. To activate, select “Spell
Checker” from the Tools menu.
• Use present tense.
Instead of: Clicking on Help will activate the Help menu.
Use: Click Help to activate the Help menu.
• Avoid there is/are.
Instead of: In the menu, there are five selections.
Use: The menu contains five selections.
• Use parentheses to introduce abbreviations, not to provide ex-
planations.
Instead of: Place the localizable elements in separate files (known as re-
source files).
Use: Place the localizable elements in separate files, known as resource files.
• Using pre-organizers such as graphics, icons, and lists helps the
user as well as the translator.
Instead of: File menu contains New, Open, …
Use: File menu contains:
• New
• Open
•
•
• Avoid semicolons.
Instead of: The first item in the menu is available; all the others are
grayed-out.
Use: The first item in the menu is available. All other menu items are
grayed-out.
• Avoid personal pronouns (other than you).
Instead of: We suggest using drag-and-drop to move items.
Use: Use drag-and-drop to move items.
• Use can or might instead of may. Use may only for granting permission.
Instead of: These options may be selected.
Use: You can select these options.
• Replace contractions with complete words.
34 AYKIN AND MILEWSKI

Instead of: If you’re available.

Use: If you are available.

Terminology.

• Eliminate culture-specific metaphors.

• Avoid acronyms and abbreviations.
• Avoid jokes, humor, and idioms (all are culture-specific and may be
offending or difficult to understand and/or translate).
• Avoid gender-specific references.
• Avoid colloquial language.
• Use standard and consistent terminology throughout a document.
Create a glossary and permit one definition per item. Using one
word for the same object/event eliminates translation problems
and provides one-to-one matching for that word. The example in
the next list item shows the dangers of using different terminolo-
gies for the same thing.
• Consider terminology differences between countries using the
same language. For example, wall outlet, socket, and electrical recepta-
cle in the United States can mean ‘plug socket,’ ‘electrical point,’
and ‘power point’ in the United Kingdom. On the other hand, in
the United Kingdom, receptacle can mean a ‘container’ or ‘trash
bin,’ whereas outlet can mean ‘exit.’
• Use a spell checker to ensure correct translation and spelling.
• Ensure that the translated version retains the technical accuracy of
the original language version.
• Test your translators by retranslating the sections of the document
back to its original language.

DATA OBJECT FORMATTING

Formatting of data objects and fields is another problem area for interna-
tionalization. Here are some issues and guidelines for the data objects most
often encountered in web site design.

Date and Time

Date formats are sometimes associated with specific languages or locales.

Although ISO 8601 specifies a standard for date and time presentation,
many countries keep their own date and time formatting instead of adopt-
ing the international standards. However, this international standard is a
good way of keeping internal representation of date within a software appli-
cation, and helps to synchronize servers around the world.
2. INTERNATIONAL INFORMATION DISPLAY 35

The international standard date notation is:

YYYY-MM-DD

where YYYY is the four-digit year, MM is the two-digit month of the year be-
tween 01 (January) and 12 (December), and DD is the two-digit day of the
month between 01 and 31. For example, 1988-01-11 represents the 11th of
January in 1988. This date can be seen in various notations in around the
world: 1/11/88, 11/1/88, 88/11/1, 11 January 1988, January 11 1988, and so
on. The delimiter to separate the year, month, and day can be a forward
slash (/), space, hyphen (-), or period (.).
Many software platforms, tools, and languages such as Java and .NET
provide how-to’s and resources for presentation of dates and times in differ-
ent languages.
The differences in date formatting can be observed in the long or short date
representations. Dates are most commonly displayed as day, month, and year.
In some countries, the month may be displayed before the day and the year; in
other countries, the year may be displayed before the month and the day.
Here are the proper formats for some key locales:

MM DD YYYY - United States

DD MM YYYY - Most European countries, Australia, Canada
YYYY MM DD - South Africa, Japan, Korea, China, Taiwan

Table 2.2 shows long and short date formats for different countries.
In long date formatting, the month is usually spelled out. Using a
three-letter abbreviation for month may work well for English-speaking
countries, but can cause problems in other languages. For example, the
months June and July are Juin and Juillet in French. If the three-letter ab-
breviation is forced, then both months could be abbreviated as Jui.
The following are guidelines on how to treat date formatting for differ-
ent locales:

• Use the correct date representation for the chosen locale.

• Make use of the properties of the tools, platforms, and languages in
providing support for date and time presentations.
• If you must choose only one date format for display for use around
the world, use dd Month yyyy (e.g., 20 August 2000). Even if it is not
formatted correctly for a given locale, the content is unambiguous.
• Make provisions to accommodate different delimiters for date,
month, and year. Include a forward slash (/), period (.), space, or hy-
phen (-) as options.
 TABLE 2.2
Long and Short Date and Time Formats
for Monday April 14, 2003 in Different Countries

36
2. INTERNATIONAL INFORMATION DISPLAY 37

TABLE 2.2 (continued)

• Do not assume a three-character space for month and day abbrevia-

tions because this may not work in some languages.

Time

As ISO 8601 specifies, the international standard notation for the time of
day is

hh:mm:ss

where hh represents the hours (00–24), mm represents the minutes

(00–59), and ss represents the seconds (00–60).
Many countries use the 24-hour time notation (e.g., 15:30) rather than
12-hour AM and 12-hour PM notation (e.g., 3:30 PM). In the United States,
the 24-hour notation is referred to as “military time.” The delimiters to sep-
arate hours, minutes, and seconds can be a colon (:) or a period (.). Table 2.2
also includes the time formatting for different countries.
ISO 8601 has been adopted as European Standard EN 28601 and is
therefore now a valid standard in all E.U. countries, and all conflicting na-
tional standards have been changed accordingly (Kuhn, 2001).
38 AYKIN AND MILEWSKI

Calendar/Holidays/Start of Week

There are a few aspects regarding calendars that we need to consider when
designing for the global markets.

• Calendar Issues: Although the Gregorian calendar is used in al-

most every country, there are still cases in which a different calen-
dar might be considered. These include:
• Arabic calendar (Note: In Islamic countries, the date changes at
sunset, not midnight)
• Jewish calendar
• Iranian calendar
• Japanese Imperial calendar
• Holidays: It is important to know the holidays of the target country,
especially if you are doing business with them. Do not assume that
everyone celebrates Christmas. Many non-Christian countries do
not have a Christmas holiday. Even new year celebrations may vary.
The Chinese New Year does not coincide with the new year associ-
ated with the Gregorian calendar.
• Start of the week (Monday vs. Sunday): In the United States, the
week begins on Sunday. In most other parts of the world it begins
on Monday. For applications that deal with scheduling, this can be
an issue. The user interface showing weeks for scheduling may look
different for the U.S. users than for the European users.

Numeric Formatting

Number formatting differs from country to country. The elements of nu-

meric formatting includes:

• Thousands and decimal separator: Separators can be a character

or a space. Table 2.3 shows different separators for numeric repre-
sentation.
• Number of digits between separators: The number of digits between
separators can vary from country to country. For example,
123,456,789.00 in the U.S. can be shown as 12,34,56,789.00 in Hindi.
• Negative numbers: Negative numbers may be indicated by using a
minus sign before or after the number, or by enclosing the number
in parentheses or brackets.

-123
123-
(123)
[123]
 TABLE 2.3
Numeric Formatting and Currency Representation for Different Countries

(continued on next page)

39
40 AYKIN AND MILEWSKI

TABLE 2.3 (continued)

Names and Addresses

In the world of e-business, there are many formats for address forms. In a
typical correspondence, an address consists of

• Title
• First/given name
• Last Name/Family Name/Surname
• Street number, building number, street name
• City or town
• State/province/region
• Country
• Zip or postal code (can include alphanumeric characters)

Labels given for each of the these address entries may vary from coun-
try to country. Here are sample address formats for Mexico and the
United States.

Mexico

Name (Paternal Name, Maternal Name, First Name)

Street and Number
Building, Floor, Suite #
Colony
City, State
Postal Code
2. INTERNATIONAL INFORMATION DISPLAY 41

United States

Name (First Name, Middle Initial, Last Name)

Number and Street
City, State Zip Code1

Differences in address formatting could be seen in the labels of the name

fields. For example, First Name is mainly used in the United States, whereas
most of the world uses Given Name. Last Name in the United States becomes a
Family Name or Surname in other countries. And having multiple names or
last names adds complexity to the problem. For example, people in Mexico
have many names, including two last names.
Titles are commonly used in addresses outside the United States. Even in
personal letters, the use of a title shows a level of respect for the addressee.
In addition to Mr./Mrs./Ms., the titles Dr. and Professor are extremely im-
portant in some cultures. For example, the corporate directory for AT&T, a
U.S.-based company, includes names and job titles. The corporate direc-
tory for Siemens, a Germany-based company, includes names, job titles,
and professional titles.
The other fields in the address format include the street, building, town,
city, country, zip/postal code, and so on. The formats for these fields vary
from country to country. The differences could be in the number of address
lines, order of address lines, number of names (first and middle initial and
last/given/surnames), street addresses, level of hierarchy in terms of town,
city, state/province/region, and country, and of zip/postal codes. For each of
these fields, there should be no assumption on format and length. For ex-
ample, in the United States the zip code consists of only numeric values;
however, in many parts of the world it might include alphanumeric entries
and might be longer than a 5-digit zip code.
However, there are many attempts to generate a generic address form so
that there is no need to create a different address form for each country.
This is especially important for online registration/shopping/ordering
practices. Many times, it is impossible to identify which address format to
use for each request. Therefore, it becomes very meaningful to provide a
generic address format which spans multiple locales.
Figure 2.1 shows a good example (from www.amazon.com) of a ge-
neric address format that can handle many of the different address for-
mats in the world.

1
For more examples of country addresses, see Microsoft (2002), which provides different
address formats for a set of selected countries.
42 AYKIN AND MILEWSKI

FIG. 2.1. Sample universal address format.

Telephone Numbers

Like other formatted fields, the format for telephone numbers varies from
one country to another. The differences between telephone numbers are:

• The total number of digits (varies from six to fifteen)

• Separators (e.g., hyphens (-), periods (.), parentheses (( )), and spaces)
• Groupings of numbers (two to six digits per group)
• Country codes (one to three digits)
• Long distance access code to access national or international calls
(could be 00, 011, 001, 0, etc.)
• Code to access an outside line on a business phone (in the United
States, “9” is commonly used to place a call outside the office)
• Extensions (some phone numbers include extensions that should
be taken into consideration during design).

Table 2.4 provides a few examples of different telephone number for-

matting:
In addition to different formats in telephone numbering, in some coun-
tries, letters are not used on the keypad. This can make it difficult to dial
numbers that are given as mnemonics, as in 1-800-WEATHER. Radio Shack
used to sell templates for travelers to use on telephones without letters.
ITU-T Recommendation E.164 (The International Public Telecommuni-
cation Numbering Plan; 1997), defined by the International Telecommuni-
cations Union (ITU), formerly Comité Consultatif International de Tele-
2. INTERNATIONAL INFORMATION DISPLAY 43

TABLE 2.4
Sample Telephone Number Formats
Australia 649-800-445-768
Austria 1234 56 78 90
Belgium 12-345 67 89
Denmark 12 34 56 78
Germany (123) 4 56 78 90
Italy (12) 3456789
Portugal 123-456 78 90
Japan 123-45-6-789-0000
United States (123) 456-7890

graphique et Telephonique (CCITT), details the format and structure of

the Public Switched Telephone Network (PSTN) numbering plan. Within
this Recommendation, it was proposed that all switches and networks
should have the capability of handling international telephone numbers
with a maximum length of 15 digits. However, given that the separators
(e.g., spaces or parentheses) can end up being part of the telephone num-
ber entry, the length of the telephone numbers may exceed 15 digits. It is
important not to limit the digits in a phone number entry in fields and also
not to parse the telephone numbers into area codes and phone numbers as
there may not be an area code, but only a city code in some countries. Users
should also be able to enter telephone numbers in free format (can use (, ), -,
/). Consider these notations in determining field sizes.

Currency

There are many different currencies in the world. Table 2.3 shows a few ex-
amples of currency representation. Although some currencies have special
symbols (such as U.S. dollar as $, Japanese Yen as ¥, British Pound as £),
many use abbreviations of the currency name (such as TL for Turkish Lira).
ISO 4217 (Codes for the Representation of Currencies and Funds) defines
three-letter abbreviations for world currencies.
The general principle used to construct these abbreviations is to take the
two-letter country abbreviations defined in ISO 3166 (Codes for the Repre-
sentation of Names of Countries) and append the first letter of the currency
name (e.g., GBP for the Great Britain Pound, and USD for the United
States Dollar). In the case of currencies defined by supra-national entities,
44 AYKIN AND MILEWSKI

ISO_4217 assigns two-letter entity codes starting with “X” to use in place of
country codes (e.g., XCD for the Central Caribbean Dollar; Allen & Hall,
2003). Similarly, because there is no specific country associated with Euro,
the code for Euro is EUR, with a well-recognized symbol, i.
It is common practice to use the symbols for well-known currencies (such
as $ or £) when monetary amounts are displayed for an international audi-
ence. For local audiences, it is again a practice to use the currency symbol
since the users are familiar with their own currency symbol. However, it is
advisable to use the three-letter code if the audience is international and the
symbol is not well recognized worldwide.

Monetary Values

The format for representing monetary value differs from country to coun-
try. Most locales use their numeric formatting rules in terms of a decimal
and thousands separator. But this is not always the case. For example, in Es-
tonia, the period is used as the decimal separator for Estonian Kroons
(123.45 kr), but the comma is used as the decimal separator for other nu-
meric representations (123,45; Microsoft, 2002).

Placement of Currency Symbol and Negative Sign

The placement of the currency symbol and the negative sign also differs
from country to country. Some countries, such as United States, use paren-
theses to indicate negative monetary value.

U.K. –£ 1,324.55
U.S. ($1,324.55)
Netherlands 1,324.55 i –
France –1,324.55 i

Sizes and Measurements

The United States uses the imperial system for the measurements, whereas
most countries use the metric system. For computer applications, this is es-
pecially important for files that rely on measurements (e.g., page size, icons,
clip art) that are shared across regions. Sizes and measurements that may
change from country to country include:

• Typographic units (point, pixel, measurement units).

• Temperature measurements (Fahrenheit or Celsius)
2. INTERNATIONAL INFORMATION DISPLAY 45

• Paper and envelope sizes (see page 28)

• Clothing sizes

Metric Versus Imperial Units

The Imperial system uses the foot, the pound, and the Fahrenheit (Fº) scale
to measure length, mass (weight), and temperature, respectively. The met-
ric system uses the meter, the kilogram, and the Celsius (Cº) scale.
Most of the world uses the metric system, except for Burundi, Liberia,
Yemen, Rwandese Republic, Union of Myanmar, and the United States
of America.

Typographic Units

Vekulenko (2000) and Kuhn (1999) summarize the differences between the
typography units used around the world. To name a few widely used ones:

• 1 point (Didot) = 0.3759 mm = 1/72 of a French Royal inch (27.07

mm) = about 1/68 inch
• 1 point (ATA) = 0.3514598 mm = 0.0138366 inch
• 1 point (TeX) = 0.3514598035 mm
• 1 point (Postscript) = 0.3527777778 mm = 1/72 inch
• 1 point (l’Imprimerie nationale, IN) = 0.4 mm

When sharing images between different locales, be certain to use the

same point system to avoid design mistakes.

Temperature

Before the 1970s, most English-speaking countries used the Fahrenheit

scale. Since then, most countries adopted the Celsius temperature mea-
surement scale, except in the United States, which uses the Fahrenheit
scale. Many travel sites now show temperatures in both scales to help
travelers.

Clothing Size

It is especially important for online merchandise catalogs to show metric

and imperial shoe and clothing sizes. Even if the target market is the United
States, there are many people in the United States from different countries
who will be pleased to see the sizes given in the system that they are used to.
Many travel agencies’ sites, such as American Express, provide clothing size
46 AYKIN AND MILEWSKI

charts to show the conversions for different countries. A useful bit of infor-
mation for travelers! Table 2.5 shows the conversion for clothing sizes.

COLOR

Color has very different meanings around the world and therefore it is a sig-
nificant element, not just for computer user interfaces, but for all forms of
international interaction. Table 2.6 shows different meanings of colors for
various cultures (adapted from HP, 2003; Morton, 2003; Prime & Wilson,
2002; Jinsheng, 2001). For example, Holzschlag (1999) and Morton (2003)
point out that the color purple represents death in Catholic parts of Europe
and that Euro Disney should have done their homework well before choos-
ing purple as a signature color.

TABLE 2.5
Clothing Sizes for the United States, Europe, and the United Kingdom
Women’s Dress Sizes Women’s Shoe Sizes
United United United
United States Europe Kingdom States Europe Kingdom
6 36 8 5 35 3.5
8 38 10 5.5 36 4.0
10 40 12 6.0 36 4.5
12 42 14 6.5 37 5.0
14 44 16 7 37 5.5
16 46 18 7.5 38 6.0
18 48 20 8.0 38 6.5
20 50 22 8.5 39 7.0
22 52 24 9.0 39 7.5
Men’s Suit Sizes Men’s Shoe Sizes
United United United
United States Europe Kingdom States Europe Kingdom
36 46 36 8 41 7
38 48 38 8.5 42 7.5
40 50 40 9.5 43 8.5
42 52 42 10.5 44 9.5
44 54 44 11.5 45 10.5
46 56 46 12.0 46 11
 TABLE 2.6
Color Meanings in Various Cultures
Color Meaning for Different Cultures
Red China: Good luck, prosperity, happiness, marriage
Japan: Anger, danger
Middle East: Danger, Evil
India: Purity, life
Egypt: Death
South Africa:Mourning
USA and Western Europe: Danger, anger, stop
Blue China: Immortality, heavens, depth, cleanliness
Japan: Villainy
India: Color of Krishna
USA and Western Europe: Masculinity, calm, authority, peace
Yellow East Asia: Sacred, Imperial, Royalty, Honor
Middle East: Happiness, prosperity
India: Religious color (celebration),success
Egypt: Mourning
USA and Western Europe: caution, cowardice
Green Asia: Family, harmony, health, peace, life, youth, energy, growth
Muslim countries: Religious color, fertility, strength
Ireland: National color
USA and Western Europe: Safe, go
White Asia: Mourning, death, purity
Middle East: Mourning, purity
India: Death, purity
USA and Western Europe: Purity, virtue
Black Asia: Evil influences, knowledge, mourning
Middle East: Mystery, evil
USA and Western Europe: Mourning, elegance (black dress and
tuxedo), death, evil
Brown India: Mourning
Purple Thailand: Mourning
Catholics: Death
Asia: Wealth, glamour
Western Europe: Royalty
Orange Ireland: Religious significance
China: Light and warmth
USA: Inexpensive goods

47
48 AYKIN AND MILEWSKI

There are numerous excellent resources that deal with color usage. Jill
Morton’s colormatters.com web site has extensive information on color
studies and online books on color. Spartan (Multimedia Produc-
tions,1999) gives many examples related to different cultures in their as-
sociation with colors. Here are a few examples from Spartan’s colorful
multimedia presentation:

• In the United States, Americans tend to relate blue to authority,

whereas in Canada, red is authority, as well as an inspirer of nation-
alism. In the British Isles, black is commonly used for authority fig-
ures. The use of red and black in Britain is usually reserved for
honorary positions, such as the Royal Guard.
• White is a very common color in hot Mediterranean countries, such
as Greece. The use of white dates back to ancient Greece when all
buildings, sculptures, and so on were required to be clean white.
• In China, red and gold are used for joyful, festive occasions. They
are also commonly used for religious shrines and temples, as well as
other forms of architecture.
• Orange is widely used by Indians in their attire and as a means of at-
tention. It also has religious significance in Hindi and Buddhism.
• The Japanese spectral range of colors is mainly muted, pastel
types, with the exception of bright red.
• In Korea, light blue and other pastel colors in clothing are usually
reserved for the wealthy and for festivals. Commoners usually wear
white or grays.
• In many African cultures, black is a symbol of strength, and white is
a symbol of fertility.
• Native Indians of Mexico used bright colors for their ceremonial
costumes. Possibly stemming from the Indian cultures, the colors of
Mexico tend to be bright and festive greens to reds or a pastel ver-
sion of those colors.

With world boundaries blurring, and the cultures blending more than
ever before, some of the strong color preferences and meanings are becom-
ing less apparent. We now, for example, see many brides wearing tradi-
tional white dresses in all Asian countries.
Although the strict meanings of colors may be fading across countries, it
is still a good practice to use colors carefully. We do not see brightly colored
web sites or consumer products for the Japanese culture; instead, products
are usually designed with pastel colors in mind.
If the same color palette is to be used across locales, it would be a good
idea not to use primary colors in design because they may still carry nega-
tive or positive meanings in some cultures.
2. INTERNATIONAL INFORMATION DISPLAY 49

LAYOUT

Our final internationalization design consideration is layout. It is impor-

tant not to assume anything about layout because all layout elements vary
according to language (W3C, 2003). These elements include how a piece
of text might be drawn on the screen, text alignment and justification,
how much room translated text takes up, the direction it flows, line
breaks, text wrapping, white spaces, word and letter spacing, text deco-
rating (underline, overline, line-thru, text shadows, text underline solu-
tion for East Asian vertical writing and blinking), and where on the
screen it should start.
Make sure there is enough room for text expansion. Layout has to accom-
modate different text properties and the change in context size due to trans-
lation. Therefore, it is important not to hardcode physical layout space, but
to let it adjust dynamically. Use of dynamic layout managers is preferred over
fixed-position layout because of the benefit of cross-platform automatic lay-
out management for graphical user interface components.

SUMMARY

The purpose of this chapter was to list, in one place, a good sampling of the
issues, guidelines, resources, and tips associated with international infor-
mation display. This chapter should make it possible for the user interface
practitioner to begin an internationalization project in a knowledgeable
and productive way. The list of issues described in this chapter is by no
means a complete list. But these issues will give cross-cultural designers a
foundation on which experience can build a fuller understanding of life-
styles, preferences, customs, and work patterns.

REFERENCES

Allen, P. L., & Hall, J. (2003). World currencies and abbreviations. Retrieved January 22,
2004, from http://www.jhall.demon.co.uk/currency/
Aykin, N. (2003). Internationalization and localization of Web sites. Tutorial presented at
the User Experience 2003 Conference.
GNOME Documentation Style Guide. (2003). Chapter on writing for localization. Pub-
lished by: Sun Microsystems. Retrieved January 22, 2004, from http://devel-
oper.gnome.org/documents/style-guide/
Hewlett-Packard. (2003). Printing with color—the meaning of color. Retrieved January
22, 2004, from http://www.hp.com/sbso/productivity/color/meaning.html
Hoft, N. L. (1995). International technical communication. New York: Wiley.
50 AYKIN AND MILEWSKI

Holzschlag, M. E. (1999). Color my world. New architect: internet strategies for technology
leaders. Retrieved January 22, 2002, from http://www.webtechniques.com/ar-
chives/2000/09/desi/
Hongladarom, S. (2000). Negotiating the global and the local: How Thai culture co-opts
the Internet. Retrieved January 22, 2004, from http://www.firstmonday.dk/is-
sues/issue5_8/hongladarom/
ITU-T Recommendation E.164. (1997). The International Public Telecommunication
Numbering Plan. International Telecommunications Union. Telecommunication
Standardization Bureau. Geneva, Switzerland. Retrieved January 22, 2004, from
http://www.itu.int/ITU-T/index.html
Japanese for Otakus. (2004). Retrieved January 22, 2004, from http://www.orange an-
gel.com/otaku/april11.html
Jinsheng, M. (2004). Colors and meanings in Chinese modern products. Retrieved January
22, 2004, from http://www.aedo-to.com/eng/library/aedo-ba/aedo_02/art04.html
Kuhn, M. (1999). Metric typographic units. Retrieved January 22, 2004, from
http://www.cl.cam.ac.uk/~mgk25/metric-type/html
Kuhn, M. (2001). A summary of the international date and time notation. Retrieved Janu-
ary 22, 2004, from http://www.cl.cam.ac.uk/~mgk25/iso-time.html.
Kuhn, M. (2003). International standards paper size. Retrieved January 22, 2004, from
http://www.cl.cam.ac.uk/~mgk25/iso-paper.html.
Microsoft (Dr. International Team). (2002). Developing international software.
Redmond, WA: Microsoft Press.
Morton, J. L. (2003). Color matters. Retrieved June 10, 2003, from www.colormattters.com
Morton, J. L. (2003). Global color meanings. Retrieved January 22, 2004, from
http://www.colormatters.com/bubarc8-global.html
Prime, M., & Wilson, M. (2002). World Wide Web Consortium (W3C) primer on the inter-
nationalisation and localisation of web pages: W3C UK and Ireland Office 1 Decem-
ber 2002. Retrieved January 22, 2004, from http://www.w3c.rl.ac.uk/QH/
QH_final_review_WP5/WD-int-primer.html
Russo, P., & Boor, S. (1993). How fluent is your interface? Designing for interna-
tional users. INTERCHI’93. 342–347.
Spartan (1999). Multimedia presentation on the colors of culture. Retrieved January 22,
2004, from http://www.mastep.sjsu.edu/Alquist/workshop2/color_and_cul-
ture_files/frame.htm
Vekulenko, A. (2000). Differences between point systems. Retrieved January 22, 2004,
from http://www.oberonplace.com/dtp/fonts/point.htm
Webjapanese.com. (2001). Retrieved June 10, 2003, from http://webjapanese.com/
wj/ftp/dtp/images/kanji01_640_w.gif
Wired4success. (2003). Color symbolism chart. Retrieved January 22, 2004, from
http://www.wired4success.com/colorsymbolism.htm
 Chapter !

User Interface Design and Culture

Aaron Marcus
Aaron Marcus and Associates, Berkeley, CA

INTRODUCTION

Over the past two and a half decades in the user interface (UI) design com-
munity, designers, analysts, educators, and theorists have identified and
defined a somewhat stable set of UI components, that is, the essential enti-
ties and attributes of all UIs, no matter what the platform of hardware and
software (including operating systems and networks), user groups, and con-
tents (including vertical markets for products and services). That means
these components can enable developers, researchers, and critics to com-
pare and contrast UIs that are evidenced on terminals, workstations, desk-
top computers, Web sites, Web-based applications, information appliances,
and mobile, wireless devices. (Marcus, 1998, 2001), among others, provides
one way to describe these UI components, which is strongly oriented to
communication theory and to applied theory of semiotics (Eco, 1976;
Peirce, 1933). This philosophical perspective emphasizes communication
as a fundamental characteristic of computing, one that includes perceptual,
formal characteristics, and dynamic, behavioral aspects of how people in-
teract through computer-mediated media. Picking up on Claude Lévi-
Strauss’s (2000) theory of human beings as sign makers and tool makers, my
theory understands a UI as a form of dynamic, interactive visual literature
as well as a suite of conceptual tools, and as such, a cultural artifact. The UI
components are the following:

51
52 MARCUS

Metaphors

Metaphors are fundamental (Lakoff and Johnson, 1980) concepts commu-

nicated via words, images, sounds, and tactile experiences. Metaphors sub-
stitute for computer-related elements and help users understand,
remember, and enjoy entities and relationships of computer-based com-
munication systems. Metaphors can be overarching, or they can communi-
cate specific aspects of UIs. An example of an overarching metaphor is the
desktop metaphor to substitute for the computer’s operating system, func-
tions, and data. Examples of specific concepts are the trashcan, windows
and their controls, pages, shopping carts, chatrooms, and blogs (Web logs).
The pace of metaphor invention, including neologisms or verbal metaphor
invention, is likely to increase because of rapid development and distribu-
tion through the Web and mobile devices of ever-changing products and
services. Some researchers are predicting the end of the desktop metaphor
era and the emergence of new fundamental metaphors (Gelerntner, 2000).

Mental Models

Mental models are structures or organizations of data, functions, tasks,

roles, and people in groups at work or play. These are sometimes also called
user models, cognitive models, and task models. Content, function, media,
tool, role, goal, and task hierarchies are examples.

Navigation

Navigation involves movement through the mental models, that is, through
content and tools. Examples of UI elements that facilitate such movement
include those that enable dialogue, such as menus, windows, dialogue
boxes, control panels, icons, and tool palettes.

Interaction

Interaction includes input/output techniques, status displays, and other

feedback. Examples include the detailed behavior characteristics of key-
boards, mice, pens, or microphones for input; the choices of visual display
screens, loudspeakers, or headsets for output; and the use of drag-and-
drop selection/action sequences.

Appearance

Appearance includes all essential perceptual attributes, that is, visual, audi-
tory, and tactile characteristics,. Examples include choices of colors, fonts,
3. USER INTERFACE DESIGN AND CULTURE 53

animation style, verbal style (e.g., verbose/terse or informal/formal), sound

cues, and vibration modes.
As computer-based communication has become a global enterprise,
the development, marketing, and business communities have become
aware of the impact of world cultures on UI design for global products and
services. A clear need has emerged to determine optimum characteristics
of suitably localized products and services based on market and user data
to achieve both short-term and long-term success, without having to de-
velop too many variations that might waste time and money in develop-
ment, distribution, and maintenance.
Likewise, over many more decades, cultural anthropologists have identi-
fied and defined somewhat stable, fundamental analytical dimensions by
which one may contrast and compare all world cultures. Culture, as evi-
denced at work, at home, in schools, and in families, may be defined in terms
of a group’s symbols, heroes, rituals, and values. In this chapter, culture re-
fers, in general, to well-established groups whose characteristics a member
may have significant difficulty to abandon, but may also include what might
be called “temporary, voluntary life-style affinity groups” not recognized as
cultures by some classical cultural anthropologists (Clausen, 2000).
For example, consider the international express-mail Web site page of
DHL for shipping packages to Saudi Arabia shown in Fig. 3.1. Is this imagery
appropriate for potential customers? This chapter seeks to provide initial re-
sources for considering such questions. The issue is potent if one considers
what DHL and FedEx, another international shipping company, were show-
ing at the same point in time. Both DHL and FedEx had similar layouts, but
the FedEx site showed several standard images, including an Asian woman
with bare sleeves and uncovered face. This race, gender, and dress style is not
typical for Saudi Arabia, while the DHL site showed an apparently Middle
Eastern man in conservative Western business attire. The DHL page also
showed a green logotype in Arabic script at the lower left, a typical reference
via color and script to Saudi Arabian nationality and the Moslem religion.
As an example of a culture model, one may cite the work of the cultural
analyst Geert Hofstede (Hofstede, 1997). He studied hundreds of IBM em-
ployees in 53 countries from 1978 to 1983 (see Table 3.1 for a display of his
findings), from which study he developed five fundamental culture dimen-
sions. These dimensions of culture are described in the following sections.

Power Distance

Power distance measures the extent to which people of a culture accept large
or small distances of power in social hierarchies. For example, if an employee
of a large organization within a culture typically has easy, informal access to
the “boss,” then one might assign a low power distance rating to that culture.
54 MARCUS

FIG. 3.1. A DHL express-mail Web site page for Saudi Arabia, as displayed in
2000. Note that the Web site has been updated significantly since then.

Individualism Versus Collectivism

Individualism versus collectivism measures the orientation to individual or

group achievements. If a culture typically focuses on group goals and ori-
ents promotion of achievements to the group, rather than to individuals of
the group, then one might assign a high collectivist rating to that culture.

Masculinity Versus Femininity

Masculinity versus femininity measures the degree to which a culture does or

does not separate traditional gender roles. Traditional gender roles assume
that males are tough, task-oriented, individual- or group-oriented “warriors”
going out in the world to conquer and defend, and females are tender, gentle
homemakers and childrearers who are family- and people-oriented. Mascu-
line cultures tend to support this traditional role separation; feminine cul-
tures tend to diminish the differences, that is, to merge both roles.

Uncertainty Avoidance

Uncertainty avoidance measures the degree to which a culture is uncom-

fortable with uncertainty and seeks to reduce uncertainty, often in the pur-
 TABLE 3.1
Hofstede’s Cultural Indexes
PDI IDV MAS UAI LTO
Rank Score Rank Score Rank Score Rank Score Rank Score
Arab Countries 7 80 26/27 38 23 53 27 68
Argentina 35/36 49 22/23 46 20/21 56 10/15 86
Australia 41 36 2 90 16 61 37 51 15 31
Austria 53 11 18 55 2 79 24/25 70
Bangladesh 11 40
Belgium 20 65 8 75 22 54 5/6 94
Brazil 14 69 26/27 38 27 49 21/22 76 6 65
Canada 39 39 4/5 80 24 52 41/42 48 20 23
Chile 24/25 63 38 23 46 28 10/15 86
China 1 118
Columbia 17 67 49 13 11/12 64 20 80
Costa Rica 42/44 35 46 15 48/49 21 10/15 86
Denmark 51 18 9 74 50 16 51 23
East Africa 21/23 64 33/35 27 39 41 36 52
Equador 8/9 78 52 8 13/14 63 28 67
Finland 46 33 17 63 47 26 31/32 59
France 15/16 68 10/11 71 35/36 43 10/15 86
Germany FR 42/44 35 15 67 9/10 66 29 65 14 31
Great Britain 42/44 35 3 89 9/10 66 47/48 35 18 25
Greece 27/28 60 30 35 18/19 57 1 112
Guatemala 2/3 95 53 6 43 37 3 101
Hong Kong 15/16 68 37 25 18/19 57 49/50 29 2 96
India 10/11 77 21 48 20/21 56 45 40 7 61
Indonesia 8/9 78 47/48 14 30/31 46 41/42 48
Iran 29/30 58 24 41 35/36 43 31/32 59
Ireland (Rep of) 49 28 12 70 7/8 68 47/48 35
Israel 52 13 19 54 29 47 19 81
Italy 34 50 7 76 4/5 70 23 75
(continued on next page)

55
 TABLE 3.1 (continued)
PDI IDV MAS UAI LTO
Rank Score Rank Score Rank Score Rank Score Rank Score
Jamaica 37 45 25 39 7/8 68 52 13
Japan 33 54 22/23 46 1 95 7 92 4 80
Malaysia 1 104 36 26 25/26 50 46 36
Mexico 5/6 81 32 30 6 69 18 82
Netherlands 40 38 4/5 80 51 14 35 53
Philippines 4 94 31 32 11/12 64 44 44 21 19
Poland 13 32
Portugal 24/25 63 33/35 27 45 31 2 104
Salvador 18/19 66 42 19 40 40 5/6 94
Singapore 13 74 39/41 20 28 48 53 8 9 48
South Africa 35/36 49 16 65 13/14 63 39/40 49
South Korea 27/28 60 43 18 41 39 16/17 85 5 75
Spain 31 57 20 51 37/38 42 10/15 86
Sweden
Sweden 47/48 31 10/11 71 53 5 49/50 29 12 33
Switzerland 45 34 14 68 4/5 70 33 58
Taiwan 29/30 58 44 17 32/33 45 26 69 3 87
Thailand 21/23 64 39/41 20 44 34 30 64 8 56
Turkey 18/19 66 28 37 32/3 45 16/17 85
Uruguay 26 61 29 36 42 38 4 100
USA 38 40 1 91 15 62 43 46 17 29
Venezuela 5/6 81 50 12 3 73 21/22 76
West Africa 10/11 77 39/41 20 30/31 46 34 54
Yugoslavia 12 76 33/35 27 48/49 21 8 88
Zimbabwe 19 25

Note. PDI = power distance index; IDV = individualism index; MAS = masculinity index; UAI
= uncertainty avoidance index; LTO = long-term orientation index. Adapted from Cultures and Or-
ganizations: Software of the Mind by G. Hofstede, 1997, pp. 26, 55, 84, 113, 166. Original data copy-
right 1997 by Geert Hofstede.

56
3. USER INTERFACE DESIGN AND CULTURE 57

suit of the truth. Cultures that focus on punctuality, formality, and explicit
communication tend to rate high in uncertainty avoidance.

Long-Term Time Orientation

Long-term time orientation grows out of a long-term basis of some cul-

tures in Confucian thought, which emphasizes patience. Naturally, this di-
mension seems strongest among Asian cultures steeped in Confucian
philosophy. (For an example of communication influences and differ-
ences, see Oliver, 1971.)
An example of the interrelations of two cultural dimensions, power distance
versus individualism-collectivism, for several countries appears in Figure 3.2.
De Mooij (2001) has argued for culture differences as a basis for differ-
ences of Internet and mobile behavior separate from economic factors.
These culture dimensions, examples of culture bias in Web site designs, and
the implications of culture dimensions on Web site and Web-based applica-
tion designs are discussed at length elsewhere (Marcus, 2001; Marcus &
Gould, 2000).

FIG. 3.2. Chart of power distance versus individualism-collectivism for selected

countries. Adapted from Cultures and Organizations: Software of the Mind by G.
Hofstede, 1997, p. 54. Copyright 1997 by McGraw-Hill. Adapted with permission
of McGraw-Hill.
58 MARCUS

MAPPING DIMENSIONS TO COMPONENTS

Although these definitions of dimensions and components are not univer-

sally accepted, they are stable enough to consider a next phase of under-
standing the relationship of culture to UIs: mapping components of UIs to
dimensions of culture, and vice versa. Being able to understand the relation-
ships of UI components to a particular culture dimension is a step forward to
mapping particular UI designs to one or more cultures. Reliably being able
to understand and predict the use, acceptance, and enjoyment of particular
UIs marks significant progress in developing consistently successful products
and services. In addition, with greater understanding of the relationships, it
becomes more feasible to develop databases, such as databases of culture
phenomena or catalogs of UI design patterns (Tidwell, 1999) in which cul-
ture dimensions play a more significant role in the attributes and tools.
Such tools might enable designers to eventually “semiautomatically” ad-
just UI designs per culture dimension, to sets of countries with similar-cul-
ture characteristics (e.g., adjusting UI designs to be suitable for all countries
that were once part of the British Empire), and ultimately, per culture. These
“culture-design” tools someday may be ubiquitous and conventional, similar
to the color adjustment tools today of a graphics product like Adobe
Photoshop™, which permits a designer to choose color sets that are safe for
the Web. A future scenario might find designers (from one or more cultures)
designing a UI for a specific culture, then asking the computer to adjust the
entire UI design for another culture, say Culture X, to make it “Culture
X-safe.” This design might be analyzed by the designers or submitted to users
of the target culture for feedback. In an interactive process, the design might
be adjusted, including the adjustment of culture parameters to reflect ongo-
ing changes in the target cultures, which are not themselves static.
To explore possible relations of culture dimensions to UI components, I,
sometimes in conjunction with Emilie Gould of Rensselaer Polytechnic In-
stitute (goulde@rpi.edu), have been giving tutorials about cross-cultural
communication since 1999 in several countries, including North America,
Europe, South Africa, and Japan. In the tutorials, we ask participants to ini-
tiate analysis and discussion of the relation of culture dimensions to UI
components. The conceptual challenge is to fill in a matrix like that of Fig.
3.3. To stimulate further study, an initial mapping appears here. This map-
ping, in greater detail, with the appropriate conditional influences, would
be the basis for detailed heuristics, design pattern catalogs, and, eventually,
a kind of artificial-intelligence-based application that could assist in design-
ing UIs to be culturally conservative or radical according to the needs/de-
sires of those doing the designing.
3. USER INTERFACE DESIGN AND CULTURE 59

FIG. 3.3. Matrix mapping culture dimensions to user interface components.

In the analysis that follows, for a given culture dimension, each subsec-
tion provides characteristic examples, as determined by personal experi-
ence, not yet by detailed study, of UI elements listed per UI component, as
described previously. Although this is an initial categorization by assertion,
I believe it nevertheless has pedagogic and practical value.

Power Distance

Metaphors

High: Government or corporate institutions and buildings; objects

with clear hierarchies, e.g., the human body, schools, government, mon-
uments, etc.
Low: Informal, or popular institutions, buildings, and objects, that empha-
size equality, options; Montessori schools, play/games, public spaces, etc.

Mental Models

High: Complex, highly organized, highly categorized, highly populated

structures, e.g., large corporate and/or government organizational models
or charts; reference data with little or no relevancy ranking
Low: Simple, informally organized and categorized structures; less struc-
tured data with some or much relevancy
60 MARCUS

Navigation

High: Restricted access, choices; authentication; passwords; prescribed

routes
Low: Open access, multiple options, sharable paths

Interaction

High: Severe error messages: “Entry Forbidden,” “You are wrong;” wiz-
ards or guides lead usage
Low: Supportive error messages, cue cards; many user-driven options
available

Appearance

High: Images of leaders; national/corporate/government themes, slo-

gans, insignia, logos, symbols, typefaces, layouts, colors; official music, an-
thems; formal speech
Low: Images of people, groups; daily activities; popular music, symbols,
typefaces, layouts, colors; informal speech

Individualism Versus Collectivism

Metaphors

Individualist: Action-oriented, tool-oriented

Collectivist: Relationship-oriented, content-oriented

Mental Models

Individualist: Product- or task-oriented

Collectivist: Role-oriented

Navigation

Individualist: Individual paths; popular choices, celebrity choices; stable

across roles; customizable
Collectivist: Group-oriented, official choices; changes per role
3. USER INTERFACE DESIGN AND CULTURE 61

Interaction

Individualist: Keyword searches; active-oriented; multiple devices; cus-

tomizable;
Collectivist: Limited, official devices; role-driven

Appearance

Individualist: Images of products, people; low context; hyperbolic, dy-

namic speech; market-driven topics, imagery, language; customizable; di-
rect, active verbs
Collectivist: Images of groups, organizations; images of roles; high con-
text; official, static terminology; institution-driven topics, imagery, lan-
guage; passive verbs

Masculinity Versus Femininity

Note: This subsection’s contents are differently organized and may be mis-
interpreted without considering the fact that the examples list items for tra-
ditional gender roles, not for masculine cultures or feminine cultures per se.
Masculine cultures would maintain clear separation of such examples; fem-
inine cultures would tend to combine or merge them, making fewer gen-
der-oriented distinctions.

Metaphors

Masculine: Sports-oriented; competition-oriented; work-oriented

Feminine: Shopping carts; family-oriented, people-oriented

Mental Models

Masculine: Work/business structures; high-level, “executive views;”

goal-oriented
Feminine: Social structures; detailed views; relationship-oriented

Navigation

Masculine: Limited choices, synchronic

Feminine: Multiple choices; multitasking, polychronic
62 MARCUS

Interaction

Masculine: Competitive-game-oriented; mastery-oriented; individual-

oriented
Feminine: Practical, function-oriented; cooperation-oriented; team ori-
ented

Appearance

Masculine: “Masculine” colors, shapes, sounds

Feminine: “Feminine” colors, shapes, sounds; acceptance of cuteness

Uncertainty Avoidance

Metaphors

High: Familiar, stable, clear references to daily life; representation

Low: Novel, unusual references; abstraction

Mental Models

High: Simple, explicit, clear articulation; limited choices; binary logic

Low: Tolerance for ambiguity, implicit structures or relations, complex-
ity; fuzzy logic, i.e., multivariate, non-binary logic.

Navigation

High: Desire for limited, clear organized options; tolerance for complex,
fine tuning controls to “master” or “control” a situation, e.g., advanced
searches on the Web, consumer electronics controls.
Low: Tolerance for ambiguous, possibly redundant options; tolerance
for risk, gambling; tolerance for simple controls, e.g., simple searches on
the Web, or www.Google.com’s “I Feel Lucky” button.

Interaction

High: Precise, complete, detailed input and feedback of status; devices

that permit finetuning.
Low: General, limited, or ambiguous input and feedback of status; de-
vices that may have gross tuning.
3. USER INTERFACE DESIGN AND CULTURE 63

Appearance

High: Simple, clear, consistent imagery, terminology, sounds; highly re-

dundant coding of perceptual cues.
Low: More varied, ambiguous, less-consistent imagery, terminology,
sounds; tolerance for more perceptual characteristics involved in purely or-
namental or aesthetic use; less redundant coding of perceptual cues.

Long-Term Time Orientation

Metaphors

Long: Stable family references; paternalistic imagery and references: Fa-

ther, Mother; stable institution/organization references, e.g., universities,
major businesses, the Mafia, Chinese state businesses, IBM in 1950s, etc.;
concrete metaphors
Short: Interchangeable roles, jobs, objects; less concrete, more abstract
metaphors

Mental Models

Long: Love/devotion; social coherence, responsibility, support

Short: Liberty: social incoherence, social irresponsibility, efficiency

Navigation

Long: Tolerance for long paths, ambiguity; contemplation-oriented, re-

liance on personal advisor; context-oriented
Short: Bread-crumb trails, clear taxonomies; quick-results; logical infer-
ence-oriented, function or action-oriented

Interaction

Long: Preference for face-to-face communication, harmony; personal-

ized messages; more links to people; live chats; interaction as “asking”
Short: Distance communication accepted as more efficient; anonymous
messages tolerated; conflict tolerated, even encouraged; performance-crit-
ical communication
64 MARCUS

Appearance

Long: Cultural markers: flags, colors, national images; soft focus; warm,
fuzzy images; pictures of groups inviting participation, suggestions of inti-
macy and close social distance
Short: Minimal and focused images; short borders, lines, edges; concen-
tration on showing task or product
This initial taxonomy represents an attempt to group UI attributes in a
manner that highlights the connection to culture dimensions. A more de-
finitive mapping clearly requires extensive effort, but this initial construct
points the way.

FUTURE STUDY OF CULTURE IN RELATION

TO USER INTERFACE DESIGN

The preceding analysis is not the only way to relate UI design and culture.
Recent publications highlight other cultural issues that also should be con-
sidered in understanding how cultural differences might impact UI design.
These other reference dimensions could/should be mapped to culture di-
mensions and to UI components in future research.
First of all, Clausen (2000) argued that classical culture in the United States
no longer exists, at least as defined by cultural anthropologists in the past. Cul-
ture in earlier circumstances represented group environments that were diffi-
cult to escape. Now, many uses of the term culture refer simply to matters of
choice, not requirement, in other words, to lifestyles. The strict use of the term
would not be applied to, for example, consumer product affinity groups, such
as owners of Volvo automobiles. Consequently, designers may need to debate
how strictly they apply the concepts of traditional anthropology to current
product design and to performance, preference, and enjoyment differences
among target groups. As noted earlier, this chapter adopts a looser definition
of culture, rather than a stricter one, for practical purposes in UI design.
Some research about how different cultures use different UI constructs
has appeared. Choong and Salvendy (1998), for example, studied the use of
icons in Chinese UIs. For their article “Implications for Design of Com-
puter Interfaces for Chinese Users in Mainland China,” Choong and
Salvendy (1999) investigated the impact of cultural differences on com-
puter performance for 40 Chinese subjects residing in mainland China and
40 U.S. users and the design of appropriate UIs for the Chinese users. Their
results are also commented on by Carroll (1999).
Choong and Salvendy found, in general, that U.S. users rely much
more on logical inference and categorization. These users tend to classify
by functions, analyze components, and infer common features. In con-
trast, the Chinese users relied more on relations and contexts. These users
3. USER INTERFACE DESIGN AND CULTURE 65

tended to classify by interdependence within wholes, to rely on subjective

experience without sharp differences between oneself and others or be-
tween facts and concepts.
They investigated sorting the contents of one’s house and found contrast-
ing sorting styles. The Chinese tend to sort by thematic groupings whereas
the U.S. users tend to sort by functions. They found that when users were
given sorting tasks and the culturally appropriate sorting methods, each had
the lowest error rates, and in contrast with the wrong kind of tool, high error
rates. Furthermore, test results of Chinese users using their preferred the-
matic structures show better memory performance, speed, and accuracy. The
Chinese also performed better using concrete metaphors. These investiga-
tions are in general accordance with the heuristic analysis provided earlier
and seem to have strong implications for significant performance differences
of users given culturally inappropriate UI components.
Robert Cialdini, professor at the University of New Mexico, has written
about the dimensions of persuasion (Cialdini, 2001) by which people con-
vince others to think or act in a particular way. He identifies the following
dimensions of persuasion, that is, factors that might persuade one to follow
another’s request:

• Authority: One might act out of respect or fear of another’s author-

ity, power, or position.
• Consistency: One might act out of belief in the consistent following
of rules of legal or sanctioned behavior under the conditions of the
request and appropriate rules of behavior.
• Liking: One might act out of consideration of positive personal
feelings toward the person making the request.
• Reciprocation: One might act in consideration of one’s own requests
in the past or desired reciprocity when making future requests.
• Scarcity: One might act out of perceived scarcity of referred-to ob-
jects, persons, and the perceived benefits of following the request.
• Social validation: One might act out of consideration of value, sta-
tus, or other social benefit accrued by following the request.

Although several of these dimensions combine in any situation to lead to

someone’s actions, Cialdini explains that different cultures typically em-
phasize one factor over others. For example, he comments that an Ameri-
can might decide to honor a request from another person in an office based
on whether that person had recently done him or her a favor, whereas a
German would consider the rules regarding the request, a Spaniard would
consider whether the person is a friend, and a Chinese office worker would
consider the person’s authority to issue the request. Although Cialdini does
not consider Web site persuasion and does not elaborate extensively on cul-
66 MARCUS

tural differences that might affect e-commerce and m-commerce, his analy-
sis clearly opens the door to such considerations.
As an example of additional dimensions in relation to UI design and cul-
ture, consider Bailey, Gurak, and Konstan (2001), who developed a set of
dimensions for establishing trust on the Web. They assert that the following
are the key dimensions:

• Attraction: Attractive people are trusted more.

• Dynamism: Greater activity builds trust (e.g., moving hands, text).
• Expertness: Relevant skills are trusted.
• Faith: Belief in a predictable future builds trust.
• Intentions: Revealed objectives and goals are presumed to be hon-
est and are trusted.
• Localness: Local personalities, sites, terminology, references are
presumed to have similar values and behavior; hence, they are
trusted more.
• Reliability: Dependable, predictable, consistent behavior or out-
comes promote trust.

Bailey (personal communication, June, 2001) commented that he feels

these dimensions vary in their importance among different cultures glob-
ally, and need to be examined in the light of these differences in order to
predict the influence of each within varying cultures.
In Frames of Mind, The Theory of Multiple Intelligences (1985), Gardner
presents his theory of the dimensions of intelligence, considers much edu-
cation and communication biased to verbal comprehension, and urges
greater diversity in communication, evaluation, and learning. His dimen-
sions are the following:

• Verbal/Image comprehension.
• Word/image fluency.
• Numerical/graphical fluency.
• Spatial visualization.
• Associative memory.
• Perceptual speed.
• Reasoning.
• Image: Self/Other awareness.

It seems reasonable that different cultures might place different empha-

ses on which dimensions are valued within that culture and emphasized in
education, at work, and at home. A future task for UI design theory and
practice is to explore how UI components, and specific patterns of combi-
3. USER INTERFACE DESIGN AND CULTURE 67

nations, would relate to these dimensions of intelligence and to the varying

roles given to them by different cultures.
Richard Nisbett, a social psychologist from the University of Michigan,
has studied how culture molds habits of thought. He uses examples of the
differences in descriptions of what Japanese and North Americans observe
in describing simple scenes, for example, what happens in a fish tank
(Nisbett, 2001, p. 6). The Japanese observers note relations among the fish,
the aquatic flora, and the simple environment. North American observers,
on the other hand, tend to focus only on the actions and attributes of the
primary fish. The contrast is stark: One culture is much more oriented to
figure-field relationships; the other culture concentrates on the figure. Sim-
ilar differences have been noted by Kaiping Peng at the University of Cali-
fornia at Berkeley. Peng (2000) noted that Chinese students seem less eager
to resolve contradictions than U.S. students. Cultural factors seem to run
deeply in mental processes. One possible conclusion is that Western schol-
ars may be in error to assume that there is, or should be, a universal focus on
reasoning, categorization, and linear cause-and-effect explanations of situ-
ations and events. Such conclusions differ markedly from classical assump-
tions that would affect the design of UIs.
David Landes, professor of history at Harvard University, argues in
“Culture Makes Almost All the Difference” (2000) that social attitudes are
more important than politics and economics in determining why some soci-
eties are richer than others. This version of the thesis that culture matters
greatly is at odds with assumptions of market economists and liberal and
Marxist philosophers, who believe that political and economic factors are of
primary importance. Hofstede (1997) has a more mixed approach. He rec-
ognizes that both cultural as well as political/economic factors determine
whether certain cultures are dominant at different points in history.
Hofstede comments that cultural orientations are extremely deeply em-
bedded in cultures over hundreds and thousands of years; consequently, he
feels that even modern communication media have not dislodged these cul-
tural orientations. In fact, cultures tend sometimes to concentrate harder
on preserving their approaches given the encroachment of alien cultures.
On the other hand, recent news events seem to argue otherwise. For exam-
ple, a recent article comments that Japan’s 10-year economic decline has
stimulated the rise of individualism (Ono & Spindle, 2000).

CONCLUSIONS

The design of products and services for the Web and mobile devices fosters
the need for good cross-cultural communication in UI design. Eng-
lish-speaking countries constitute 8% of the world’s population, but by
68 MARCUS

2005, approximately 75% of Internet users will be non-English speaking. It

seems likely that cultural factors will need to be considered more fre-
quently. Already, 80% of corporate Web sites in Europe offer more than
English even though launching multilanguage Web site portals with 11 Eu-
ropean languages is a significant burden to operations.
Consequently, cross-cultural analysis and design issues need to be con-
sidered more integrally in planning stages, and developers need checklists
and guidelines to assist them in their design phases. Having a better under-
standing of the mappings of culture dimensions to UI components, as well
as to such dimensions as trust or intelligence, may inform designers to make
better decisions about usability, aesthetics, and emotional experience. Ulti-
mately, computer-assisted tools may result. As indicated earlier, increased
performance, preference, and enjoyment seem likely outcomes of cultur-
ally appropriate UI design.

ACKNOWLEDGMENTS

I acknowledge the assistance of Emilie Gould, cited earlier, in organizing

and conducting the tutorials, and the tutorial participants, through which
portions of the material in this chapter were developed. I also acknowledge
two works of mine (Marcus, 2001, 2002), edited portions of which appear in
this chapter.

REFERENCES

Bailey, B. P., Gurak, L. J., & Konstan, J. A. (2001). An examination of trust production in
computer-mediated exchange. Paper presented at the seventh Human Factors and
the Web 2001.
Cialdini, R. (2001, February). The science of persuasion. Scientific American, 284(2),
76–81.
De Mooij, M. (2001). Internet and culture. Paper presented at. Internet, Economic
Growth and Globalization organized by Institute for International and Regional
Economic Relations, Gerhard-Mercator University, Duisburg, Germany, 8 Au-
gust 2001.
Carroll, J. M. (1999). Using design rational to manage culture-bound metaphors for
international user interfaces. Proceedings, International Workshop on International
Products and Services, Rochester, NY, pp. 125–131.
Choong Y., & Salvendy, G. (1998). Designs of icons for use by Chinese in mainland
China. In Interacting with computers. International Journal of Human–Computer
Interaction, 9:4, February 1998, pp. 417–430. Amsterdam: Elsevier.
Choong, Y., & Salvendy, G. (1999). Implications for design of computer interfaces
for Chinese users in Mainland China. The International Journal of Human-Com-
puter Interaction, 11:1 1999, pp. 29-46. Amsterdam: Elsevier. Cialdini, Robert.
3. USER INTERFACE DESIGN AND CULTURE 69

(2001). The Science of Persuasion, Scientific American, Vol, 284, No. 2, February
2001, pp. 76–81 (www.influenceatwork.com).
Clausen, C. (2000). Faded mosaic: The emergence of post-cultural America, Ivan R.
Dee.
Eco, U. (1976). A theory of semiotics. Bloomington: Indiana University Press.
Gardner, H. (1985). Frames of mind: The theory of multiple intelligences. New York: Basic
Books.
Gelerntner, D. (2000). The second coming: A manifesto. http://www.edge.org/docu-
ments/archive/edge70.html
Hofstede, G. (1997). Cultures and organizations: Software of the mind. New York:
McGraw-Hill.
Lakoff, G., & Johnson, M. (1980). Metaphors we live by. Chicago: University of Chi-
cago Press.
Landes, D. (2000). Culture makes all the difference. In L. E. Harrison & S. P. Hun-
tington (Eds.), Culture matters: How values shape human progress (pp. 2–13). New
York: HarperCollins.
Lévi-Strauss, C. (2000). Structural anthropology. Trans. Claire Jacobson and Brooke
Schoepf. New York: Basic Books.
Marcus, A. (1998). Metaphors in user-interface design. Journal of Computer Documen-
tation, 22(2), 43–57).
Marcus, A. (2001). Cross-cultural user-interface design. In M. J. Smith & G.
Salvendy (Eds.), Proceedings of the Human–Computer Interface International Confer-
ence (Vol. 2, pp. 502–505). Mahwah, NJ: Lawrence Erlbaum Associates.
Marcus, A., & Gould, E. W. (2000). Crosscurrents: Cultural dimensions and global
web user-interface design. Interactions, (7)4, 32–46.
Nisbett, Richard E., Kaipeng Peng, Incheol Choi, & Ara Norenzayan (2001). Culture
and systems of thought: Holistic vs. analytical cognition, Psychological Review,
108, 291–310.
Oliver, R. T. (1971). Communication and culture in ancient India and China. Syracuse,
NY: Syracuse University Press.
Ono, Y., & Spindle, B. (2000, December 30). Japan’s long decline makes one thing
rise: Individualism. Wall Street Journal, p. A1.
Peirce, C. S. (1933). Existential graphs. In C. Hartshorne & P. Weiss (Eds.), Collected
papers of Charles Sanders Peirce (Vol. 4, pp. 293–470). Cambridge, MA: Harvard
University Press.
Peng, K. (2000). Readings in cultural psychology: Theoretical, methodological and empirical
developments during the past decade (1989–1999). New York: Wiley.
Tidwell. J. (1999). Common ground: A pattern language for human–computer interface de-
sign. Retrieved February 20, 2002, from www.mit.edu/~jtidwell/interaction_pat-
terns.html

Additional Useful References

Atchison, Jean. (1996). The atlas of languages: The origin and development of languages
throughout the world, Facts On File. New York, 1966.
70 MARCUS

Alvarez, G. M., Kasday, L. R., & Todd, S. (1998). How We Made the Web Site Interna-
tional and Accessible: A Case Study. Proceedings (CD-ROM), Fourth Human Factors
and the Web Conference, Holmdel, New Jersey.
American Institute of Graphic Arts (AIGA). (1981). Symbol signs. Visual Communica-
tion Books. New York: Hastings House.
Associated Press. (2001). Saudi Arabia issues edict against Pokemon, San Francisco
Chronicle, 27 March 2001, p. F2.
Barber, W., & Badre, A. (1998). Culturability: The merging of culture and usability.
Proceedings of the Fourth Human Factors and the Web Conference.
Bliss, C. K. (1995). Semantography. Sidney, Australia: Semantography Publica-
tions.
Campbell, George L. (2000). Concise compendium of the world’s languages. Sandpoint,
ID: MultiLingual Computing and Technology.
Comrie, Bernard, Stephen Matthews, and Maria Polinsky (Eds.). (2000). The Atlas of
languages. Sandpoint, ID: MultiLingual Computing and Technology.
Coriolis Group. (1998). How to build a successful international web site. Scottsdale, AZ:
Coriolis Group.
Cox, Jr. Taylor (1994). Cultural diversity in organizations, San Francisco:
Berrett-Koehler Publishers.
Crystal, David. (1987). The Cambridge Encyclopedia of Language. Cambridge: Cam-
bridge University Press.
Daniels, Peter T., & William Bright. (Eds.). (2000). The World’s Writing Systems.
Sandpoint, ID: MultiLingual Computing and Technology.
Day, Donald L., Elisa M. del Galdo, and Girish V. Prabhu. (Eds.). (2000). Designing
for Global Markets 2, Second International Workshop on Internationalisation of
Products and Systems, 13–15 July, Baltimore, MD. Rochester, NY: Backhouse
Press.
Day, Donald. (2000). Gauging the Extent of Internationalization Activities. In the
Proceedings of the Second International Workshop on Internationalization of Products
and Systems, 13–15 July 2000, Baltimore, MD, pp. 124–136. Rochester, NY: Back-
house Press.
DelGaldo, E., & Nielsen, J. (Eds.) (1996). International user interfaces. New York: John
Wiley and Sons.
Doi, Takeo. (1973). The anatomy of dependence. New York: Kodansha Interna-
tional.
Doi, Takeo. (1986). The anatomy of self: The individual versus society. New York:
Kodansha International.
Dreyfuss, Henry. (1966). Symbol sourcebook. New York: Van Nostrand Rhinehold.
Elashmawi, Farid, and Harris, Philip R. (1998). Multicultural management 2000:
Essential cultural insights for global business success. Houston: Gulf Publishing
Co.
Fernandes, T. (1995). Global interface design: A guide to designing international user inter-
faces. Boston: AP Professional.
French, Tim, and Smith, Andy. Semiotically enhanced web interfaces: Can semiotics
help meet the challenge of cross-cultural design? In the Proceedings of the Second
International Workshop on Internationalization of Products and Systems, 13–15 July,
Baltimore, MD, pp. 23–38. Rochester, NY: Backhouse Press.
3. USER INTERFACE DESIGN AND CULTURE 71

Frutiger, A. (1997). Signs and symbols: Their design and meaning. Cham, Switzerland:
Syndor Press GmbH.
Goode, Erica. (2000). How culture molds habits of thought. New York Times, 8 August
2000, p. D1 ff.
Gould, Emilie. W., Zakaria, Norhayati, and Shafiz Affendi Mohd. Yusof. (2000). Ap-
plying culture to website design: A comparison of Malaysian and US websites. In
the Proceedings of the IPCC/SIGDOC Conference, Boston, 25 September 2000.
Graham, Tony. (2000). Unicode™: A Primer. Sandpoint, ID: MultiLingual Com-
puting and Technology.
Hall, E. (1969). The hidden dimension. New York: Doubleday.
Harel, Dan, and Girish Prabhu. (1999). Global User Experience (GLUE), Design for
cultural diversity: Japan, China, and India. Proceedings of the First International
Workshop on Internationalization of Products and Systems, 20–22 May 1999, pp.
205–216. Rochester, NY: Backhouse Press.
Harris, John, and McCormack, Ryan. (2000). Translation is Not Enough: Consider-
ations for Global Internet Development. San Francisco: Sapient. Retrieved Feb-
ruary 18, 2002, from http://www.sapient.com/pdfs/strategic_viewpoints/
sapient_globalization.pdf
Harris, Philip R., and Moran, Robert T. (1993). Managing cultural differences: High
performance strategies for a new world of business. Houston, TX: Gulf Publishing Co.
Helander, Martin G., et al. (Eds.). (1997). Handbook of human–computer interaction.
Amsterdam, Netherlands: Elsevier Science.
Hendrix, Anastasia. (2001). The nuance of language. San Francisco Chronicle, 15
April, 2001, p. A10.
Herskovitz, Jon. (2000). J-Pop Takes Off: Japanese music, movies, TV shows enthrall
Asian nations. San Francisco Chronicle, 26 December 2000, p. C2.
Hoft, Nancy L. (1995). International technical communication: How to export information
about high technology. New York: John Wiley and Sons, Inc.
Honold, Pia. (1999). Learning how to use a cellular phone: Comparison between
German and Chinese users. Journal of Society of Technical Communication, 46(2),
May 1999, pp. 196–205.
International Organization for Standardization (ISO). (1996). ISO 7001: Public In-
formation Symbols of The American National Standards Institute. Geneva: ISO.
International Organization for Standardization (ISO). (1993). ISO 7001: Public In-
formation Symbols: Amendment 1 of The American National Standards Institute.
Geneva: ISO.
International Standards Organization (ISO). (1989). Computer Display Color: Draft
Standard Document 9241-8, Geneva: ISO.
Iwayama, M., Tokunaga, T., & Tanaka, H. (1990). A method of calculating the mea-
sure of salience in understanding metaphors. In the Proceedings of the American As-
sociation for Artificial Intelligence (AAAI), 1990, pp. 298–303.
Kimura, Doreen. (2002). Sex differences in the brain. Scientific American, 267(3), pp.
118–125.
Konkka, Katja, and Koppinen, Anne. (2000). Mobile devices: Exploring cultural
differnces in separating professional and personal time. In the Proceedings of the
Second International Workshop on Internationalization of Products and Systems, 13–15
July, Baltimore, MD, pp. 89–104. Rochester, NY: Backhouse Press.
72 MARCUS

Kurosu, M. (1997), Dilemma of usability engineering. In Salvendy, G., Smith, M.,

and Koubek, R. (Eds.): Design of Computing Systems: Social and Ergonomics Consider-
ations (Vol. 2), Proceedings of the Seventh International Conference on Hu-
man–Computer Interaction. San Francisco, USA, August 24–29, 1997, pp.
555–558. Amsterdam: Elsevier.
Lanham, R. A. (1991). A handlist of rhetorical terms (2nd ed.). Berkeley: University of
California Press.
Leventhal, L., et al. (1996). Assessing user interfaces for diverse user groups: Evalua-
tion strategies and defining characteristics. Behaviour and Information Technology,
15(3), pp. 127–139.
Lewis, Richard. (1991). When cultures collide. London: Nicholas Brealey.
Lingo Systems. (1999). The guide to translation and localization. Los Alamitos, CA:
IEEE Computer Society.
The Localization Industry Standards Association (USA). (1999). The localization in-
dustry primer. Féchy, Switzerland: LISA.
Marcus, Aaron. (1983). Visual rhetoric in a pictographic-ideographic narrative. In
the Proceedings of the Second Congress of the International Association for Semiotic
Studies, Vienna, July 1979. Tasso Borbé (Ed.), 3(6), pp. 1501–1508. Berlin: Mou-
ton Publishers,
Marcus, Aaron. (1992). Graphic design for electronic documents and user interfaces. Read-
ing: Addison-Wesley.
Marcus, Aaron. (1993). Human communication issues in advanced UIs. Communica-
tions of the ACM, 36 (4), pp. 101–109.
Marcus, Aaron. (1993). Designing for diversity. In the Proceedings of the 37th Human
Factors and Ergonomics Society. Seattle, Washington, 11–15 October 1993, Vol. 1,
pp. 258–261.
Marcus, Aaron. (1995). Principles of effective visual communication for graphical
user-interface design. In Baecker, Grudin, Buxton and Greenberg (Eds.), Read-
ings in Human–Computer Interaction (2nd ed., pp. 425–441). Palo Alto, CA: Mor-
gan Kaufman.
Marcus, A. (2000). International and intercultural user interfaces, in User interfaces
for all, ed. Dr. Constantine Stephanidis, Lawrence Erlbaum Associates Publishers,
New York, 2000, pp. 47–63.
Marcus, Aaron, and Chen, Eugene. (2002). Designing the PDA of the future. Interac-
tions, 9:1, pp. 32–44.
Marcus, Aaron (2002). User-interface design and culture dimensions, Workshop on
Internationalization, Proceedings. CHI 2002, Minneapolis, MN, 21-24 April
2002. in press.
Marcus, Aaron, et al. (1999). Globalization of user-interface design for the web,
Proc., 5th Conference on Human Factors and the Web, 3 June 1999, NIST,
Gathersburg, MD, avail. from www.tri.sbc.com/hfweb.
McLuhan, M. (1964). Understanding media: The extensions of man. New York:
McGraw-Hill.
Neale, D. C. and John M. C. (1997). The role of metaphors in user-interface design.
In Helander, M., Landauer, T. K. and Prabhu, P. (Eds.). Handbook of human–com-
puter interaction (2nd ed), pp. 441–462. Amsterdam: Elsevier Science.
3. USER INTERFACE DESIGN AND CULTURE 73

Neustupmy, J. V. (1987). Communicating with the Japanese. Tokyo: The Japan Times.
Nielsen, J. (Ed.) (1990). Designing user interfaces for international use (Vol. 13). Ad-
vances in human factors/ergonomics, Amsterdam: Elsevier Science.
Nisbett, R. E. (2003). The geography of thought: How Asians and Westerners think differ-
ently … and why. New York: Free Press.
Lee, Ook, The Role of cultural protocol in media choice in a confucian virtual work-
place, IEEE Transactions on Prof. Comm., Vol. 43, No. 2, June 2000, pp. 196–200.
Olgyay, N. (1995). Safety symbols art. New York: Van Nostrand Reinhold.
Ota, Yukio. (1987). Locos: Lovers communications system (in Japanese), Pictorial Insti-
tute, Tokyo, 1973. The author presents the system of universal writing that he in-
vented.
Ota, Yukio, Pictogram design. Tokyo: Kashiwashobo.
Perlman, Gary. (2000). ACM SIGCHI Intercultural Issues. In the Proceedings of the
Second International Workshop on Internationalization of Products and Sysems, 13–15
July, Baltimore, MD, pp. 183–195. Rochester, NY: Backhouse Press.
Pierce, T. (1996). The international pictograms standard. Cincinnati, OH: ST Publica-
tions.
Pirkl, James J. (1994). Transgenerational design: Products for an aging population. New
York: Van Nostrand Reinhold.
Prabhu, G. V., Chen, B., Bubie, W., and Koch, C. (1997). Internationalization and lo-
calization for cultural diversity. In Salvendy, G., Smith, M., & Koubek, R. (Eds.):
Design of Computing Systems: Cognitive Considerations (Vol. 1), Proceedings of the
Seventh International Conference on Human–Computer Interaction, San Fran-
cisco, USA, August 24–29, 1997, pp. 149–152. Amsterdam: Elsevier.
Prabhu, Girish and Dan Harel. (1999). GUI Design preference validation for Japan
and China: A Case for KANSEI Engineering? In the Proceedings of the Eighth Inter-
national Conference on Human–Computer Interaction, Munich, Germany 22–26 Au-
gust 1999, pp. 521–525. Amsterdam: Elsevier.
Prabhu, Girish V., and delGaldo, Elisa M. (Eds.). (1999). Designing for global mar-
kets 1, Proceedings of the First International Workshop on Internationalization of
Products and Systems, 20–22 May 1999, Rochester, NY. New York: Backhouse Press.
Rowland, Diana. (1985). Japanese business etiquette. New York: Warner Books.
Shahar, Lucy, and Kurz, David. (1995). Border crossings: American interactions with Is-
raelis. Yarmouth, Maine: Intercultural Press.
Sheppard, Charles, and Jean Scholtz (1999). The effects of cultural markers on
website use. In the Proceedings of the Fifth Conference on Human Factors and the
Web.
Stephanidis, C., et al. (Eds.). (1998). Toward an information society for all: An Inter-
national R&D agenda. International Journal of Human–Computer Interaction, 10(2),
pp. 107–134.
Stephanidis, Constantine. (Ed). (2000). User interfaces for all. New York: Lawrence
Erlbaum Associates Publishers.
Stille, Alexander. (2001). An old key to why countries get rich: It’s the culture that
matters, some argue anew. New York Times, 13 January 2001.
Storti, Craig. (1994). Cross-cultural dialogues: 74 brief encounters with cultural difference.
Yarmouth, ME: Intercultural Press.
74 MARCUS

Tannen, Deborah. (1986). That’s not what I meant! New York: Ballantine Books.
Tannen, Deborah. (1990). You just don’t understand: Women and men in conversation.
New York: William Morrow and Company.
Trompenaars, F., & Hampden-Turner, C. (1998). Riding the waves of culture. New
York: McGraw-Hill.
Vickers, Ben. (2000). Firms push to get multilingual on the web. Wall Street Journal,
22 November 2000, p. B11A.
Victor, David A. (1992). International business communication. New York:
HarperCollins Publishers.
Waters, Crystal. (1997). Universal web design: Reach more viewers, improve site appear-
ance, and enhance your marketability through efficient design. Indianapolis, Indiana:
New Riders Publishing.
Yardley, Joanathan. (2000). Faded mosaic nixes idea of ‘cultures’ in U.S. San Fran-
cisco Examiner, 7 August 2000, p. B3.
Yeo, Alvin Wee. (2001). Global-software development lifecycle: An exploratory
study. In the Proceedings of Special Internet Group for Computer–Human Interaction
(SIGCHI), 31 March–5 April 2001, Seattle, WA, pp. 104–111.

URLs and Other Resources

Selected URLs and other resources from a list maintained by the author
and available by contacting him at <Aaron.Marcus@AMandA.com> are
the following:

ACM/SIGCHI Intercultural Issues database: www.acm.org/sigchi/intercultural/

ACM/SIGCHI Intercultural listserve: chi-intercultural@acm.org. Moderator: Don-
ald Day <d.day@acm.org.>.
African-American Websites: www.bet.com, www.netnoir.com, www.blackfamilies.com
Bibliography of Intercultural publications: www.HCIBib.org//SIGCHI/Intercultural
Color: www.colortool.com
Cultural comparisons: www.culturebank.com
Culture resources: www.webofculture.com, www.acm.org/sigchi/intercultural/
Digital divide: www.digitaldivide.gov, www.digitaldivide.org, www.digitaldividenet
work.org
Global business: www.globalbusinessmag.com
Glossary, six languages: www.bowneglobal.com/bowne.asp?page=9&language=1
Indian culture: www.indiagov.org/culture/overview.htm
Internationalization providers: Basis Technology <www.basistech.com>, CIJ
America <www.cij.com>, www.Logisoft.com
Internationalization resources: www.world-ready.com/r_intl.htm, www.world
ready.com/biblio.htm
Internet statistics by language: www.euromktg.com/globstats/index.html, www.world
ready.com/biblio.htm
Java Internationalization: http://java.sun.com/docs/books/tutori
Localization: www.lisa.org/home_sigs.html
3. USER INTERFACE DESIGN AND CULTURE 75

Localization providers (selected): www.Alpnet.com, www.Berlitz.com, www.global

sight.com, Lernout & Hauspie <www.lhsl.com>, www.Lionbridge. com,
www.Logisoft.com, www.Logos-usa.com, www.translations.com, www.
Uniscape.com
Machine translation providers: Alta Vista’s Babelfish <www.babelfish.altavista.
com>, www.IDC.com, www.e-Lingo.com, Lernout & Hauspie <www.lhsl.com>,
www.Systransoft.com
Microsoft’s global development page: www.eu.microsoft.com/globaldev/fareast/
fewinnt.asp
Microsoft Planning for and testing global software: http://www.microsoft.
com/GLOBALDEV/Non%20mirror/back%20up/gbl-gen/INTREFNEW.HTM
Microsoft Windows Internationalization: http://www.microsoft.com/globaldev/
gbl-gen
Native-American-oriented Website: www.hanksville.org/NAresources/
Simplified English: userlab.com/SE.html
Internet users survey, Nua: www.nua.ie/surveys/how_many_online
Unicode: www.unicode.org/, IBM Unicode Glossary: www-4ibm.com/software,/de-
veloper/library/glossaries/unicode.html
Women: wow.com, oxygen.com, chickclick.com

Resources

This section lists organizations providing information about international

standards and intercultural issues.

• ACM SIGCHI Intercultural Issues Database

Gary Perlman maintains the ACM SIGCHI Intercultural Issues Database,
which is summarized in (Perlman).

Gary Perlman, OCLC Online Computer Library Center

5565 Frantz Road, Dublin, OH 43017
Email: Gary Perlman <perlman@TURING.ACM.ORG>
URL: http://www.acm.org/sigchi/intercultural/

• American National Standards Institute (ANSI)

This organization analyzes and publishes U.S. standards, including those
for icons, color, terminology, user interfaces, and so on. Their contact data
are the following:

American National Standards Institute (ANSI)

11 West 42nd Street, 13th Floor, New York, NY10036
Tel: 212-642-2000
76 MARCUS

Email: info@ansi.org
URL: www.ansi.org

• China National Standards Organization

This organization analyzes products to be imported into China. The con-
tact data are the following:

China Commission for Conformity of Electrical Equipment

(CCEE) Secretariat
Postal address: 2 Shoudu Tiyuguan, NanLu, 100044, P.R. China
Office address: 1106, 11th floor, 2 Shoudu Tiyuguan, NanLu,
Beijing, P.R.China
Tel: +86-1-8320088, ext. 2659, Fax: +86-1-832-0825

• East-West Center
This organization, formerly funded by the U.S. Congress, is a center for
technical and cultural interchange among Pacific Rim countries. The cen-
ter’s research and publications cover culture and communication. The con-
tact data are the following:

East-West Center
1601 East-West Road, Honolulu, HI 96848-1601
el: 808-944-7111, Fax: 808-944-7376
Email: ewcinfo@ewc.hawaii.edu
URL: http://www.ewc.hawaii.edu

• Information Technology Standards Commission of Japan (ITSCJ)

Information Processing Society of Japan

Kikai Shinko Building, No. 3-5-8 Shiba-Koen, Minato-ku,

Tokyo 105, Japan
Tel: +81-3-3431-2808, Fax: +81-3-3431-6493

• Institute for International Research

This organization conducts research and organizes conferences about glob-
alization and localization:

Institute for International Research, Inc.

2009 N. 14th Street, Suite 303, Arlington, VA 22201
3. USER INTERFACE DESIGN AND CULTURE 77

Tel: +1-703-908-9010, Fax: +1-703-908-9018

URL: www.iir-ny.com

• International Standards Organization (ISO)

This organization analyzes and publishes world standards for all branches
of industry and trade, including standards for icons, color, terminology,
user interfaces, etc. Their contact data are the following:

International Standards Organization (ISO)

Geneva, Switzerland
Tel: +41-22-749-0111, Fax: +41-22-733-3430
Email: central@iso.ch
URL: http://www.iso.ch/

Of special interest is ISO 8601’s international time and date standards.

Information about ISO’s 8601 standards, and the particular one about time
and date standards, may be found at the following URLs:

http://www.aegis1.demon.co.uk/y2k/y2kiso.htm
http://www.roguewave.com/products/resources/exchange/iso8601.html

• Japan National Standards Organization

These organizations analyze and publish Japanese standards, including
those for icons, color, terminology, user interfaces, and so forth. Their con-
tact data are the following:
Japanese Industrial Standards Committee (JISC)
Agency of Industrial Science and Technology
Ministry of International Trade and Industry
1-3-1, Kasumigaseki, Chiyoda-ku, Tokyo 100, Japan
Tel: +81-3-3501-9295/6, Fax: +81-3-3580-1418

• Localization Industry Standards Organization (LISA), The

This organization provides information and publications relevant to local-
ization professionals:
The Localization Industry Standards Association
7, rue du Monastère, 1173 Féchy, Switzerland
Tel: +41-21-821-32-10, Fax: +41-21-821-32-19
URL: http://www.lisa.org
78 MARCUS

• Web of Culture, The

This Web site publishes news and resources related to cross-cultural com-
munication:
Ms. Eileen Sheridan, President
113 Anita Road, Suite 6, Burlingame, CA 94010
Tel/Fax: +1 408 273 6074
URL: <http://www.webofculture.com>

• World Wide Web Consortium

This organization provides information relevant to globalization issues,
including accessibility. Two URLs of interest are:
http://www.w3.org/International, for information about internalization.
http://www.w3.org/WAI, for information about accessibility.
 Chapter "

Synthesizing the Literature

on Cultural Values

Emilie W. Gould
Rensselaer Polytechnic Institute, Troy NY

INTRODUCTION

I was having lunch with a Malaysian friend who was having problems with
her doctoral advisor. “When I say I want to study culture and information
systems, he says, ‘That’s too broad!’ When I ask him what he means, he
says, ‘Culture is everywhere—national, ethnic, religious, corporate, fam-
ily—but it’s not consistent. You can’t make useful generalizations based on
culture. Even if you could, it would be impossible to develop separate in-
terfaces for each different culture. Focus on the psychology of the individ-
ual user instead!’”
I said, “Yeah, I’ve had that problem too. A lot of people in the field of
human–computer interaction (HCI) really resist the notion of culture be-
cause it seems like such a vague, anthropological concept. When HCI first
evolved from experimental psychology, you could run a test in the usabil-
ity lab and claim to know what every novice user might need. Then people
designing cooperative work discovered qualitative research. But HCI is
just coming to grips with culture. I tell people, ‘I’m applying a set of
well-accepted management theories to people’s design preferences on the
Internet.’ They like that better.”
Legitimating the study of culture has been a problem for anyone inter-
ested in the intersection of national difference, personal preference, and

79
80 GOULD

computer interfaces. We all know individuals from other cultures who seem
more like us than like their compatriots. They make it hard to deny the
claim that “People are where the interface is, not culture!” But you can say
the same for any social science. People are different, yet general psychologi-
cal profiles have been developed. Organizations vary, but in a regular way.
Countries can be grouped together in useful typologies—democratic, so-
cialist, authoritarian, and so forth. Within cultures, individuals cover a wide
spectrum of belief and behavior but, in the aggregate, they cluster together
and these clusters display a surprising amount of stability.
So, what is the best way to analyze national culture to improve human in-
terface design? There are many ways of thinking about the problem, but
generally some theories stress how cultures differ while others examine
what they have in common—the “emic/etic” distinction of linguist Kenneth
L. Pike (1954). Pike contrasted the universality of phonetics (language
sounds) with the diversity of phonemics (interpretations of those sounds as
coherent systems of communication). His analysis has been widely applied
to social phenomena like writing and rhetoric that are based on a universal
system like the alphabet but require selective interpretation by “insiders.”
Cultural anthropology can provide useful insights into designing inter-
faces for specific countries, but theories from the field of intercultural com-
munication are generally better for culturally diverse audiences. Most
designers do not have the mandate to develop entirely different products
for each national or ethnic market. Intercultural communication theory
makes it possible for them to focus on a few crucial variations. Since the
1990s, when the concept of culture finally emerged as a concern for inter-
face design, most design ideas have come from intercultural communica-
tion, most notably, in articles and workshops by Aaron Marcus (1993, 2001;
Marcus & Gould, 2000) and in the essays in Elisa M. del Galdo and Jakob
Nielsen’s International User Interfaces (1996).
For the rest of this chapter, I (a) further discuss the advantages of
intercultural communication theories for HCI design, (b) list key communi-
cation theorists whose work has heuristic value for the field, and (c) identify
the most important variables for interface development.

THE ADVANTAGES OF INTERCULTURAL COMMUNICATION

THEORIES FOR HCI DESIGN

Though cultural anthropology seems like it should be valuable for interface

development, it has several limitations as a design guide. First, it tends to
focus on specific cultures to the detriment of comparison and theory and,
second, it often emphasizes authenticity over interaction. Some of these
problems are inherent in its chosen methodology; others derive from its
historical development as an academic field.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 81

Many anthropological studies use “thick description” to record and

highlight behaviors and attitudes that contrast with the anthropologist’s
own culture. Unfortunately, this approach emphasizes a degree of cultural
uniqueness that may make it seem impossible to bridge cultural differences.
Although each culture has its own symbols and patterns of action, not all dif-
ferences are significant. Some of the interface development literature on
cultural markers illustrates this problem. Tony Fernandes (1995) first (de-
lightfully) introduced the problem of translating cultural symbols, taboos,
and aesthetics into interface design but did not identify ways to determine
which symbols were important. Wendy Barber and Albert Badre (1998) in-
troduced the term cultural marker to define interface elements that “signify a
cultural affiliation.” They talked about developing a directory of flags, col-
ors, and national symbols that would allow interfaces to be automatically
recompiled with different cultural markers for different markets. Charles
Sheppard and Jean Scholtz (1999) tried to test this idea, but their results
were not significant. Their effort demonstrates the difficulty of working
with an essentially atheoretical concept.
By contrast, the field of intercultural communication focuses on the reg-
ularities between cultures, identifying a few critical variables that help peo-
ple craft appropriate messages and collaborate. In particular, the literature
on value orientations reduces cultural differences to a manageable number.
In 1950, Florence Rockwood Kluckhohn first systematically com-
pared cultures according to their values. The following year, her hus-
band Clyde Kluckhohn (1951) used the term value orientations in an essay
on theories of action. For the next decade, Harvard University spon-
sored Florence Kluckhohn and Fred L. Strodtbeck in the Values Project.
This longitudinal study compared five distinct communities—Zuni, Na-
vaho, Mexican, Mormon, and Texan—living near one another in the
American Southwest.
Their results were published in 1961 as Variations in Values Orientations. The
book’s most notable contribution was its characterization of each community
on the basis of five existential (and thus presumably universal) questions:

What is the purpose of human nature (and can people Human-nature orientation
change)?
How do people relate to nature (and the supernatural)? Man-nature orientation
How do people manage time? Time orientation
What is the purpose of human activity? Activity orientation
How do people relate to each other? Relational orientation

Three possible variations were assigned to each question. Taken to-

gether, these value orientations allowed Kluckhohn and Strodtbeck to
82 GOULD

identify opportunities for conflict and cooperation and to analyze each

community equitably.
However, few other cultural anthropologists followed their lead.
Through the rest of the decade, many studies continued to emphasize the
distance between “traditional” societies and “evolved” cultures. Eventually,
postcolonialist critics like Edward Said (1978/1994) began to document the
role of academic disciplines in facilitating cultural imperialism. “Oriental-
ism” and its ilk were seen to justify empire by emphasizing the inferiority of
non-Western cultures. Anthropology came under attack for its focus on iso-
lated, marginalized groups.
The field took these criticisms to heart; this crisis reformed its study of
culture. Anthropologists today use a variety of techniques to avoid the un-
equal power relationships inherent in studies of “exotic” people. Members
of the culture collaborate in the research (much as do union members in
Participatory Design); lightly edited (but carefully selected) personal narra-
tives and field notes may be published in place of abstract analysis; and, an-
thropologists increasingly investigate their own cultures.
But such work too often remains unitary and singular. In 1946, Ruth
Benedict said that a culture might share 90% of its observances with its
neighbors, but the differences, “however small in proportion to the whole,
turn its future course of development in a unique direction” (1946/1989,
p. 9.) Postmodernist and postcolonialist studies of “difference” often
make it appear that nothing important is shared except the experience of
oppression. Unfortunately, the moral imperative to avoid stereotyping
and the desire to recognize “authenticity” are antithetical to a utilitarian
focus on collaboration.
Intercultural communication came through a similar crisis but devel-
oped a more useful methodology. The field emerged after World War II
and was shaped by the collapse of Europe, the emergence of the pax
Americana, and political and economic globalization. In the 1940s, the diffi-
culties facing many American diplomats and representatives of non-
governmental organizations led Congress to reorganize the Foreign
Service and found the Foreign Service Institute (FSI). Linguists and anthro-
pologists were hired to provide language and culture training. Historians
and communications researchers collaborated with descriptive linguists
like George L. Trager (who focused on nonverbal paralanguage) and Ray L.
Birdwhistell (who developed kinesics). In 1959, the year after William J.
Lederer and Eugene Burdick’s Ugly American made the best-seller list, Ed-
ward T. Hall proposed a solution to the problem of “culture.”
The Silent Language was based on examples from the FSI and showed how
ignorance of cultural patterns and nonverbal communication underlay nu-
merous cases of intercultural miscommunication. The book was easy to
read and helped explain to millions of Americans why the country was be-
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 83

ing demonized while trying to do good. As a result of its popularity and the
increasingly theoretical focus of his subsequent books, Edward T. Hall is of-
ten described as the founder of the field of intercultural communication. As
he himself claimed, “The complete theory of culture as communication is
new and has not been presented in one place before” (1959/1973, p. 32).
Intercultural communication was soon recognized as a separate com-
munication specialty. Subsequent researchers looked for, and found,
many patterns of cultural variation. Generally these variations were
based on analyses of face-to-face communication between people from
different cultures.
This interpersonal focus makes the field a good fit with HCI. According
to Horton and Wohl’s (1956) theory of parasocial communication and
Reeves and Nass’ (1996) “media equation,” our interactions with “new me-
dia” are just an extension of our face-to-face interactions with one another.
Consequently, theories that deal with communication between people from
different cultures should transfer directly to communication between peo-
ple from one culture using interfaces developed in another.
Hall has been followed by Edward C. Stewart and Milton J. Bennett
(1972); John C. Condon and Fathi S. Yousef (1975); Geert Hofstede (1980;
Hofstede & Bond, 1988); Harry C. Triandis (1975, 1983, 1995); William B.
Gudykunst (1987; Gudykunst et al., 1996); Fons Trompenaars and Charles
Hampden-Turner (1998); and David A. Victor (1992). Next, I review each of
these researchers and discuss ways to apply their theories to interface design.

EDWARD T. HALL: CHRONEMICS, PROXEMICS, AND CONTEXT

Hall’s background preparing diplomats at the Foreign Service Institute

shaped his work. He felt that assignees should be tested for their psycholog-
ical adaptability, trained in the appropriate language of their new country,
and introduced to its formal culture. But, “training in language, history,
government, and customs is only the first step. Of equal importance is an in-
troduction to the nonverbal language of the country” (1959/1973, p. xiii).
Until people understood this “language of behavior,” they could not begin
to be effective, and they needed to understand their own “silent language”
to anticipate their reception.
In The Silent Language, Hall compared American cultural patterns
with those of Japan, Germany, France, Greece, the island of Truk, and
the Middle East. He focused on two nonverbal features that remain im-
portant in intercultural communication theory: time (chronemics) and
space (proxemics).
In Beyond Culture (1976), he identified another factor, which governs
people’s responses to verbal discourse, that is, context.
84 GOULD

Hall’s discussion of chronemics is quite detailed. Time is organized in

two ways. The first is the subjective division of time into technical, formal,
and informal systems. The second is the connection of time with activity.
Technical time reflects the underlying physical context of our experi-
ence of time. The solstices, phases of the moon, sunrise, and atomic de-
cay can all be used to separate time into units. Formal time is often based
on technical time. It consists of conventional systems of measurement
that help people plan, schedule, and manage time. Formal time also
leads to a sense of determinacy—the feeling that events are linked and
causal. However, nothing requires a day to have 24 hours or 7 days to
make a week.
Informal time is cultural and situational. It exists side by side with techni-
cal and formal time and frequently uses the same vocabulary, but it involves
culturally different perceptions of the rate time passes. For instance, if I say,
“I will see you in 5 minutes” in Germany, I need to hurry to meet you. If I say
the same thing in Ecuador, you will not be concerned if I don’t show up for
an hour. Informal time also deals with the significance of events in the past,
present, and future. Hall noted that people in the Middle East and South
Asia are more likely to ground decisions in historic events than are people
in the United States.
The connection between time and activity led Hall to his notions of
monochronic and polychronic time. In monochronic societies like the United
States, Canada, and northern Europe, people do one thing at a time. They
prefer to work sequentially and try to resist interruptions. In polychronic
societies like Latin America, the Middle East, and parts of Asia, people work
on several things at once. This may mean promising to meet different peo-
ple in different places at the same time or working on two or three projects
simultaneously. Monochronic time enhances determinacy but, in poly-
chronic time, things tend to “happen.”
Hall’s second contribution to intercultural theory is proxemics, the social
use of space. Once again, he analyzed the expression of this construct in dif-
ferent cultures in terms of technical, formal, and informal systems. Physical
constraints (mass, edge, volume) underlie technical systems of space; the
process of architecture gives rise to various formal systems; and people in-
terpret informal patterns of proximity and arrangement to determine sta-
tus and group orientation.
With Beyond Culture (1976), Hall developed the idea of context. Although
this construct is also used in literary theory for situating text and verbal dis-
course, Hall wanted to highlight the importance of nonverbal communica-
tion. In high-context societies, features of the social system shape meaning
through chronemics, proxemics, and other nonverbal behaviors; a mes-
sage’s actual text is secondary. In low-context societies, the information is
explicitly expressed in the text of the message.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 85

The United States tends to be low-context; Americans generally say ex-

actly what they want you to understand. Many Asian societies (Japan is the
classic example) are high-context; the message is inherent in the occasion,
the physical setting, and the relationships between the participants. People
within the culture know how to interpret each other’s intent, but many
American business people have learned the hard way that “yes” (said po-
litely but unenthusiastically) really means “no.”

Hall as a Guide for Interface Development

How do Hall’s constructs relate to interface design? First, his definition of

technical, formal, and informal systems can be mapped to cultural percep-
tions of applications and the Web. Computers also have certain technical lim-
its based on hardware and software, but the formal layout of screens (and
sequences of screens) can still be read in different ways. For instance, the “ar-
chitecture” of a web page may differ in terms of color, mass, and balance be-
tween verbal and visual elements. Many American web sites currently sport
subdued hues, arrange content in explicit hierarchies of indentation, and
emphasize text over graphics. Web sites from other countries appear visually
striking if they upset American expectations by using vivid colors, an asym-
metric arrangement, or more graphics than text. Appropriate content may
also vary. Text that lacks formal historical grounding will be unpersuasive to
people who rely on tradition to justify their actions. Moreover, presentation
and content may interact. Pictures that are read individually in one country
can become part of an unintentional collage of meaning in another.
Second, Hall’s notion of time has implications for interaction. Different
applications may express monochronic or polychronic time. A monochronic
application can support a user’s time sensitivity by focusing narrowly on a
task or (somewhat paradoxically) by providing a wide range of choices to re-
duce the search hierarchy. A polychronic application can present less struc-
tured options or support incomplete sequences of tasks. For instance, a user
may look up a train schedule on the Internet but expect to return to the home
page (or go to the station) to complete the transaction.
Third, Hall’s notion of context is important because it says that high-con-
text users may not focus on the explicit verbal message; thus, a developer
cannot depend on written text for clarity. Context may be expressed in the
nonverbal elements of the screen, or context may be out of the developer’s
control and inherent in the status of the software provider or in local proce-
dures that constrain the user’s access to the application. As a result, a web site
for a low-context audience may not be appropriate for a high-context one. A
site may lack credibility because it does not provide branding or information
on the social structure of the sponsoring organization. A site can also fail be-
cause it does not recognize the social status of the user or because it addresses
86 GOULD

a high-status user with insufficient respect. Under certain circumstances, ap-

plications can actually be “too easy.” Lee (2000) found that Korean subordi-
nates prefer to fax messages to their superiors because e-mail did not require
enough effort to show sufficient respect.

EDWARD C. STEWART AND MILTON J. BENNETT:

SUBJECTIVE AND OBJECTIVE CULTURE

Edward C. Stewart and Milton J. Bennett believed that communication style

and cultural values arise from the process of human perception. They wrote
their classic book on American Cultural Patterns (1972/1979) to help U.S. citi-
zens working abroad. Before such sojourners can properly interpret mes-
sages and events in other cultures, they need to recognize how they have been
conditioned by their own. Like Hall, Stewart and Bennett believed that
intercultural communicators must be culturally self-aware to be effective.
Culture has two fundamental aspects. Objective culture is social and ma-
terial, encompassing a society’s political and economic system, aesthetics,
customs, art, architecture, and institutions. Subjective culture is psychologi-
cal, reflecting on society’s values, expectations, theories of action, and pat-
terns of thought. Objective culture is tangible (or at least easily talked
about); subjective culture tends to be experienced unconsciously.
Cultural clashes arise from mismatches in the process of human percep-
tion. Subjective culture can be brought forward into mindfulness by under-
standing how the “deep mind” interprets its environment. An understanding
of the different subjective levels of thought helps identify the sources of cul-
tural miscommunication.
All people have the same senses, but people from different places often
draw different conclusions from the same stimuli. Their minds are condi-
tioned to filter out or ascribe different messages to similar sensory informa-
tion. For instance, people may convert the same stimuli into different
percepts. A small mammal that barks is food for one person, a pet for an-
other. As part of the process of distinguishing figure and ground, different
people foreground different information for their idea of a “dog.”
At the next level of mindfulness, people develop patterns of thinking. Per-
sonal development within a family, community, and culture helps shape their
cognitions. Memory, emotion, stories people have read, and stories they
have heard from others bring to mind dinners in Shanghai or walks in the
park. This is the stage at which “the mind extends its mastery of objective re-
ality by creating a cognitive, or subjective, reality where meaning is assigned
and relationships elaborated” (Stewart & Bennett, 1972, p. 22). Patterns of
thinking apply to all types of sensory data—graphic and nonverbal, as well as
text and language. Screen layout and expectations of interaction are as sub-
ject to the values and assumptions of culture as are words and pictures.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 87

Complex symbol systems make up the deepest layer of the Stewart and
Bennett model. These systems—pictorial, musical, linguistic, and mathe-
matical—give rise to metaphor, grammar, and semantics. People learn
these systems within the larger societies that use them. (They can also be
learned formally but it takes an insider to explain that = “dog.”)
So, how does language influence values? The two are intertwined.
Stewart and Bennett acknowledge a debt to Benjamin Whorf and the
structural linguists. In analyzing patterns of language use in the United
States, they tie grammar to behavior:

• The subject/predicate form of English is linked to a heightened

sense of causality (“it as implied agent”).
• The use of dichotomies and implied negative constructions is
linked to critical thinking.
• Modifier/noun patterns are linked to a preference for process and
precision.

These syntactical elements enhance active, independent behaviors and

work against indirection or collaboration. People who use such language
are likely to be part of or develop an individualistic culture.
Stewart and Bennett make a good case for looking beyond the literal con-
tent of visual imagery and language and grounding cultural analysis in per-
ception and the underlying grammar of symbol systems. More than 30
years later, Richard E. Nisbett (2003) has identified similar cognitive differ-
ences that affect the way Westerners and Asians “read” pictures—and com-
puter screens.

Stewart and Bennett as a Guide for Interface Development

Most computer companies base their internationalization efforts on the

premise that all users perceive computer systems in the same way. If every-
one engages the system at an equivalent semantic level, then language lo-
calization should be sufficient to address national differences. At worst, a
company might be required to change certain metaphors, for example, a
U.S. mailbox becomes a British post box.
But Stewart and Bennett’s distinction between objective and subjective
culture increases the possibility that companies must go beyond surface text
and images. The grammar of the interface is hard to recognize but may be
fundamental in cultivating trust and acceptance. Design elements like
alignment, proximity, repetition, and contrast (Williams, 1994) affect the
massing of text on a page and the overall balance of the screen. Such layout
features should not be ignored when localizing screens because they can
contribute to intercultural miscommunication. For instance, text that is
88 GOULD

read from left to right trains users to look for important information in the
top left and lower right quadrants of a screen. But bidirectional text is read
from right to left and Asian scripts may be read from top to bottom. So
where should the most important information be placed? When translating
a process, the order of the graphics should be reversed.
HCI researchers need to identify similar cultural differences in percep-
tion and interpretation. A grammar of visual rhetoric, like that of Gunther
Kress and Theo van Leeuwen (1996), will not be stable across cultures. An
image of a young man gazing back from the screen may promote engage-
ment with the interface in Britain but intimidate a new user in East Asia.
Cultural messages are embedded throughout the user interface.
Charles Kostelnick (1995) has also emphasized the link between percep-
tion and aesthetics. Interface designers tend to be so embedded in the vi-
sual language of modernism that they are incapable of recognizing their
own bias. It is necessary to understand how people in different cultures ex-
perience an interface to address their hidden (and seemingly irrational)
motivations. Such cultural knowledge must include their artistic prefer-
ences as well.
Nancy Hoft (1996) championed the value of Stewart and Bennett’s
“metamodel” for constructing narrowly targeted cultural models for inter-
face localization. Designers should first build a model of their own culture
to have a standard for comparison. Then, thinking about unconscious levels
of perception will help determine what parts of the interface to focus on.
Fundamental differences in objective and subjective culture will mandate
extensive change in localized screens and content.
By using a model of culture, Hoft promised that designers could identify:

• Global information that can be put into the interface without re-
quiring future translation.
• Cultural bias in the existing application.
• Parts of the application that should be localized for a specific cul-
ture.
• Compelling cultural metaphors and cultural markers.
• Potential cultural problems.
(In addition, cultural models provide guidance for developing ap-
propriate methods for usability evaluation.)

However, Hoft might have taken Stewart and Bennett’s perspective

even further. She talks about unspoken and unconscious rules in her own
“Iceberg” metamodel and highlights Hall’s contributions to decoding
nonverbal communication, but she does not fully apply these insights her-
self. For instance, she recommends identifying international variables for
cultural models by surveys and observation, but the first depends on
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 89

self-report and the second on an observer being able to recognize the trig-
gers for unconscious behavior.

JOHN C. CONDON AND FATHI S. YOUSEF:

EXTENDING THE VALUE ORIENTATIONS APPROACH

In 1975, John Condon and Fathi Yousef published An Introduction to

Intercultural Communication. As their editor noted, this was the first survey to
pull together the diverse themes associated with interpersonal, nonverbal,
and speech communication; combine them with value orientations; and
provide the reader with a single, unified explanation of cultural differences
in social structure, language, and argumentation. (Even Hall developed his
theories piecemeal.)
Condon and Yousef cited everyone from Hall, Bateson, and Goffman (for
interpersonal communication) to Birdwhistell (nonverbal communication
and paralanguage) to Toulmin and Aristotle (rhetoric), Sapir, Whorf, and
Chomsky (linguistics), and McLuhan (mass communication). But the focus of
their book is their extension of Kluckhohn and Strodtbeck’s five existential
questions. They defined 25 value orientations under the headings of:

• Self—individualism/interdependence, age, sex, and activity.

• Family—relational orientations, authority, role behavior, and mobility.
• Society—social reciprocity, group membership, intermediaries, for-
mality, and property.
• Human Nature—rationality, good and evil, happiness, and mutability.
• Nature—relationship of humans to nature, ways of knowing nature,
structure of nature, and concept of time.
• The Supernatural—relationship of humans to the supernatural,
meaning of life, providence, and knowledge of the cosmic order.

These categories are best represented as a Venn diagram (see Fig. 4.1)
with overlapping spheres representing self, society, and nature with the in-
tersections representing family, human nature, and the supernatural.
Condon and Yousef’s (1975) definition of cultural values is flexible. The
number of value orientations is not absolute. There could be more—or fewer:

Each of the value orientations … arises from empirical data obtained from
many cultures. The greater danger is not that of hypothesizing irrelevant
categories but in not including enough. Since the scheme is not intended to
be exhaustive, this is not a major problem; one is always free to add more or,
if desired, to combine categories or to ignore some entirely. (p. 58)

Cultural values develop early in life. Half of the values deal with “what is
thought to be good” and the other half with “what is thought to be true.”
90 GOULD

FIG. 4.1. Condon and Yousef’s model of culture.

Though standards of goodness are often easier to recognize and articulate,

diverging beliefs about truth tend to have more profound consequences for
intercultural communication. Opposing notions of reason, pleasure,
change, and purpose may hinder individual perception and prevent peo-
ple from agreeing on common logics, aesthetics, strategies, or goals.

Condon and Yousef as a Guide for Interface Development

The expansion of Kluckhohn and Strodtbeck’s value orientations to social

structure, language, paralanguage, and argumentation moves their theory
from existentialism into the realm of pragmatic communication. Because
three options are associated with each value, a broad range of behaviors is
described by the 75 heuristics.
Values like interdependence and the use of specialist intermediaries
could have a direct bearing on deciding whether to make agents a promi-
nent part of an interface. Preference for youth versus age, women over men,
or individuals with tighter or looser relationships to their families might in-
fluence representations of users. The activity orientations and relationship
of humans to nature (human domination, harmony, or the supremacy of
nature) should be factored into any decision to develop utilities to change
application defaults and system parameters.
But the greater importance of Condon and Yousef’s conceptual scheme
lies in their emphasis on rhetoric and persuasion. How narrowly should an
application focus on specific tasks? Should an interface be organized de-
ductively or inductively to appeal to people from a different culture? What
kind of information will be persuasive? Once developers recognize that
their own patterns of rationality are not universal, there is the possibility of
designing new interfaces with more appeal to those voluntarily on the op-
posite side of the digital divide.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 91

Condon and Yousef also highlight the importance of communication

roles. People communicate with images of each other: You communicate
with an image of me and I communicate with an image of you. In addition,
my image of “me” also influences my expectations for interaction with your
image of “you.” Walter Ong (1975) once said that an author’s audience is al-
ways a fiction. Interface designers are no less guilty of inventing their users.
In fact, our projections are often hyperpersonal. Joe Walther (1996)
found that unacquainted people communicating at a distance make more
extreme, and generally more positive, judgments of one another than peo-
ple communicating face-to-face. We contribute to the digital divide when
we have unrealistic and uncritical expectations that everyone who uses a
computer must be much like us. Unfortunately, when they are not, their re-
actions can become overly negative. In general, bad experiences diminish
willingness to try new software. Such problems take on a geopolitical di-
mension because computer interfaces and Internet web sites function as
communication partners in themselves. When television was new, Horton
and Wohl (1956) described people’s strong engagement with TV stars and
other media figures. Reeves and Nass (1996) found no difference in classic
social psychology experiments rerun with humans and a computer surro-
gate. People who interact with our software feel that they are interacting
with our culture. Poor usability may be seen as confirmation of a vendor’s
(and a culture’s) lack of respect.
Thus, our interpretation of users from other countries is often based on
unrealistic images of them. And their (often negative) projections of us,
based on their interactions with our products, may be equally unfounded.
With the best of intentions, people in HCI tend to underestimate the effects
of culture even as we struggle to close the digital divide.

GEERT HOFSTEDE: THE DURABILITY

OF THE “SOFTWARE OF THE MIND”

Hofstede’s (1980) theory of universal cultural dimensions has become the

best known, and probably most applied, intercultural communication the-
ory. Various researchers have applied it to leadership, motivation, compen-
sation, and teamwork within domestic and multinational firms, and Nancy
Hoft (1996) used it in her article on developing a cultural model.
Hofstede based his work on a series of surveys from 1968 to 1972 with
IBM employees from 72 national subsidiaries; after factor analysis, four di-
mensions explained half the differences between the 116,000 respondents.
Not only were the factors supposed to be universal, but Hofstede claimed
that they were highly resistant to change because people learned them as
children. Adults might be exposed to opposing values and superficially
adapt, but they could never erase this “software of the mind.”
92 GOULD

Aaron Marcus (chap. 3, this volume) thoroughly describes these di-

mensions:

• Power Distance: the extent to which less powerful people expect and
accept that power is distributed unequally.
• Individualism vs. Collectivism: the extent to which people are inte-
grated into tight social networks.
• Masculinity vs. Femininity: the relative desirability of material success
versus quality of life and of assertive versus modest behavior.
• Uncertainty Avoidance: the extent to which people tolerate ambigu-
ity and risk or feel threatened by change.

In addition, Marcus demonstrates how they can be mapped to a model of

the user interface to generate culturally appropriate designs. (See also
Marcus & Gould, 2000.)
Hofstede’s dimensions have become a touchstone for comparisons of na-
tional differences because of the size and scope of the original survey, but
they have also been attacked. Three major criticisms have been raised
against Hofstede’s four factors:

1. IBM employees are not representative of any national culture. As

employees of a multinational corporation with a conservative (in
those days) corporate culture, they should not have been used to
sample national traits.
2. The factors are not stable. Cultures change over time and even IBM is
not the company it was in 1972 when Hofstede finished his research.
3. Hofstede’s survey was itself biased. Because he developed and
tested his questions in Europe and the United States, and adminis-
tered them in English, they were permeated with Western values.

Hofstede acknowledged that IBMers are not representative but rejected

the implication that they can not be used to identify cultural values. Instead,
he claimed that their very bias makes them a functionally equivalent sam-
ple. That is, IBM’s distinct workplace culture gives employees so much in
common that any differences between them must be the result of differ-
ences in national culture.
Hofstede also rejected the criticism that his factors might change over
time. Instead, he emphasized the stability of childhood acculturation. Re-
cent work by Hofstede fellow and colleague, Marieke de Mooij (2001), reit-
erates this point. But Hofstede did admit that his country rankings omit
some nations (e.g., People’s Republic of China) that have become more im-
portant in the three decades since he did his research.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 93

Hofstede also accepted that his original survey on employee attitudes

does have a Western bias. Michael Bond (a Canadian-born professor in
Hong Kong) convinced him to administer the Chinese Value Survey,
which had been developed by Taiwanese and Hong Kong researchers in
the early 1980s. After reviewing the factor analysis, Hofstede and Bond
(1988) found a fifth cultural dimension: long- versus short-term orienta-
tion (also known as Confucian dynamism, the acceptance of traditional
Confucian values).
The Chinese Value Survey supported the validity (and universality) of
power distance and individualism and collectivism. However, it modified
the definition of masculinity to include more “chivalric” values like pa-
tience, courtesy, kindliness, and compassion.
This redefinition highlights a major problem with the third of
Hofstede’s dimensions. Notions of masculine and feminine behavior vary
widely and are themselves culturally determined. A more useful way to in-
terpret this dimension might be to examine the extent that people are
bound by general social roles. In masculine societies, it is difficult for men
to look after children, but it is also difficult for anyone to change their reli-
gious or political affiliation. In feminine societies, people are less bound
to social roles in general.
More recent research repeats these criticisms of Hofstede’s methodology
while discovering surprising confirmation for his strongest dimensions. For
instance, Shalom Schwartz (1994) published an alternative set of value di-
mensions developed from a new sample of 41 cultural groups in 38 nations.
As shown in Fig. 4.2, his factor analysis juxtaposes two qualities found in
individualistic societies—Intellectual Autonomy (self-direction) and Affec-
tive Autonomy (stimulation and hedonism)—with the Conservatism (con-
cern for security, conformity, and tradition) found in homogeneous
(collectivist) societies. Further, Schwartz was able to identify factors associ-
ated with Self-Enhancement that contrasted with values promoting Self-
Transcendence.
The two Self-Enhancement factors are Hierarchy (the legitimacy of
power hierarchies) and Mastery (self-assertion, self-promotion, and active
control of the social environment)—factors similar to Hofstede’s Power Dis-
tance and Masculinity. The two Self-Transcendence factors are Egalitarian
Commitment (the motivation for altruism and volunteer service in individ-
ualistic societies) and Harmony (with nature)—factors that expand
Hofstede’s notion of Femininity.
Schwartz better explains why individualists often act collectively, for in-
stance, how Sierra Club members in Los Angeles and subsistence farmers
in traditional African villages can share a common environmental per-
spective. But, overall, Hofstede’s dimensions hold up surprisingly well.
94 GOULD

FIG. 4.2. Shalom Schwartz’s culture-level value structure.

Hofstede as a Guide for Interface Development

Hofstede’s tables of cultural differences provide a methodology for antici-

pating cultural differences. Each of the 50 countries (and three regions) in
his original survey is ranked in terms of the original four work orientations,
with another 20 listed for Confucian Dynamism. Thus, if you are a Cana-
dian developing a web site for South Koreans, you can determine that your
audience will have higher power distance, lower individualism, greater
masculinity, higher uncertainty avoidance, and a more long-term orienta-
tion than a domestic audience.
When you associate interface elements with these dimensions, you may see
some overlap. Support for mastery (masculinity) has much in common with
allowing people to learn sites by trial and error (low uncertainty avoidance).
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 95

Hofstede admits that most of his factors are correlated with one or two
others, but considers this redundancy helpful, not problematic. Not all cul-
tural differences are equally critical for each country. Only significantly dif-
ferent cultural orientations warrant attention, though each can percolate
throughout the interface.
For example, Canada and South Korea are quite different, and a Cana-
dian designer might need to develop a radically different site for a Korean
audience. But a German designer modifying a domestic web site for Dutch
customers would discover that the two countries share similar values for
power distance, individualism, and uncertainty avoidance; only masculinity
(Germany) and femininity (the Netherlands) are ranked far apart. As a re-
sult, the German could focus on the representation of gender and social
roles and leave the rest of the site alone.

HARRY C. TRIANDIS: ADDING CULTURE

TO SOCIAL PSYCHOLOGY

In the same decade that Hofstede started researching culture, Harry C.

Triandis began to consider its effects on social psychology. He first exam-
ined the interplay between norms, attitudes, and behaviors. By the end of
the 1960s and early 1970s, he was looking for differences in group identity
to explain people’s actions—analyzing subjective (as opposed to material)
culture in Greece, race in the United States, and social distance between
American, German, and Japanese students. He linked the influence of
“in-groups” and “out-groups” to the issue of culture; in fact, he saw culture
as the ultimate group.
At the time, social psychology was coming to realize that many of its topics
were culture-bound and had little explanatory value. Studies replicated with
people from different ethnic and national backgrounds showed radically dif-
ferent results. Variables often measured trivial amounts of difference.
In 1975, Triandis wrote an influential article that reviewed these prob-
lems and urged the field to focus on higher order constructs, like associa-
tion and dissociation, whose expression could be compared across cultures.
He acknowledged that intercultural studies would be difficult to develop
because they require broad-based collaboration, multiple methods, and
customized test materials for each culture, but he encouraged people to
work cross-culturally.
Shortly thereafter, Triandis promoted models of universal social behav-
iors: “Eventually social psychological theories will be computer programs”
(1978, p. 14). In another article, “Dimensions of Cultural Variation as Pa-
rameters of Organizational Theories,” he urged social scientists to apply
cultural values (most from intercultural communication) to then-current
theories of productivity, job satisfaction, and leadership:
96 GOULD

Organizational theorists naturally develop theories about the part of the so-
cial world that they know and understand. As a result, many such theories are
inappropriate and inadequate when applied to other cultures. We create or-
ganizations that work well for us, and for people like us. But these organiza-
tions are usually a failure for people with a different perspective.… What is
likely to be most effective in one culture is often different from what is most
effective in another. (1983, pp. 140–141, 165).

Triandis’ promotion of these research goals makes him foundational to

the field of cross-cultural psychology. In 1976, he was elected president of
the International Association of Cross-Cultural Psychology (itself founded
only 4 years earlier.) In the 1980s, he edited the multivolume Handbook of
Cross-Cultural Psychology.
Over the last 15 years, he has taken his own advice and focused on a sin-
gle cultural syndrome: individualism and collectivism. (This was one of
Kluckhohn and Strodtbeck’s original value orientations and is the most
common cultural dimension in intercultural communication theory.)
Drawing on philosophy, political science, religion, and history, as well as
communication, anthropology, and psychology, Triandis has examined in-
dividualism and collectivism in depth. Much of his work considers such puz-
zling aspects as the strength of in-groups, the paradox of individualists in
collectivist societies (and vice versa), and interactions with power distance.
In reconceptualizing the dimension, Triandis (1988) asked himself why
people in collectivist societies are disproportionately influenced by their
in-groups. He decided that people in individualistic societies have greater
freedom of action and the opportunity to belong to more groups than peo-
ple in collectivist societies. As a result, no single group dominates them,
though they often spend more time overall conforming to group norms. By
contrast, people in collectivist societies have fewer choices and belong to
fewer groups. Because their societies are also more homogeneous, these
groups (family, religion, and state) tend to hold similar values and reinforce
one another. Consequently, groups exert more control over people’s ac-
tions. People in collectivist societies learn to be cautious and gather a great
deal of information about new individuals and situations because group re-
lationships make such large demands on their time, energy, and autonomy.
However, outside their groups, people can act quite independently.
A year later, noting that people function on an individual (not cultural)
level, Triandis (1989) incorporated a theory of social identity into the di-
mension. People growing up in different cultures have more or less
well-developed private, public, and collective selves that they rely on for
appropriate social responses. In different situations, one self usually dom-
inates and influences the development of personal relationships and pat-
terns of communication. The individual level of individualism is idiocentric
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 97

and the individual level of collectivism, allocentric. These individual-level

values may contrast with the overall culture-level values assigned to a spe-
cific country.
Triandis (1995) even found a way to redefine power distance as a subor-
dinate dimension of individualism and collectivism (building on Hofstede’s
recognition that the two strongly covary.) Horizontal (low power distance) in-
dividualism is linked to the freedom to define and pursue one’s goals, and
vertical (high power distance) individualism to competition and accomplish-
ment. Horizontal collectivism is tied to interdependence, in-groups, sharing,
and consensus, whereas vertical collectivism is associated with service and the
internalization of group norms.
Each category has its negative side. With horizontal individualism can
come loneliness; with vertical individualism, personal stress. Horizontal
collectivism can lead to distrust of people outside the group; vertical collec-
tivism, to the emergence of excessive bureaucracy and rigid observance of
rules. People are most satisfied with their lives when there is a good fit be-
tween their personal self-concepts and their social environment.

Triandis as a Guide for Interface Development

Triandis’ skepticism about the ability of Western social science to identify

universal principles of human behavior or universal methods of human re-
search is a critique HCI should consider. All science is embedded in cul-
ture—computer science and interface design no less than social psychology.
Triandis urged his own field to embrace collaboration and multi-site testing
as the solution. This methodology is one that HCI should also adopt.
In addition, Triandis strengthened the explanatory value of individual-
ism and collectivism. His explanation of the power of in-groups suggests
that interface designers must work harder to establish trust with new users
from collectivist cultures. His division of individualism and collectivism into
multiple levels (individual, organizational, and societal) explains how a
person born in Japan can be more independent than one born in the
United States. (A society may be predominantly individualistic or collectiv-
ist, but individuals within that society can vary.) His theory of social identity
acknowledges that a person (or culture) can be simultaneously individualis-
tic and collectivist.
These refinements make individualism and collectivism elements in a
personal repertoire and improves our understanding of them as design
heuristics. Moreover, it improves them as experimental variables. Cul-
ture-level dimensions are particularly problematic in usability testing with
the small number of subjects in the average usability lab. Many studies
based on Hofstede’s constructs have returned ambiguous results because
the researchers assumed that everyone in a culture shares the same cultural
98 GOULD

orientation. Individual-level variables provide better information about

their users and more realistically reflect the range of beliefs within cultures.
Hazel Rose Markus and Shinobu Kitayama (1991) incorporated
Triandis’ notions into their theory of independent (individualistic) and in-
terdependent (collectivist) self-construals. These personal attitudes are eas-
ily measured with Singelis’ (1994) Self-Construal Scale (since modified) by
Cross, Bacon, & Morris, 2000) and make a better research variable for HCI
than assignments based on Hofstede’s country rankings.

WILLIAM B. GUDYKUNST: BUILDING A UNIFIED

THEORY OF COMMUNICATION

William B. Gudykunst’s work in the field of communication parallels Harry

Triandis’ advocacy of a cross-cultural perspective for social psychology.
Like Triandis, Gudykunst first saw cultural differences as fundamental but
now works to reconcile them within higher level constructs. He too has ex-
plored the importance of social identity. However, unlike Triandis, much of
his work is based on the analysis of specific communication contexts. In-
stead of focusing on a single cultural syndrome, Gudykunst has investigated
the effect of culture on initial encounters and relationship building. Re-
cently, he has also looked at communication style.
Gudykunst’s practical experience as an Intercultural Relations Specialist
for the U.S. Navy in the early 1970s first helped shape his career. “While
conducting intercultural training in Japan, I thought intercultural commu-
nication (i.e., communication between people from different cultures) was
different from intracultural communication (i.e., communication with
members of our own culture)” (1994, p. ix).
Gudykunst’s first papers show this training bias. He examined the way
people from different cultures encounter and learn to trust one another by
applying Hall’s and Hofstede’s dimensions to classic communication pro-
cesses like homophily (similarity) and uncertainty reduction (1983a,
1983b). But by 1984, he and Young Yun Kim concluded that intercultural
and intracultural communication vary along a single continuum from total
strangeness to total familiarity.
Initial interactions are particularly critical when people from different
cultures meet because they are likely to have less in common. They begin by
disclosing information about themselves and asking questions. These strat-
egies help develop attraction by identifying common values, discovering
common relationships, reducing uncertainty, and increasing attributional
confidence. The process of interaction becomes an opportunity for coordi-
nation and affiliation.
At first, social stereotypes and information about group memberships
are particularly important. As the relationship develops, people make more
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 99

predictions based on knowledge of each other’s personal characteristics.

Eventually, the need for similarity decreases (old friends tolerate greater
differences). This pattern is consistent across cultures, but the information
requirements differ. People from high-context cultures ask more questions
about family and background; people from low-context cultures are more
immediately sociable, expressive, and direct.
Gudykunst and Tsukasa Nishida (1986) found a correlation between
Hofstede’s dimensions and levels of intimacy in six types of relationship.
Gudykunst, Elizabeth Chua, and Alisa J. Gray (1987) confirmed that people
look for greater evidence of similarity when beginning new intercultural
friendships. In 1988, Gudykunst pulled together his work on individual and
group representations and relationship development and published it as Anx-
iety/Uncertainty Management (AUM) theory (updated in 1993 and 1995.)
In addition to collapsing inter- and intracultural communication,
Gudykunst can also be credited with introducing new theories from Europe
into interpersonal communication. In 1986, he edited Intergroup Communi-
cation and wrote an introduction with Tae-Seop Lim:

This book is intended to bring together the study of intergroup relations in

Europe and the study of interpersonal communication in the United States.
Both areas have developed separately, with little borrowing of concepts or
theories. There are advantages to both areas, however, in their integration.
It is through communication that intergroup relations are established and
communication plays a central role in their improvement or deterioration.
The intergroup relations perspective brings a focus upon group member-
ship and its influence on the study of interpersonal communication that vir-
tually has been ignored to date in the United States. (1986, p. 1)

Individuals come together in groups for a variety of reasons (perceptions

of similarity, common interests, interdependence, or the desire for interac-
tion.) Members learn to identify with one another and this social identity
becomes part of their self-concept. People can belong to many groups and
develop multiple social identities. Personal identity makes up the remain-
der of their self-concept.
When social identity is dominant, intergroup communication takes
place. When personal identity is dominant, interpersonal communication
takes place. Situations like labor negotiations activate both types of identity;
anonymous encounters on the street activate neither. Gudykunst and
Tae-Seop Lim’s (1986) model (see Fig. 4.3) highlights the influence of com-
munication context on identity. The contact situation often determines
whether social or personal identity is made salient.
The final theme in Gudykunst’s work has been his research on communi-
cation style. He used Hall’s theory of high- and low-context communication
100 GOULD

FIG. 4.3. Interpersonal and intergroup communication.

in his work on initial interactions and relationship development, but he has

also looked for the predictors of style. In 1988, Gudykunst, Stella
Ting-Toomey, and Chua developed a model of interpersonal communica-
tion. This model shows how culture indirectly mediates communication
style and content through social cognitive processes, situational factors, af-
fect, and personal habits.
A few years later (1996), the model was simplified by incorporating
Markus and Kitayama’s self construals (see Fig. 4.4).
Gudykunst and Ting-Toomey collaborated with Yuko Matsumoto,
Nishida, Kwangsi Kim, and Sam Heyman to survey high- and low-context
styles in the United States, Japan, Korea, and Australia. When defined at
the level of culture, individualism and collectivism only partially predicted
communication style; however, when defined on an individual level (inde-
pendent vs. interdependent self construals plus personal values), the di-
mension fully predicted communication style. People who considered
themselves unique and independent used a low-context style, whereas peo-
ple who considered themselves part of larger social relationships used a
high-context style.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 101

FIG. 4.4. Model of communication styles.

Gudykunst as a Guide for Interface Development

Gudykunst’s research on initial encounters, identity, and communica-

tion style has direct relevance for HCI. In addition, it suggests some in-
teresting intersections between intercultural and computer-mediated
communication.
“Meeting” new software (or a new web site) invokes the same process of
prediction, attribution, and uncertainty reduction that people use when
meeting a stranger. Therefore, interface designers need to provide appro-
priate relationship-building information because people will make these
attributions even when the uncertainty reduction process is unacknowl-
edged. Group memberships are particularly important information for
people who use a high-context style; personal expressiveness is particularly
persuasive for people who prefer low-context communication.
Global access makes it likely that Web designers must include both types
of information. The Web also provides some interesting technical opportu-
nities for enhancing subsequent interactions. Cookies permit personaliza-
tion of a site visited before. Usability, task orientation, and efficiency can be
optimized to show consideration to low-context communicators.
Interactivity by itself will enhance liking, that is, as long as the experience is
positive. (See Rafaeli, 1988; Rafaeli & Sudweeks, 1998.) But new users may
take different amounts of time to develop a sense of trust. People from
high-context cultures are likely to need more information from the inter-
face before they commit themselves; on the plus side, they are probably will-
ing to explore more of the web site or application to find it.
102 GOULD

Gudykunst and Lim’s discussion of the salience of social and personal

identity also has some interesting implications for HCI. Different contexts
of use, for example, learning online presentation software at work, buying
printer paper from the cheapest supplier on the web, or picking a university
from a distance, may make different parts of one’s identity salient. People in
a work environment make the effort to learn a new application because it is
a necessary part of their corporate identity. People engaged in an online
transaction that involves little risk and no status or sense of participation
simply send their order to the lowest vendor. But people researching an or-
ganization that requires a long-term commitment (or carries high risk) may
need different information, depending on whether their preference is for
high- or low-context communication. For instance, an Asian technician de-
ciding to come to Europe for a training program could require more text on
the history and management of the institute than a North American. But
the same Asian would not need that level of information when picking a ski-
ing vacation. Considering the effects of an application on social versus per-
sonal identity should become a new design heuristic.
Several computer-mediated communication theories focus on the same
issues. Initial interactions and social identity come together in Martin Lea
and Russell Spears’ (1991) SIDE model (Social Identity model of
Deindividuation Effects). This model suggests that people working to-
gether across computer networks make attributions about one another de-
spite the lack of nonverbal cues present in face-to-face meetings. When
their mutual group identity is enhanced and made salient, they will collabo-
rate productively; when their individual identities are enhanced, they may
“act out” (deindividuate) and ignore the needs of others. Joseph Walther
and Judy Burgoon (1992) found that people in computer-mediated teams
achieve the same level of liking and support as people in face-to-face
groups, but it takes more time. Due to Walther’s hyperpersonal effect, peo-
ple may even like one another more because their impressions are based on
positive projections and stereotypes. In deciding what kind of information
to front-load on a home page, web designers need to consider whether peo-
ple will be visiting the site once or returning repeatedly and whether they
need to gather information to personalize subsequent interactions.
Finally, Gudykunst’s work on communication style with Matsumoto,
Ting-Toomey, Nishida, Kim, and Heyman (1996) has direct application to
interface development. Eight factors emerged when respondents from Aus-
tralia, Japan, Korea, and the United States were asked to rate 158 commu-
nication behaviors related to high- and low-context communication:

1. Certainty in inferring others’ meanings, intentions, needs, and

feelings (associated with individualism because it reflects a con-
scious strategy).
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 103

2. Use of indirect or ambiguous communication (collectivist).

3. Interpersonal sensitivity, tact, and preference for harmony (collec-
tivist).
4. Use of dramatic communication (individualist).
5. Openly expressed use of feelings to guide behavior (individualist).
6. Personal openness and willingness to initiate communication with
others (individualist).
7. Precision in communication (individualist).
8. Positive perceptions of silence in communication (collectivist).

Some of these styles reflect the way the respondents process informa-
tion; others, the way they expect information to be delivered. The eight
factors were correlated with independent and interdependent self
construals and with personal values derived from individualistic and col-
lectivist cultures. Because people can have both independent and inter-
dependent self construals, the authors note that specific situations may
activate opposite self-concepts. For instance, people might use a differ-
ent style communicating with an in-group than with an out-group they
had no interest in joining.
Although limited to four cultures, this study provides direction on the
most important features to emphasize in developing interface text and
graphics for people with individualist and collectivist values.

FONS TROMPENAARS: CULTURE HITS THE BEST-SELLER LIST

Fons Trompenaars’ best seller, Riding the Waves of Culture, was first pub-
lished in 1994; the second edition, with Charles Hampden-Turner as coau-
thor, came out 4 years later. The book made Trompenaars an international
management guru. (According to his web site, he was listed as a top man-
agement consultant in 1999, along with Michael Porter, Tom Peters, and
Edward de Bono.) His discussion of cultural differences as potential busi-
ness assets began as a critique of existing Western management theories.
Trompenaars’ career has been in industry—working for Shell’s Person-
nel Department in the 1980s and founding his own consulting firm (now
Trompenaars Hampden-Turner) in 1989. His practice specializes in
cross-cultural mergers, executive coaching, vision statements, and training.
Clients fill out surveys on their cultural orientations and cultural compe-
tence, and Trompenaars has accumulated a database of over 55,000 re-
sponses to support his analysis of culture.
For Trompenaars, management theory is not culture-neutral. Business
practices in different countries can look the same but be based on vastly dif-
ferent assumptions:
104 GOULD

The issue is not whether a hierarchy in the Netherlands has six levels, as does a
similar company in Singapore, but what the hierarchy and those levels mean
to the Dutch and Singaporeans. Where the meaning is totally different, for ex-
ample, a “chain of command” versus “a family,” then human-resources poli-
cies developed to implement the first will seriously miscommunicate in the
latter context. (Trompenaars & Hampden-Turner, 1998, p. 6.)

Culture defines the way each society solves the problems of its social,
temporal, and physical environment. Trompenaars uses an analytic frame-
work of seven cultural dimensions: the five pattern variables of Talcott Par-
sons (1951), which explain how people relate to one another within a social
system, and the cultural dimensions for time and the environment from
Kluckhohn and Strodtbeck (1961) and Edward T. Hall (1959/1973).
The seven dimensions are:

1. Universalism vs. Particularism: the extent to which people base their

actions on rules or relationships. In universalist societies, people believe
it is possible to define universal principles which will be “true” in all situa-
tions. Governments are based on the rule of law; economics is a social
“science”; and people conduct business as a rational enterprise. In
particularistic societies, people temper their actions to fit the social situa-
tion. They may support legal principles in theory but bend the truth in a
court of law to help a friend; the term social science is an oxymoron; and
business is a creative art. Universalists base their business dealings on
contracts; particularists renegotiate their business relationships when
personnel change. Abstract reasoning and statistics persuade universal-
ists; the practical effects of policies persuade particularists.
2. Individualism vs. Communitarianism: the extent to which people act
on the basis of their own needs or the needs of their social group. Individ-
ualists believe that society benefits when all have the freedom to act in
their own self-interest. Communitarians believe that society benefits
when people discipline themselves and consider the needs of others. In-
dividualists like to work (and play) alone; they expect credit for the per-
sonal responsibility they take on. Communitarians prefer to work closely
with others; they may vacation in a group; they expect credit to be given
to the project team as a whole (not just the team leaders.) Personal incen-
tives (like cash bonuses) motivate individualists; higher-order principles
(like quality) also motivate communitarians.
3. Neutral vs. Expressive Communication Styles: the extent to which
people hide or display their emotions. At a fundamental level, people
everywhere feel the same basic emotions (see Gudykunst and
Ting-Toomey, 1988), but they do not express them in the same way. Peo-
ple in neutral cultures restrain their feelings; people in affective cul-
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 105

tures openly show them. When someone in a neutral culture loses

control, their loss of composure is shocking; when someone in an affec-
tive culture feels something extreme, their depth of feeling may not be
noticeable. When someone in a neutral culture uses irony, the humor is
likely to be lost on people who have a passionate communication style;
in fact, the ironist may even be thought deceitful. Different cultures sep-
arate their emotions from objective, logical decision-making or display
their emotions as a legitimate part of the process. In face-to-face com-
munication, each style affects eye contact, rate of speech, tone of voice,
facial animation, use of silence, and overall preference for nonverbal or
verbal communication.
4. Specific vs. Diffuse Relationships: the extent to which relationships
with other people are limited or unconditional. People in specific societ-
ies may consider coworkers close friends but never see them out of the of-
fice; people in diffuse societies become close to fewer people at work but
include them in their private life. These different expectations of inti-
macy reflect culturally different notions of public and private space and
of the costs of friendship. Consequently, people in specific cultures often
feel free to criticize friends because the friendship doesn’t involve a deep
level of intimacy. People in diffuse cultures often seem indirect (even eva-
sive) because they know that their words have greater emotional signifi-
cance. They also take longer to make friends because the cost is
unconditional acceptance. Once that relationship has been established,
the relative social roles tend to stay the same forever—if you become
friendly with a teacher, that person can give you advice even after you es-
tablish yourself successfully. This dimension influences both the depth
and breadth of emotional involvement and is closely related to commu-
nication style. There is also a close link between specific and diffuse rela-
tionships and several of Hofstede’s dimensions.
5. Achievement vs. Ascription: the extent to which status is accorded on
the basis of personal performance or social role. Cultures with an
achievement orientation value performance and action; doing some-
thing is better than doing nothing. Cultures where status is ascribed on
the basis of gender, age, family connections, or profession assume that
these characteristics are important indications of effectiveness; doing
nothing—waiting for opportunities to present themselves—is better
than acting in haste. Achievement cultures value youth for its own sake
and are often informal; people expect to change jobs and employers
during the course of their career. Ascription cultures value experience
and maturity; people are loyal to their company and demonstrate re-
spect for the organization by absolute respect for superiors. Networks
in achievement cultures are personal; in ascription cultures, they dupli-
cate the official power structure. Titles and degrees are important in as-
106 GOULD

cription cultures since everyone needs to know the sources of an

individual’s status.
6. Orientation to Time: the time horizon (past, present, or future) and
preference for sequential or synchronic activities. Trompenaars com-
bines the time dimensions of both Kluckhohn and Strodtbeck and Ed-
ward T. Hall. People view their actions in historical terms, in the “here
and now,” or in terms of the future. As a result, they may be resistant to
change, follow current trends, or defer immediate gratification for
long-term advantage. An orientation to the past or distant future is of-
ten found in ascription cultures; an orientation to the present (or near
future) is found in achievement ones. Ascription cultures may see time
as circular; achievement cultures see it as linear. Different groups also
use time in different ways. Sequential cultures make schedules and do one
thing at a time (Hall’s monochronic cultures). Synchronic cultures are
open to the situation; they let events develop their own momentum and
often end up juggling multiple activities at the same time (Hall’s
polychronic cultures).
7. Orientation to Nature: the extent to which people feel they can con-
trol, or are controlled by, nature and the social environment. Like
Kluckhohn and Strodtbeck, Trompenaars believes that people see them-
selves as masters of nature or living in harmony with nature (and occa-
sionally mastered by it.) These contrasting positions lead to varying
appreciations of human agency and the natural world. People who are in-
ner-directed believe they can control external events and are attracted to
technology and stories of human achievements; people who are outer-di-
rected value nature for its own sake and prefer to work within the status
quo. When plans fail, inner-directed people may react aggressively and
try again, while outer-directed people bide their time and wait for new
opportunities. A control orientation can affect a culture’s willingness to
change and innovate while an orientation to harmony may promote flex-
ibility and service to others.

Because of the popularity of Riding the Waves of Culture, Trompenaars’

seven cultural differences are well-known. The framework is very appropri-
ate for intercultural studies but does have one problem—unlike Hofstede’s
dimensions, it has been applied to (not derived from) a statistical analysis of
Trompenaars’ survey data.
Trompenaars does not summarize his data by culture; in addition, coun-
tries are rank ordered for individual questions but not for each dimension
overall. As a result, the distinctions between many of the dimensions are un-
clear: for instance, one question on neutral or expressive communication
style seems similar to another on specific or diffuse relationships. Individu-
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 107

alism and collectivism, universalism and particularism, and achievement

and ascription all appear related. This makes it difficult to use
Trompenaars’ set of cultural values experimentally.

Trompenaars as a Guide for Interface Development

Nonetheless, Trompenaars’ seven cultural differences make excellent

design heuristics—especially when combined with principles of graphic
design like the “grammar” of Kress and van Leeuwen (1996). Should the
mode of address in messages, text, and graphics be especially polite and
reassuring or more direct (and potentially brutal)? How must trust do
people need to develop before they invest hours of work in an applica-
tion? Do they want to be in control or can a wizard help them achieve
their goals?
It is easy to speculate on specific effects, particularly for the Web. For in-
stance, sites for people with universalistic values might focus on products,
functionality, and price, whereas sites for people with particularistic values
might focus on business relationships and service.

Universalism Particularism
Overall focus on products, function- Overall focus on business relationships and
ality, and price service
Little information about company Strong branding
Price comparisons Customer testimonials
Formal rhetorical style Familiar rhetorical style (2nd person ad-
dress)
Sharp focus images that highlight Soft focus images that highlight the cus-
the product tomer
Symmetric templates Asymmetric templates
Consistent interaction style Use of different interaction models to
achieve different goals

In selling to a universalist market, a company like Honda might use the

Web to emphasize the innovative engineering used in its new gas-hybrid
cars while emphasizing its service and reputation in selling to a
particularist market.
Sites for people from individualistic cultures might be designed to en-
hance individual performance, whereas those for people from communi-
tarian cultures might emphasize group goals and behavior.
108 GOULD

Individualism Communitarianism
Focus on individual self-interest and Focus on collaborative behavior and
enhanced performance group communication
Text reinforces self-image, status, or Text discusses benefits to others
material success
Emphasis on innovation Emphasis on reliability and quality
Images of individuals using the prod- Images of groups of people using the
uct (or the product alone) product (product may never be dis-
played)
Images of people at play (or working Images of formally dressed people work-
in informal settings) ing with others

When appealing to individualists, a food company like Starbucks might

emphasize the rarity, taste, and price of its shade-grown coffee but focus on
the benefits to the environment when pitching to communitarians.
Sites targeted at people with a neutral communication style might use a
more subdued color palette, reserved rhetorical style, and larger propor-
tion of text to graphics than sites targeted at people with an affective style.

Neutral Communication Style Expressive Communication Style

Subdued color palette Bright, expressive colors
Reserved rhetorical style Exaggerated claims, elaborate words, and
emphatic punctuation
More text than graphics Equal or larger amount of graphics to text
Rhetorical strategy based on objec- Rhetoric anchored in personal stories or
tive, logical analysis narratives
Uncluttered layout Dense layout
Long- and middle-range images Close-up images
People in images make reserved eye Strong eye contact or none at all
contact
Use of irony and word play Use of comic images and animations

A cruise line selling in northern Europe (where most people use a neutral
communication style) might show an overall image of one of its boats and em-
phasize its ports of call. The same line might show the sumptuous buffet (or a
collage of its tourist facilities) when selling to a more expressive U.S. audience.
People from specific cultures may prefer web sites that are task-oriented
and direct, whereas diffuse cultures may require more information (much
of it relational) even for nominally simple web transactions.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 109

Specific Relationships Diffuse Relationships

Task-oriented information Information that references the busi-
ness relationship
Direct rhetorical style Indirect rhetorical style
Evocation of personal satisfaction Evocation of public principles
Images of happy, smiling people Images of reserved, serious people
Specific to general logic General to specific logic

A U.S. university recruiting students from a specific culture might talk

about the starting salaries of its graduates and show groups of happy students
(at play or graduation). Students from a diffuse culture would look for infor-
mation about the prestige of its programs and graduates and be persuaded to
come by images of students hard at work on the best new equipment.
People from achievement cultures are likely to be persuaded by sites that
help you “do something” (buy a product or recommend a book), whereas
people from ascription cultures may require information about who is do-
ing the selling before they will act. It is not that these users don’t want to use
the Web productively—they do, but first they need to know why they should
buy from one organization rather than another.

Status by Achievement Status by Ascription

Programmatic sites that “do some- E-commerce functions secondary to or-
thing” ganizational information
Focus on outcomes Focus on reputation and qualifications
Visual angle giving power to the viewer Visual angle enhancing size of object
Images of youth (both male and fe- Images of older men
male)
Leaders made to appear approachable; Leaders given full job titles and honor-
names used without titles (but qualifica- ifics
tions often given indirectly)
Frequent updates and revisions Site relatively static (consistent over
time)
Many links (broad action hierarchy) Fewer links (deep action hierarchy)
presenting many choices with choices more strongly structured

An established company like IBM would have a built-in advantage over

a relatively new company like Dell when selling to people from an ascrip-
tion culture. It could maximize this advantage by reminding people of its
technical leadership and focusing on the CEO. In an achievement culture,
110 GOULD

such information would be secondary to providing advanced functions

(like the ability to build your own PC through the web site). Too much rela-
tional information might even be viewed as “information overload” and
become detrimental.
Differences in time horizons can also influence the amount of text and
the choice of graphic design. Sites with a past orientation need to give a lot
of historical context and use traditional color schemes; sites with a present
orientation should focus on quick action and use bright, active colors; and
sites with a future orientation ought to talk about mission and future plans
and consider using a pale (indecisive) palette.

Orientation to Past Orientation to Present Orientation to Future

Dark colors with histori- Bright, active colors Futuristic or dreamy col-
cal connotations ors
Background information Current information only Information on mission
before new information and future plans
Emphasis on text and his- Emphasis on empowering Promotion of organiza-
torical information users to act quickly tional innovation and use
of novel interface features
Images of historical fig- Images of contemporary Nonrepresentational or
ures or older people people doing things idealized images

These color schemes and rhetorical strategies could also help a site tar-
get different domestic audiences, such as conservative investors or early
adopters of new technology.
Sequential and synchronic sites might use different interaction models.
For instance, a sequential site might use a careful hierarchy to lead people
through tasks one step at a time. (To buy a train ticket, a user would first be
required to choose a destination. Then the system would provide a list of
departure times. Finally, the system would ask for the number of people in
the party and let the user choose from available classes of travel.) A
synchronic site might ask all those questions on the same page, knowing
that some queries could fail and force the user to choose again.

Sequential Use of Time Synchronic Use of Time

Strongly structured task hierarchies “One-click” interaction model
Wizards Tolerance for failure
“Breadcrumb trails” and site maps Total reliance on search engines

Finally, people from cultures with a control (inner-directed) orientation

to nature are likely to celebrate human agency and technology, whereas
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 111

people from cultures with a harmony (outer-directed) orientation might fa-

vor graphics of nature and social cooperation.

Inner-Directed Cultures Outer-Directed Cultures

Focus on action and technology Focus on nature and human community
Spare, instructional text Figurative, metaphoric text
Images of human achievements (e.g., a Images of nature (small-scale images
picture of the Hoover dam) that evoke nature in microcosm or pan-
oramas that emphasize its power)
Bright, active colors Pale, “natural” colors
Symmetric templates Asymmetric templates

These variations are all speculative. Unlike Hofstede’s work dimensions,

Trompenaars’ cultural differences have not been extensively used in re-
search, but his work does fill gaps in the analytic frameworks of other cul-
tural values researchers. In addition, five of his seven dimensions apply
directly to communication styles and relational goals.

DAVID VICTOR: CAPTURING INTERCULTURAL

COMMUNICATION IN A MNEMONIC

One last theorist remains to be discussed. David Victor takes a more histori-
cal and categorical approach to intercultural communication, focusing on
specific elements of the communication context that can facilitate or hinder
mutual understanding.
In his 1992 text, he warned that international business communication is
increasingly subject to two contradictory trends: the homogenization of
consumer needs (and products) and the intensification of cultural hetero-
geneity. On the one hand, consumer demographics have converged so that
you can buy blue jeans in Japan, Pokemon in Africa, CDs by Ali Farka Toure
in the United Kingdom, and Christmas crackers in the United States. On
the other, people have begun to resist homogenization by reemphasizing
their cultural uniqueness. Strong political-social movements in Quebec,
Eastern Germany, and much of the Arab world seek to protect cultural iden-
tity. Even Japan, an enthusiastic consumer of Western popular culture, re-
cently instituted a Council on the Japanese Language to investigate the
proliferation of loan words in katakana.
Victor promises that anyone can learn to communicate in any culture by
using his “culture-general” approach (though they will need to work at it.) A
mnemonic—LESCANT—captures the seven most important aspects of in-
ternational business communication:
112 GOULD

1. Language
2. Environment and technology
3. Social organization
4. Contexting
5. Authority conception
6. Nonverbal communication
7. Temporal conception.

The pragmatics of communication tend to be more important to Victor

than cultural theories. However, unlike popular books on “Do’s and Ta-
boos” (specific things to do, or avoid, when dealing with people from partic-
ular countries), his analysis is detailed and comprehensive. For instance,
the chapter on language ranks the world’s leading languages, discusses the
Sapir–Whorf hypothesis, gives examples of linguistic ethnocentrism that
can isolate the non-speaker, identifies problems related to linguistic (but
not cultural) fluency, lists strategies for working with translators, and cau-
tions against stereotyping based on broken speech.
Unlike most of the theorists, Victor also comments on the role of tech-
nological factors in international communication. He describes how a
sales opportunity was lost when a business person expected a potential cli-
ent to have easy access to a phone. Software designers often make the same
error, expecting users to have access to the latest hardware and software
and the desire to work with the latest release. In many countries, people
deliberately choose to work with back-level systems to prevent technical
problems and reduce risk.
Victor’s discussion of “contexting” (communication that relies on infor-
mation outside the message) merges Hall’s theory with Hofstede’s notion of
uncertainty avoidance. Communication in high-context cultures is embed-
ded in strong interpersonal relationships: Strangers must develop trust be-
fore they can successfully negotiate with one another. By contrast, people in
low-context cultures focus on the end product, not the provider. Perhaps as a
result, they depend to a greater degree on explicit verbal and written com-
munication, contracts, and commercial law. In addition, people in high-con-
text cultures rely more on nonverbal communication and become skilled in
inference. Silence and the settings in which people speak often communicate
more meaning than the polite, indirect speech they use. However, this am-
biguous style does not create any tolerance for ambiguity or risk.
People in high-context cultures avoid uncertainty by acting conserva-
tively and gathering as much information about their communication
partners as possible. This relates back to the emphasis on personal rela-
tionships, but it comes with a downside: “Face” becomes central to the
communication process. One is forced to protect one’s own public per-
sona and the dignity of everyone else. Consequently, people in high-con-
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 113

text cultures prefer indirection, harmony, and politeness to argumen-

tation, confrontation, and self-disclosure. Everything works well until
low-context people attempt to communicate with them. Then, the differ-
ences in communication style may end in shame and mutual loss of face—
although the low-context partner often does not know it. (Stella Ting-
Toomey [1996]; Ting-Toomey, Oetzel, & Yee-Jung [2001] discussed face
negotiation in more detail.)
Victor’s wealth of information can be overwhelming. Like the “thick de-
scription” of anthropology, it is hard to use the LESCANT system to predict
areas of concern because you have to be immersed in the other culture to do
the analysis. But if you plan to target just one other culture, Victor’s frame-
work is an excellent way to identify areas of agreement and cultural faultlines.

Victor as a Guide for Interface Development

Victor’s focus on the pragmatics of business communication has many

points of intersection with interface design. For instance, localization often
treats the issue of language as a simple technical problem, that is, leave
room for expansion and translate the text. But Victor (1992) notes that ap-
parent fluency has its own drawbacks:

For example, a U.S. businesswoman may represent a company in negotiat-

ing with a Japanese firm. The Japanese negotiating team with whom she in-
teracts has little knowledge of English. Recognizing that the woman speaks
Japanese, the negotiating team is likely to judge her by Japanese standards.
When the U.S. businesswoman breaks Japanese gender-linked social expec-
tations and does not behave as they might expect a Japanese woman to be-
have, she is less likely to be forgiven than a foreigner unable to speak
Japanese might be. (p. 30).

Localized products often evoke similar responses. Users appreciate be-

ing communicated within their own language, but untranslated cultural as-
sumptions undermine their good impression. Problematic interface
elements include informal text based on different norms of politeness and
icons or graphics that violate national and religious conventions. (Paradox-
ically, interfaces in the original language reduce user expectations and may
increase user tolerance for social gaffes.)
Interfaces need to be “culturalized” as well as translated; icons, graphics,
rhetoric, and interaction style must also suit the target audience. A German
telecommunications company found that investing in product differentia-
tion generated higher acceptance. Pia Honold (1999) ran a series of focus
groups and tests on user documentation for cellular phone customers in
114 GOULD

Germany and China. German users preferred a thick book identifying all
the phone functions; they wanted explicit information in as much detail as
possible. Chinese users asked for online documentation; they would lose
face if other people saw them looking at a book to solve a problem.
Corporate guidelines for localization should include a LESCANT check-
list. The interplay between cultural values and effective communication in-
volves a wide range of pragmatic issues that are easy to overlook.

SUMMARY

So, after discussing all these theorists, what are the most important cultural
values to consider when designing software?
As always, “it depends!” If I were adopting just one theorist’s model of
culture and communication, I would use that of Geert Hofstede or Shalom
Schwartz. Hofstede’s (1997) work dimensions have a great deal of face va-
lidity and are widely used in research, but they are not entirely independent
of one another. In addition, his country rankings omit some important na-
tions. By contrast, Schwartz’s (1994) cultural dimensions provide an alter-
native to Hofstede’s work and also include rankings. But his seven values
are more difficult to grasp and some of the countries included in his list are
atypical (Estonia, Bulgaria, etc.) .
If I were picking a set of values from those that show up over and over
again in various theories and studies, I would start with these elements of
the social environment: (a) individualism and collectivism, (b) power dis-
tance, (c) human action (mastery and harmony), and (d) time (particularly
the time horizons).
The primacy of the individual or the group seems to be the key consider-
ation for intercultural relationships. Kluckhohn and Strodtbeck, Hall, Con-
don and Yousef, Hofstede, Triandis, Gudykunst, Trompenaars, and Victor
all explicitly reference individualism and collectivism. Schwartz claims to
go “beyond individualism/collectivism,” but his cultural dimensions reflect
the same social dichotomy: The two autonomy dimensions and self-en-
hancement (individualism) are the obverse of conservatism and self-tran-
scendence (collectivism).
Conceptions of social power (power distance, hierarchy, ascription, and
achievement) strongly covary with individualism and collectivism. They
should be considered in conjunction with it or folded into the dimension as
vertical and horizontal individualism and collectivism (Triandis, 1995).
Understanding how humans interact with computers, and with one another
across cultures through applications or web sites, also requires an under-
standing of the ordering of social relationships.
Human action supports a set of themes associated with mastery and
harmony. Cultures that support mastery focus on personal empowerment
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 115

and activity for its own sake, whereas those that support harmony tend to
be more passive and perhaps more fatalistic. There is some overlap be-
tween mastery, masculinity, and individualism and harmony, femininity,
and collectivism, but country rankings do not always bear this out. For in-
stance, Schwartz (1994) ranks France at the bottom of his country list for
mastery but Trompenaars and Hampden-Turner (1998) put it near the
top of their list for control of one’s fate. Nonetheless, the social role ac-
corded technology in different cultures is an important aspect of the hu-
man–computer interface.
The fourth dimension is time, though its expression is multifaceted. Cul-
tural views of time show up in decision making predicated on short-term
rather than long-term goals, rhetorical strategies that ignore or appeal to
history, specific or diffuse relationships, and monochronic or polychronic
scheduling of events. However, in general, cultures that focus on the pres-
ent tend to ignore long-term consequences (whether past or future) and
work with other people through more instrumental (specific) relationships.
To round out my selection, I would also pick the two cultural dimensions
that most clearly deal with communication style: (a) context and (b) affec-
tive (expressive) and neutral communication.
Both overlap aspects of the first four values I mentioned, but more im-
portantly they supplement one another. Considering how people in differ-
ent cultures show or restrain their emotions enriches Hall’s theory of high-
and low-context communication. Hall linked individualistic cultures to ex-
plicit messages and collectivist cultures to nonverbal messages context-
ualized within a social setting, but both types of messages can show varying
amounts of affect. Most low-context countries, like the Netherlands, use a
neutral, unemotional style, but so do some high-context countries, like Ja-
pan. Arab culture tends to be expressive and high-context. Affective and
neutral communication should be considered in regard to communication
context, just as power distance and hierarchy have been folded into individ-
ualism and collectivism.

FUTURE DIRECTIONS

Despite the development of virtual teams and multinational business prac-

tices, culture is recognized as a far more potent force outside the United
States than in it. The expansion of the Common Market, the success of Jap-
anese management (before the collapse of the “bubble economy”), and the
globalization of the Chinese and Southeast Asian economies have
prompted the rest of the world to reexamine the validity of Western social
science and management theories. International journals of intercultural
communication, cross-cultural psychology, and management and technol-
ogy contain work with immediate application to HCI.
116 GOULD

Many researchers in China, Japan, Korea, and Southeast Asia are partic-
ularly active in identifying alternative cultural paradigms. Michael Bond (in
Hong Kong) helped Geert Hofstede identify Chinese values; Tsukasa
Nishida, Young Yun Kim, Yuko Matsumoto, Kwangsu Kim, and Taae-Seop
Lim have been exploring the validity of other generalizations about com-
munication. Cultural values, with their built-in dichotomies, work well
within the European and American academic tradition, but a great deal of
innovative research is being done throughout the world to develop more
holistic concepts of culture.
Western software development theories are also being subjected to anal-
ysis. People like Alvin Yeo have investigated the assumptions underlying
design and evaluation practices. Researchers dealing with the transfer of
collaborative technologies are finding that basic principles of usability do
not have the same utility in all cultures.
In Malaysia, Yeo (2001) discovered that usability tests and other stan-
dard methodologies associated with Western software engineering fre-
quently don’t work. Malaysian users testing a spreadsheet he developed
found it difficult to express their dissatisfaction; because of their collectivist
orientation, they censored themselves to preserve his face. Yeo received the
most accurate feedback from people already known to him (who knew he
needed their negative comments) and from computer experts (whose ex-
pertise authorized them to speak critically). He suggests that interface test-
ing in Asia must be based on objective performance measures rather than
on subjective user feedback. The International Workshops on Internation-
alisation of Products and Systems (IWIPS) showcase similar war stories
about interface development from people working around the world.
Western bias may even underlie some of our most cherished HCI beliefs.
Individualistic forms of decision making and task orientation are not uni-
versal. In the late 1980s and early 1990s, group decision support systems
(GDSS) were introduced into a number of countries in Asia and Africa.
Richard T. Watson, Teck Hua Ho, and K. S. Raman (1994) found that
Singaporean groups began with more consensus and showed less change
than U.S. groups. They speculated that the need for group harmony under-
mined individual problem solving. Although the effect was weak—coming
after group size, member proximity, and task type—they considered cul-
ture a necessary factor in interpreting GDSS outcomes.
Four years later, Gert-Jan De Vreede, Noel Jones, and Rabson J. Mgaya
(1998) reported on the introduction of GDSSs in Malawi, Zimbabwe, and
Tanzania. They found that incentives for participation were positively as-
sociated with endorsement by top management and negatively affected by
anonymous input. In these vertically collectivist states, people depended
on leaders to make initial judgments. In addition, participants wanted the
system to identify the contributions of individuals so that they could de-
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 117

cide whose ideas could be developed. Paradoxically, then, successful tech-

nology transfer depended on hierarchical influence and claims to “own”
ideas, two usability features that the software had been designed in the
United States to circumvent. (See Jay F. Nunamaker et al., 1991, for a list
of these design goals.)

CONCLUSION

Designing effective interfaces is not just good business. Software for inter-
cultural audiences can bring people together or drive them apart. Inappro-
priate design that violates cultural expectations perpetuates the digital
divide (Shneiderman, 2000, p. 86); software that offends local norms fuels
opposition to globalization. The Web is an emerging focal point for con-
flict. K. Viswanath and Liren Benjamin Zeng (2002) recently documented
the role of transnational advertising in anti-globalization protests. Simi-
larly, European privacy advocates have attacked Microsoft for its strategies
on personal data.
Interface designers need to think about ways to reduce alienation and
cultural bias in software that, by design, cuts across national and cultural
boundaries. Incorporating cultural values into our design practices is a
good way to start. We also need to test software in a variety of countries and
factor culturalization (as well as translation) into product development.
Perhaps in the future, software design will be reformed by a new paradigm
from Asia, Latin America, or Africa that transcends these problems. Until
then, we need to be mindful of the effects of culture and do our best to pro-
vide people everywhere with interfaces that support them.

REFERENCES

Barber, W., & Badre, A. (1998). Culturability: The merging of culture and usability.
In Proceedings of the Fourth Conference on Human Factors and the Web. Retrieved Jan-
uary 24, 2004, from www.research.microsoft.com/users/marycz/hfweb98/bar-
ber/index.htm
Benedict, R. (1989). The chrysanthemum and the sword: Patterns of Japanese culture.
Boston: Houghton Mifflin. (Original work published 1946)
Condon, J. C., & Yousef, F. S. (1975). An introduction to intercultural communication.
New York: Macmillan.
Cross, S. E., Bacon, P. L., & Morris, M. L. (2000). The relational-interdependent
self-construal and relationships. Journal of Personality and Social Psychology, 78,
791–808.
de Mooij, M. (2001). Internet and culture. Proceedings of the Conference on Internet ,
Economic Growth, and Globalization (August 8, 2001). Duisburg, Germany: Insti-
tute for International and Regional Economic Relations, Gerhard-Mercator Uni-
versity.
118 GOULD

De Vreede, G-J., Jones, N., & Mgaya, R. J. (1998). Exploring the application and ac-
ceptance of group support systems in Africa. Journal of Management Information
Systems 15, 197–234.
del Galdo, E. M., & Nielsen, J. (1996). International user interfaces. New York: Wiley.
Fernandes, T. (1995). Global interface design: A guide to designing international user inter-
faces. Boston: AP Professional.
Gudykunst, W. B. (1983a). Similarities and differences in perceptions of initial
intracultural and intercultural encounters: An exploratory investigation. The
Southern Speech Communication Journal, 49, 58–65.
Gudykunst, W. B. (1983b). Uncertainty reduction and predictability of behavior in
low- and high-context cultures: An exploratory study. Communication Quarterly,
31, 49–55.
Gudykunst, W. B. (Ed.). (1986). Intergroup communication. London: Edward Arnold.
Gudykunst, W. B. (1987). Cross-cultural comparison. In C. R. Berger & S. H. Chaffee
(Eds.), Handbook of Communication Science (pp. 847–889). Beverly Hills, CA: Sage.
Gudykunst, W. B. (1988). Uncertainty and anxiety. In Y. Y. Kim & W. B. Gudykunst
(Eds.), Theories in Intercultural Communication (pp.123–156). Newbury Park, CA:
Sage.
Gudykunst, W. B. (1993). Toward a theory of effective interpersonal and intergroup
communication: An anxiety/ uncertainty management (AUM) perspective. In R.
L. Wiseman & J. Koester (Eds.), Intercultural Communication Competence (pp.
16–32). Newbury Park, CA: Sage.
Gudykunst, W. B. (1994). Bridging differences: Effective intergroup communication
(2nd ed.). Thousand Oaks, CA: Sage.
Gudykunst, W. B. (1995). Anxiety/ uncertainty management (AUM) theory: Current
status. In R. L. Wiseman (Ed.), Intercultural Communication Theory (pp. 9–58).
Thousand Oaks, CA: Sage.
Gudykunst, W. B., Chua, E., & Gray, A. J. (1987). Cultural dissimilarities and uncer-
tainty reduction processes. Communication Yearbook, 10, 456–469.
Gudykunst, W. B., & Kim, Y. Y. (1984). Communicating with strangers: An approach to
intercultural communication. New York: Random House.
Gudykunst, W. B., & Lim, T-S. (1986). A perspective for the study of intergroup com-
munication. In W. B. Gudykunst (Ed.), Intergroup communication (pp. 1–9). Lon-
don: Edward Arnold.
Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., & Heyman,
S. (1996). The influence of cultural individualism-collectivism, self construals,
and individual values on communication styles across cultures. Human Communi-
cation Research, 22, 510–543.
Gudykunst, W. B., & Nishida, T. (1986). The influence of cultural variability on per-
ceptions of communication behavior associated with relationship terms. Human
Communication Research, 13, 147–166.
Gudykunst, W. B., & Ting-Toomey, S. (1988). Culture and affective communication.
American Behavioral Scientist, 31, 384–400.
Gudykunst, W. B., & Ting-Toomey, S., with Chua, E. (1988). Culture and interpersonal
communication. Newbury Park, CA: Sage.
Hall, E. T. (1973). The silent language. Garden City, NY: Anchor Books. (Original
work published 1959)
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 119

Hall, E. T. (1976). Beyond culture. Garden City, NY: Anchor Press.

Hofstede, G. (1980). Culture’s consequences: International differences in work-related val-
ues. Beverly Hills, CA: Sage.
Hofstede, G. (1997). Cultures and organizations: Software of the mind (Rev. ed.). New
York: McGraw-Hill.
Hofstede, G., & Bond, M. H. (1988). The Confucius connection: From cultural roots
to economic growth. Organizational Dynamics, 16, 417–433.
Hoft, N. (1996). Developing a cultural model. In E. M. del Galdo & J. Nielsen (Eds.),
International user interfaces. New York: Wiley.
Honold, P. (1999). Learning how to use a cellular phone: Comparison between Ger-
man and Chinese users. Technical Communication, 46, 196–205.
Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction.
Psychiatry, 19, 215–229.
Kluckhohn, C. (1951). Values and value orientations in the theory of action: An ex-
ploration in definition and classification. In T. Parsons & E. Shils (Eds.), Toward a
general theory of action (pp. 390–406). Cambridge, MA: Harvard University Press.
Kluckhohn, F. R. (1950). Dominant and substitute profiles of cultural orientations:
Their significance for the analysis of social stratification. Social Forces, 28,
376–394.
Kluckhohn, F. R., & Strodtbeck, F. L. (1961). Variations in value orientations. Westport,
CT: Greenwood Press.
Kostelnick, C. (1995). Cultural adaptation and information design: Two contrasting
views. IEEE Transactions on Professional Communication, 38, 182–196.
Kress, G. R., & van Leeuwen, T. (1996). Reading images: The grammar of visual design.
London: Routledge.
Lea, M., & Spears, R. (1991). Computer-mediated communication, de-individua-
tion and group decision making. International Journal of Man-Machine Studies, 39,
283–310.
Lederer, W. J., & Burdick, E. (1958). The ugly American. Greenwich, CT: Fawcett.
Lee, O. (2000). The role of cultural protocol in media choice in a Confucian virtual
workplace. IEEE Transactions on Professional Communication, 13, 196–200.
Marcus, A. (1993). Human communication issues in advanced UIs. Communications
of the ACM, 36, 101–109.
Marcus, A. (2001). International and intercultural user interfaces. In C. Stephanidis
(Ed.), User interfaces for all: Concepts, methods, and tools (pp. 47–63). Mahwah, NJ:
Lawrence Erlbaum Associates.
Marcus, A., & Gould, E. W. (2000). Crosscurrents: Cultural dimensions and global
web user-interface design. Interactions, 7, 32–46.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cogni-
tion, emotion, and motivation. Psychological Review, 98, 224–253.
Nisbett, R. E. (2003). The geography of thought: How Asians and Westerners think differ-
ently … and why. New York: Free Press.
Nunamaker, J. F., Dennis, A. R., Valacich, J. S., Vogel, D., & George, J. F. (1991). Elec-
tronic meeting systems. Communications of the ACM, 34, 40–61.
Ong, W. J. (1975). The writer’s audience is always a fiction. Publications of the Modern
Language Association of America, 90, 9–21.
Parsons, T. (1951). The Social System. Glencoe, IL: Free Press.
120 GOULD

Pike, K. L. (1954). Language in relation to a unified theory of the structure of human behav-
ior. Glendale, CA: Summer Institute of Linguistics.
Rafaeli, S. (1988). Interactivity: From new media to communication. In R. P.
Hawkins, J. M. Wiemann, & S. Pingree (Eds.), Advancing communication science:
Merging mass and interpersonal processes (pp. 110–134). Newbury Park, CA: Sage.
Rafaeli, S., & Sudweeks, F. (1998). Interactivity on the nets. In F. Sudweeks, M.
McLaughlin, & S. Rafaeli (Eds.), Network and netplay: Virtual groups on the internet
(pp. 173–189). Menlo Park, CA: AAAI Press.
Reeves, B., & Nass, C. (1996). The media equation: How people treat computers, television,
and new media like real people and places. New York: Cambridge University Press.
Said, E. W. (1994). Orientalism. New York: Viking Books. (Original work published
1978)
Shneiderman, B. (2000). Universal usability. Communications of the ACM, 43, 84–91.
Schwartz, S. H. (1994). Beyond individualism/collectivism: New cultural dimen-
sions of values. In U. Kim, H. C. Triandis, C. Kagitcibasi, S-C. Choi, & G. Yoon
(Eds.), Individualism and collectivism: Theory, method, and applications (pp. 85–119).
Thousand Oaks, CA: Sage.
Sheppard, C., & Scholtz, J. (1999). The effects of cultural markers on web site use. In
Proceedings of the 5th Conference on Human Factors and the Web. Retrieved January
23, 2004, from http://zing.ncsl.nist.gov/ hfweb/proceedings/sheppard/in-
dex.html
Singelis, T. M. (1994). The measurement of independent and interdependent
self-construals. Personality and Social Psychology Bulletin, 20, 580–591.
Stewart, E. C., & Bennett, M. J. (1991). American cultural patterns: A cross-cultural per-
spective (Rev. ed.). Yarmouth, ME: Intercultural Press.
Stephenson, G. M. (1981). Intergroup bargaining and negotiation. In J. C. Turner &
H. Giles (Eds.), Intergroup behavior (pp. 168–198). Chicago: University of Chicago
Press.
Ting-Toomey, S. (1996). Intercultural conflict styles: A face-negotiation theory. In Y.
Y. Kim & W. B. Gudykunst (Eds.), Theories in intercultural communication (pp.
213–235). Newbury Park, CA: Sage.
Ting-Toomey, S., Oetzel, J. G., & Yee-Jung, K. (2001). Self-construal types and con-
flict management styles. Communication Reports, 14, 87–104.
Triandis, H. C. (1975). Social psychology and cultural analysis. Journal for the Theory
of Social Behavior, 5, 81–106.
Triandis, H. C. (1978). Some universals of social behavior. Personality and Social Psy-
chology Bulletin, 4, 1–16.
Triandis, H. C. (1980). (Ed.). Handbook of cross-cultural psychology. Boston: Allyn & Ba-
con.
Triandis, H. C. (1983). Dimensions of cultural variation as parameters of organiza-
tional theories. International Studies of Management and Organization, 12, 139–169.
Triandis, H. C. (1988). Collectivism vs. individualism: A reconceptualization of a ba-
sic concept in cross-cultural psychology. In G. K. Verma & C. Bagley (Eds.),
Cross-cultural studies of personality, attitudes, and cognition (pp. 60–95). London:
Macmillan.
Triandis, H. C. (1989). The self and social behavior in differing cultural contexts.
Psychological Review 96, 506–520.
4. SYNTHESIZING THE CULTURAL VALUES LITERATURE 121

Triandis, H. C. (1995). Individualism and collectivism. Boulder, CO: Westview Press.

Trompenaars, F., & Hampden-Turner, C. (1998). Riding the waves of culture: Under-
standing diversity in global business (2nd ed.). New York: McGraw-Hill.
Victor, D. A. (1992). International business communication. New York: HarperCollins.
Viswanath, K., & Zeng, L. B. (2002). Transnational Advertising. In W. B. Gudykunst
& B. Mody (Eds.), Handbook of International and Intercultural Communication (2nd
ed.). Thousand Oaks, CA: Sage.
Walther, J. B. (1996). Computer-mediated communication: Impersonal, interper-
sonal, and hyperpersonal interaction. Communication Research, 23, 3–43.
Walther, J. B., & Burgoon, J. K. (1992). Relational communication in computer-me-
diated interaction. Human Communication Research, 19, 50–88.
Watson, R. T., Ho, T. H., & Raman, K. S. (1994). Culture: A fourth dimension of
group support systems. Communications of the ACM, 37, 44–55.
Williams, R. (1994). The non-designer’s design book: Design and typographic principles for
the visual novice. Berkeley, CA: Peachpit Press.
Yeo, A. W. (2001). Global-software development lifecycle: An exploratory study. In
Proceedings of the SIGCHI Conference on Human Factors in Computing (pp.104–111).
New York: ACM Press.
 Chapter #

Managing Multicultural Content

in the Global Enterprise

Jorden Woods
Co-Founder, Global Sight Corporation, San Jose, CA

INTRODUCTION AND THE IDEAL ENTERPRISE

This chapter is primarily concerned with the application of multicultural

content best practices to a global enterprise. Specifically, the chapter deals
with how to apply multicultural best practices not only to a single product or
deliverable, but to all consumable content generated and maintained in an
international organization on an ongoing basis.
To better understand what the challenges are and how the application of
best practices affects processes within a global corporation, I first look at a
vision of a fictitious ideal enterprise in the context of multicultural content.
To clarify, this ideal enterprise is that organization which is capable of pro-
viding every significant customer with the same high quality user experi-
ence regardless of country of origin or ethnic background. The easiest way
to visualize this ideal enterprise is to view it through the eyes of its custom-
ers, partners, and employees on a global basis.

Customer Experience

Every customer experience is clearly a very personal situation that cannot

be uniquely determined or parameterized. However, it is possible to gener-
alize a customer’s experience given enough interaction in combination with

123
124 WOODS

customer profiling and user feedback. In this generalized sense, the ideal
experience can be conceived as one in which the customer is provided with
or has easy access to useful information and/or functionality that meets the
customer’s needs, is aligned with the user’s preferences, and is provided in a
timely and relevant manner.
In this ideal setting, it should be clear that be I a Zulu farmer, a princess
in the Thai royal family, or a French nuclear technician I should value my
interactions with the company’s representatives and its content. In short
each interaction will meet my particular needs, be they:

• Linguistic, expected or preferred language or dialect.

• Cultural, expected or preferred cultural nuance in terms of values,
historical allusion, use of colors and icons, relationships, hierarchy,
religion, worldview, and so on.
• Technical, expected or appropriate weights, measures, protocols
and conventions.
• Economic, expected or preferred currency, taxation, shipping
methods, and so on.
• Legal, expected or appropriate statutory or regulatory conven-
tions.
• Geographical, expected or appropriate climactic or seasonal
norms.

Unfortunately, by not fully understanding their customer’s specific

needs, global enterprises often unwittingly provide many of their global
customers with inappropriate information or functionality in a culturally
threatening or linguistically intimidating style.
To further define this ideal experience, consider the following scenario:
a U.S. company has launched an existing North American wireless product
for the Asian marketplace. Primary markets for launch include China, Ja-
pan, Korea, Taiwan, Malaysia, Thailand, Indonesia, and Australia. The list
below details the key issues that would need to be taken into account to min-
imize customer alienation and generate keen user interest:

• Localized product interface.

• Culturally appropriate product size, shape, and color.
• Culturally and linguistically appropriate name and branding.
• Competitive distribution channels and partnerships for product
placement, accessibility, and promotion.
• Timely localized information about the product in multiple for-
mats: Web, print, wireless, and so on.
• Country or region specific services which appeal to the primary
customer demographic.
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 125

• Easily accessible, in-language customer service.

• Adoption of local protocols and standards to maximize
interoperability with locally relevant systems, accessories, and in-
frastructure.

By ensuring that the product meets these requirements in each of its tar-
get markets, the customer will feel that they are purchasing a product that
appears to have been developed specifically to address his or her needs.
Based on the foregoing analysis, it should now be clear that simply re-
packaging the device with translations of the original content will be insuffi-
cient to generate an ideal customer experience.

Employee or Internal Process Experience

There are many internal process scenarios that can lead to the ideal cus-
tomer experience outlined in the previous section. For example, employ-
ees may be working under tremendous stress in a situation where no
processes exist. Such a situation, by the very nature of its inefficiency, is un-
likely to generate a return on investment. Instead, I now focus on an ideal
environment within a hypothetical global enterprise.
To facilitate the corporate employees in meeting the diverse needs of a
global customer base, the ideal enterprise makes information, processes,
and resources available for the intelligent generation and update of multi-
cultural information. So, when communicating and interacting with a Zulu
farmer, a Thai princess, or a French engineer, appropriate information is
available for each customer’s particular needs, and efficient processes and
resources are available to deliver it quickly and accurately.
More specifically, I examine the particular case of the U.S. company that
is intent on launching an existing North American mobile device into the
Asian marketplace (China, Japan, Korea, Taiwan, Malaysia, Thailand, In-
donesia, and Australia). In this case, the ideal company’s marketing and en-
gineering staff would have easy access to the appropriate linguistic,
cultural, economic, and technical information to effectively market and
modify the product for each target country prior to product launch.
The ideal company’s business and information technology (IT) re-
sources ensure the quality and timeliness of the information and functional-
ity delivered. They would do this by developing and providing a process
infrastructure for automating and streamlining the tasks dictated by the
company’s business needs. Furthermore, resources, such as experts in the
areas of cultural, legal, and tax accountancy, would be available and would
have easy access to the company’s process flow.
Therefore, when simultaneously launching a product into markets as
different as Indonesia, China, and Australia, the company’s employees are
126 WOODS

able to rely on a breadth of information, process infrastructure, and exper-

tise to rapidly and routinely carry out their tasks. Deadlines are met, ex-
penses meet estimates, the uptake of the product in the target markets is in
line with expectations, and the company’s brand is enhanced.
This scenario should be contrasted with one in which, information, re-
sources, and processes are not in place. In such a scenario, deadlines are
only rarely met, expenses balloon out of control, uptake of product is slow,
and the company’s brand is eroded or tarnished. Examples of the latter
abound in the literature with cases such as the following:

• Nike’s initial humbling losses to Adidas in the European market-

place based on a misreading of their “rebel” branding as “hooli-
ganism.”
• The failure of the Big Three U.S. automakers to significantly infil-
trate Japan’s Japanese language-centric, metric-standard, right-
side drive road system with English-language, English-system,
left-side drive cars.
• The United States’ disastrous campaign of feeding the hungry in
Afghanistan with food packs the same relative size and color as clus-
ter bombs and with information distributed to a nearly illiterate
population without pictures and only in English.
• Walmart’s slow and expensive expansion into Latin America in
which the U.S.-centric culture did not translate and stores were
stocked with electronics based on 110 voltage which would not op-
erate locally.
• The U.S. Milk Council’s unforgettable “Got Milk?” campaign,
which, for the U.S. Hispanic market, was unsuccessfully conveyed
as “Are you Lactating?”

Success, however, can be spectacular as seen in:

• The ability of the Japanese Keiritsu to make rapid progress into the
consumer electronics, automotive, and steel building sectors on a
global basis.
• SAP’s domination of the Enterprise Resource Planning (ERP)
space on a global basis from its home base in Germany.
• Nokia’s rise from the small country of Finland to dominate the
global cell phone marketplace.

With this ideal company as a backdrop, it is now useful to explore the

challenges to success in greater detail.
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 127

CHALLENGES TO SUCCESSFUL GLOBALIZATION

As the examples in the previous section have clarified, there are multiple
challenges to the successful development and management of multicultural
information and functionality for global markets. These challenges, how-
ever, once analyzed can be seen to fall into three primary categories: pro-
cess challenges, integration challenges, and technical challenges.
The following sections explain each of these in more detail.

Process Challenges

Historically, companies have globalized their operations organically, with

new countries being added on an opportunistic basis through distributor
relationships, through acquisitions, and, in some cases, through the devel-
opment of formal subsidiaries. This approach, by focusing on each country
as a separate entity, has typically created organizations that are highly de-
centralized and only loosely integrated.
As a result, it is not uncommon for each office within an extended enter-
prise to have different business goals, employ different business processes,
and use different technology infrastructure to meet the needs of the local
marketplace. Furthermore, it is generally the case that divisions within a
single country operate autonomously or with only small consideration for
the processes, goals, and technologies employed by other divisions (con-
sider marketing, sales, IT, and engineering, e.g.).
From this analysis one can quickly see that the average global company is
not only loosely integrated internationally but is also loosely integrated lo-
cally. Due to the Web, there has been more focus recently on cross-func-
tional teams; however, it is more often the case that divisions operate
independently or in serial, rather than parallel, fashion.
Today’s products and services change continuously, operate under very
short deadlines, and demand tight communication and tight integration
among all the members involved in the development, promotion, sales, and
delivery processes. Under the operating model of a highly decentralized
and only loosely integrated global organization, differences between highly
incompatible business processes and technology infrastructure are exacer-
bated by differences in local language, culture, tax law, legal approach, time
zones, and governmental regulations.
In addition, business process within individual countries will vary based
on the type or purpose of the content that is being localized. For example,
the globalization business processes associated with a press release in a sin-
gle country may be radically different from the globalization business pro-
128 WOODS

cesses associated with a product catalogue or interactive customer support

within that same country.
These innate incompatibilities and differences create serious inefficien-
cies in process and work flow. Leveraging content from one region or coun-
try to another can then only be overcome by organizations expending large
amounts of both human and capital resources.
The sum of three variables—rate of content change, volume of content,
and number of languages or countries—can quickly reach critical values
which overwhelm the process and resource infrastructure of the corpora-
tion. In short, it can become prohibitively expensive and operationally im-
possible for a company to maintain its global content so as to meet the high
standards of quality and timeliness expected by customers, partners, and
employees in the global marketplace.
The greatest casualty for a company succumbing to process bottlenecks
is that the company becomes unable to meet its time to market commit-
ments. Once this critical capability is compromised, the company can suffer
in the following ways:

• Brand and message fragmentation.

• Distribution of inaccurate, out-of-date, or contradictory information.
• Decreases in customer loyalty due to breach of trust.
• Noticeable functionality lag between applications deployed in dif-
ferent markets.
• A negative return on investment due to globalization initiatives.

Such conditions can therefore hopelessly undermine a globalization ini-

tiative and send teams back to the drawing board in search of better solu-
tions. Overcoming process challenges are therefore highly critical to
achieving success; however, there are other factors that can halt a globaliza-
tion initiative in its tracks, and these challenges fall under the category of in-
tegration challenges.

Integration Challenges

In a globalization initiative, there are two main points of technology inte-

gration which must be tackled to be successful. These areas are (a)
interoperability with source content repositories and applications to en-
able content acquisition, and (b) interoperability with service provider
tools and applications (i.e., translators, reviewers, editors). Both revolve
around technology interoperability. The first involves interoperability
with the technology infrastructure within the enterprise, whereas the sec-
ond involves interoperability both internal and external to the enterprise
with desktop applications.
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 129

As Internet technology has matured, IT organizations have deployed

greater and more diverse varieties of applications within the enterprise to
both manage and benefit from the increasing functionality of this medium.
As a result, there has been a proliferation of database- and/or file-driven ap-
plications for creating and personalizing content, such as content manage-
ment systems (CMSs), personalization systems, and application servers.
To facilitate online commerce and to generate interactivity with business
processes, companies have also deployed procurement systems, logistics
applications, and order processing systems. In addition, other applications
have been deployed to interact with legacy systems, such as databases,
mainframes, and ERP and MRP systems, as companies have increasingly
moved their business operations to an Internet-based environment.
Each of these environments contains content that may need to pass
through a globalization process such that it can be made available to an im-
portant target customer segment. Overcoming this challenge is therefore
the first step in a globalization process. However, integration does not end
with the source and target repositories; it is also critical that content pro-
vided in the globalization process be compatible with desktop tools used by
contributors in the work flow process.
For localization, the tools of choice typically include localized or specialty
word processors and translation tools such as Microsoft Word, Trados Work-
bench, Star Transit, or Déjà Vu. An inability to interoperate with these prod-
ucts may make it impossible to carry out translation and localization of
content and therefore will quickly make globalization impossible.
As this analysis illustrates, it is generally the case that the format for
content processing is often different from the storage format. As a result,
after localization, it is critical that the format be once again modified to
enable the storage repository to store and manage the content. This pro-
cess cycle, therefore, places very significant restrictions on the level of
technology interoperability. If interoperability is low, the difference can
only be made up through human labor, which is expensive and, more im-
portantly, time consuming.
Technical challenges, however, only begin with interoperability. There are
many others that must be considered to maintain the integrity of content dur-
ing a localization process. These other technical challenges are detailed next.

Technical Challenges

On close inspection, each of the information storage and processing environ-

ments within an enterprise contains content in different formats and has dif-
ferent rules or Application Programming Interfaces (APIs) for accessing it. As
global content leveraging can only begin by first sourcing content from the
organization’s content repositories, it is critical that any globalization solu-
130 WOODS

tion be capable of handling three different sets of parameters simulta-

neously. These parameters include content language encodings, content
storage formats, and application and data repository APIs.
An inability to work with any one of the multitude of permutations of these
parameters can lead very rapidly to the demise of any globalization initiative.
As an example, consider first a case of content stored in XML, being
sourced from an Oracle database, in English (ASCII-7 language encod-
ing), in the United States and being targeted to an I-mode server, in Japa-
nese, for distribution in Japan. To enable this transaction to occur, there
need to be three successful transformations. First, the content needs to be
localized into Japanese and then stored in either the Shift-JIS or ISO
2022-JP language encoding. Next, the markup format needs to be trans-
formed from XML to cHTML (condensed HTML), and a template needs
to be applied with a different screen format configuration due to the dif-
ferent device requirements. Finally, the content will need to be mapped
from the Oracle database in the United States to a Japanese version of
SQL-Server in Japan.
Now consider an alternate scenario occurring simultaneously with the
process just described: a Chinese HTML page stored in an enterprise CMS
being targeted to a WAP server in France, or JSP-based information in a
personalization server under a UNIX/Oracle configuration in the United
States being targeted to a Brazilian, NT-based ASP server. Obviously, the
combinations are enormous and interoperability across all major lan-
guages, formats, and applications/repositories is the only road to both cur-
rent and future success.
Three additional issues now become clear:

1. The systems that are interacting in this example need to support

double byte character sets, that is, character encodings such as
those for Chinese, Japanese, and Korean.
2. Transformations need to be made on items such as date and time
format, naming conventions, and weights and measures.
3. Storage of information within these systems must allow for linguis-
tically correct translation (i.e., no concatenation allowed).

In short, it is critical that these systems be internationalized so as to in-

nately handle these three issues. Without an internationalized infrastruc-
ture, there is the high probability for data corruption, loss of content
integrity, and the generation of meaningless information.
By conquering all three of these challenges simultaneously, an organiza-
tion now has the ability to create the foundation for content globalization
success. This foundation is explained in greater detail in the next section on
best practices.
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 131

BEST PRACTICES

Key Terms

I often find that any discussion of multicultural content is best understood

once five key terms are explained and agreed on. These terms and their
definitions with regard to enterprise content are as follows:

1. Globalization: a strategic undertaking whereby an organization cre-

ates and enforces standardized business processes on a worldwide basis.
2. Internationalization: a lower level strategic undertaking whereby
an organization develops and/or deploys people, processes, and systems
that perform in ways that are independent of language and culture. This
is often thought of as an architectural undertaking that is carried out
once and then maintained afterward.
3. Localization: a tactical process through which content in one lan-
guage, culture, and country is interpreted to capture the nuance and con-
ceptual meaning for another language, culture, and country.
4. Translation: a tactical process in which content in one language,
culture, and country is transformed to have a similar wording in another
language, culture, and country.
5. Personalization: a strategic initiative through which an organiza-
tion develops processes to provide specific content that is tailored to ap-
peal to the preferences of a particular individual. On a global basis,
personalization begins with language and culture.

The essence of these definitions is that there is a hierarchy to these terms.

So, localization is dependent on translation; however, it is greater than it.
Globalization and personalization are dependent on internationalization,
localization, and translation; however, they are greater than all of them
combined. Understanding the relationships inherent in this hierarchy, in-
terestingly enough, lays the foundation for discussing the primary best
practices for successful content globalization.

The Three Pillars

Based on my experiences over the last 5 years, three critical best practices can
be summed up as the Three Pillars for successful globalization. Each of these
needs to be done in succession, and when all three are in place, a stable founda-
tion is formed for a globalization initiative. The Three Pillars are as follows:

I. Strategy: Development of a high level documented strategy which

has the global buy-in of all significant stakeholders;
132 WOODS

II. Infrastructure: Development and management of a fully interna-

tionalized infrastructure;
III. Global processes: Deployment and enforcement of standardized
and coordinated content localization processes on a worldwide ba-
sis.

Pillar I: Strategy. A globalization initiative, by its very nature as a world-

wide project, is strategic in nature. No matter how small, there are ramifica-
tions in multiple countries, in multiple languages and cultures, which affect a
company’s perception and positioning. As a result, a globalization initiative of-
ten becomes a highly politicized event, with multiple groups jockeying for con-
trol, visibility, and direction.
As a best practice, setting the strategy is similar to the design phase for an
art or engineering project. This stage is critical because without it, the direc-
tion and goals of the initiative would otherwise be poorly defined. Without
well-defined goals or direction, communication across countries, lan-
guages, and functional groups becomes nearly impossible. Where possible,
the message will be, at best, confusing and will not enable groups to engage
in meaningful discussion.
Instead, the strategy for the initiative should be developed by a working
group of key, high-level individuals and then openly discussed as a living
document among all significant stakeholders. It is mandatory that the indi-
viduals driving the initiative be at as high a level as possible in the organiza-
tion to ensure that the results of the document be binding and be
enforceable throughout the organization.
Conversation among the different groups will then have the chance to
stimulate truly global thinking among the participants. Such thinking will,
in turn, uncover challenges and lead to the creation of broad-based solu-
tions. The strategy document can then be iterated until a compromise is
reached which clearly articulates benefits to the global stakeholder commu-
nity. Once approved, this vision document will form the basis for the road
map for the initiative.

Pillar II: Infrastructure. Once the business vision for the globaliza-
tion initiative is in place and blessed by the stakeholders, it is then possible
to consider the infrastructure implementation issues for the project. It is at
this point that the team queries its infrastructure and decides whether or
not it has sufficient and/or appropriate resources, systems, and processes to
support the initiative.
Unfortunately, it is all too common that teams develop a business vision
and then find out during the implementation process that their goals are
unsupportable. Due to the technical nature of electronic media, it is critical
that the business vision be coupled with technological due diligence. Any
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 133

gaps between what is called for in the vision and what is available can then
be remedied prior to project initiation.
As an example, consider a team that has decided to roll out multicultural
content into the Eastern European and Asian marketplaces. This team de-
cides on its direction, goals, and road map but then neglects to investigate
the resourcing and technical aspects of the project. Once underway, the
team realizes that it does not know how to support double byte character
sets (DBCSs) for the Asian marketplaces or the varied scripts of Eastern Eu-
rope. Also, its applications for developing content do not support storage,
input, or rendering of the character sets. Furthermore, the team’s lack of
access to multicultural resources leads unknowingly to the development of
template colors and icons that are highly offensive in the target regions.
Such a scenario can be avoided by carefully evaluating the human and
technological resource infrastructure to appropriately develop, manage,
and update content for the target regions. Thinking through these issues
and making changes to the infrastructure prior to project kick-off will then
reduce the chances for surprises during the implementation phase, or will
make the surprises introduced more manageable.

Pillar III: Global Processes. Although often overlooked during a

planning phase, the life of a globalization project is typically much longer
in duration than the initial project development cycle. As a result, prepar-
ing for and deploying the infrastructure to support updates and ongoing
work is just as critical for success as the initial strategy and infrastructure
considerations.
As was noted in the earlier section on challenges, one of the greatest hur-
dles to any globalization initiative is the process challenge. Not only can this
challenge create a massive financial and resource drain, it can nullify all ef-
forts such that no noticeable gains are produced.
The first step to overcoming process-based challenges is to design con-
tent localization processes (i.e., work flows, content pre- and postprocess-
ing, etc.) that meet the business needs of the enterprise on global, regional,
and local levels. Once this design is completed these processes can then be
automated via a standardized infrastructure. The final step, and another
key to sustainable globalization success, resides in having visibility to, and
the power to enforce, these processes on a worldwide basis.
With global processes in place, the update of content to a worldwide au-
dience becomes routine and, in most cases, predetermined. Crises are
greatly reduced, as most situations have already been addressed, and there
are mechanisms in place for dealing with problems and surprises. Further-
more, worldwide teams are tightly integrated by the global processes, which
means that work is completed quickly, efficiently, and to the satisfaction of
all parties involved.
134 WOODS

Central Oversight With Local Empowerment (COLE)

In addition to the Three Pillars, the fourth key best practice is known as cen-
tral oversight with local empowerment (COLE). To understand this con-
cept better in context, it is first instructive to discuss alternative develop-
ment strategies: centralized and decentralized development.

Centralized. Under a centralized development scenario, a company, di-

vision, or group manages all of its content creation, technology, and human
resources in one location. Advantages to this approach include the following:

• Resources are coordinated with one voice.

• All business processes receive good visibility.
• Consistency can be enforced.
• Costs and scheduling receive good oversight.
• Single technology infrastructure minimizes technical challenges.

Though appealing because of these strengths, this approach has a signif-

icant downside, based primarily on the fact that the centrally developed
content is actually to be targeted to other regions. As has been noted, these
other countries may have significantly different language, cultural, tax, le-
gal, regulatory, and economic environments that will need to be taken into
consideration. If these factors are not taken into account, then the entire
centralized initiative can be called into question.
The list below summarizes the weaknesses of the centralized approach:

• Costs for multicultural consultancy can rise exponentially.

• Content may not conform to the needs of countries targeted.
• Scheduling may not meet the needs of country offices.
• Technologies and media offered may not meet standards in target
countries.
• Country offices may attempt to redo work created centrally, leading
to rising costs.

Centralized development may work in limited cases; however, in the

long term this approach usually leads to significant problems. Another ap-
proach, which is opposite in philosophy to the centralized approach, is typi-
cally called the decentralized approach. It is discussed in more detail next.

Decentralized. Under the decentralized approach, a company, such as

a multinational, empowers each of its country offices or subsidiaries to act
autonomously to produce content for its local needs. This approach clearly
addresses the weaknesses of the centralized approach by putting each re-
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 135

gion in the driver’s seat for its content. As a result, the strengths of this ap-
proach include the following:

• Country content is focused on the specific needs of each region.

• Locale-specific issues, such as language, culture, and currency, are
addressed automatically.
• Scheduling of content releases meets the needs of each office.
• Promotions for partners, holidays, and local events are locally con-
trolled.
• Content is actively tied into the country offices’ sales and marketing
plans.

As with the centralized approach, however, there are issues that create
problems for the decentralized approach. Most of these problems revolve
around the development of severe process and integration challenges that
hinder the company’s ability to coordinate distributed teams and maintain
costs. It should be noted as well, that the greatest challenges arise in an on-
line environment due to the ease with which information from different
countries can be accessed and compared. The list below summarizes the
problems associated with the decentralized approach:

• Lack of central oversight makes coordination on a regional or

global basis nearly impossible.
• Significant incompatibilities may arise in the technological and
process infrastructure so that different offices cannot work together
effectively.
• Costs skyrocket due to redundancies in resources and technology
across offices.
• Global branding, messaging, and look and feel becoming highly
fragmented.
• Inconsistencies and contradictory information across countries can
create increased customer service and customer loyalty issues.

Interestingly, one can see that the strengths of each approach, central-
ized and decentralized, cancel out the weaknesses of the other. A hybrid
approach, which combines the best of both of these worlds, is known as
central oversight with local empowerment. This approach is detailed in
the following subsection.

Central Oversight With Local Empowerment

As discussed in the previous sections, the strength of the centralized ap-

proach lies in its ability to coordinate and oversee processes with a single
136 WOODS

voice and technology infrastructure, whereas for the decentralized ap-

proach, its strengths lie in enabling local offices to fine-tune content to meet
the needs of their locale. The optimal approach would then provide a cen-
tralized infrastructure that oversees and coordinates all processes, while all
processes would route content through the local offices to enable them to
develop and/or review content for their markets.
This approach, known as central oversight with local empowerment, has
proven to be one of the most powerful assets to companies such as HP, GE,
Xerox, and Dell, all of which have globalized effectively. It should also be
clear from this discussion that this approach is fully complementary to the
principles of the Three Pillars.
Armed with these best practices, companies now have the knowledge
and direction to enable them to develop and manage multicultural con-
tent sustainably. The following sections go into greater detail into how
these best practices can be implemented within the enterprise to generate
financial success.

THE DETAILS OF IMPLEMENTING BEST PRACTICES

So how does an enterprise, division, or group use the concepts of best

practices to move itself closer to the vision of an ideal enterprise? The fol-
lowing sections detail the steps a company should take with each of the pil-
lars in order to implement a content globalization initiative with a high
probability for success.

Pillar I: Strategy

At the highest levels, the strategy for a globalization project is quite a com-
plicated undertaking, as it incorporates large amounts of financial analysis,
risk factors, competitive issues, and so forth. These issues are not in the
scope for this chapter or this book, so I assume that the business globaliza-
tion strategy is in place prior to the development of the content globaliza-
tion strategy. To simplify the discussion of the content globalization
strategy, consider it as moving through three distinct phases: (a) definition
of goals and scope, (b) communication and cooperation between stake-
holders, and (c) road map and metrics.

Definition of Goals and Scope. The first step in creating a strategy

for a content globalization initiative is to synthesize from the business
strategy three primary constituents: the goals of the initiative; the
breadth of the initiative (i.e., the number of countries and teams in-
cluded); and the depth of the initiative (i.e., the types of products and
content to be included).
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 137

As an example, consider the following fictitious plan for a U.S. mobile

device company that is planning to increase sales in Asia:

Problem: XYZ Company has noted that the Asian market is the largest mar-
ket in the world for mobile electronic devices such as phones and pagers.
However, XYZ Company has only a small presence in Japan and Australia
and some penetration through distributorships in the other major countries.
Goal: XYZ Company will increase its market share from current levels (<
1%) to at least 5% in the primary Asian countries within 12 months of prod-
uct launch. It will carry out an extensive launch and promotion campaign
within 3 quarters by selling its key North American products. The initiative
should reach breakeven within the first 6 months and then become profit-
able with a growth rate of 10–15% afterward.
Breadth of the initiative: Research indicates that the primary Asian markets
include the following eight countries: China, Japan, Korea, Taiwan, Malay-
sia, Thailand, Indonesia, and Australia. Teams from the two offices, Japan
and Australia, as well as those in the distributorships will be leveraged. Other
offices will be set up over the course of the project.
Depth of the initiative: Three key products and all their supporting collateral
will be launched into the markets. The company home page, product, and
support sections as well as two commerce applications will also be globalized to
support the launch and ongoing marketing and sales activities.

With the goals, depth, and breadth of the project defined, enough infor-
mation exists about the initiative to begin to communicate about it with the
other country teams. It is at this point that the stakeholders for content glob-
alization can be defined, contacted, and brought into the strategic process.

Communication and Cooperation Between Stakeholders. Choosing

stakeholders is a critical step that can have political ramifications if not car-
ried out wisely. To begin, it is recommended that a senior individual from
the group leading the initiative create a committee to represent the global-
ization project. After the group has been created, it should be officially rec-
ognized and given a charter by members of senior management in order to
imbue it with legitimacy.
At this point, invitations should be made to international deputies to en-
able each country office to provide input. Individual country stakeholders
should be chosen on the basis of their position and ability to create action in
their organization. Alienating key individuals who may be threatening or
including individuals with little power is not recommended.
To kick off the strategic session, it is recommended that all important
stakeholders be brought together so that they can meet and discuss with
138 WOODS

each other in person. Such a level of interaction is critical at the kick-off

of the project but may not be necessary on an ongoing basis unless the
initiative is very large, politically charged, and/or complex. Clearly, how-
ever, ongoing interaction and communication is critical to enable the
team to work through its issues and develop a plan outlining the recom-
mended strategy.
Time should be set aside to deal with the concerns of each office and de-
velop solutions that are both practical and applicable to as many countries
as possible. Should problems persist or should issues escalate beyond the
capabilities of the team, it may be necessary to go to outside experts for help
and a path forward.
Consider once again, the example of the North American mobile device
company launching its products in Asia:

Consider that the CEO gives the mandate to the Senior Vice President of
Worldwide Marketing, Nancy Liu, to carry out the marketing side of the
product launch. Nancy would then invite her marketing counterparts in cor-
porate and product marketing, as well as channel partners, distributors, and
information services, to be stakeholders in the initiative.
With a significant company presence in Japan and Australia, these two of-
fices would clearly have input on the marketing and technology aspects, and
so several members would be invited from each team. Other stakeholders
would be drawn from the one key distributor for the remaining countries
plus the operational lead from the team in the new Chinese office.
Issues to deal with would include a strong belief from the Japanese and Aus-
tralian offices that they are self-sufficient, not needing any help or input
from the United States. The distributor would be interested in charging for
additional overhead associated with the launch. Meanwhile, the Chinese
team would claim not to have time or resources, but would most likely hire
consultants to complete the task.

Part of Nancy’s role would be to push for a strategy and process that incorpo-
rate COLE. She would need to work with the offices to have them help each
other and buy into a single overarching framework for the project. Addi-
tionally, all major cultural, linguistic, legal, technical, and regulatory issues
would need to be discussed to ensure that these would be taken into account
in the process. Once a solution is reached, the team would document the
strategy and put forth a road map for its execution.

Creating metrics and a road map enables the team to measure their per-
formance and schedule out the use of resources. It also enables the team to
synchronize their efforts more effectively and work toward common goals.
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 139

Road Map and Metrics. A team-developed road map is critical to the

success of the initiative because it ensures that the global team is focused on
the same goals during a well-defined time frame. Milestones and deadlines
keep the project on track and facilitate monitoring of team performance
and the overall viability of the initiative. Due to this structure, it is also less
likely for team miscommunication to occur as the objectives are well docu-
mented and mutually agreed on.
Metrics should be put in place both for the deployment of the initiative,
often known as the development phase, and the ongoing operations, or
maintenance phase, to provide visibility to the project’s success. Different
metrics may be of value to different initiatives and organizations; however,
there are three metrics that should be applicable in most circumstances:

1. On time/on budget metric—this metric is a measure of how well the

team conformed to its projected timeline and budget for initiative
deployment.
2. Return on investment (ROI) metric—this metric measures the fi-
nancial return as a function of the money invested in the initiative.
3. Customer loyalty index or similar—this metric measures customer
perception with regards to how well the company’s initiatives and
behavior coincide with customer desires.

Results from these three metrics allow the team to understand which ap-
proaches were successful and which were not and to develop its own specific
best practices regarding meeting customer needs, fine-tuning intracompany
collaboration, and optimizing automation of its globalization processes.
As an example, consider once again the North American mobile device
company launching its products in Asia:

After a number of good international meetings, Nancy Liu is able to under-

stand the dynamics between the different offices and decide on the identi-
ties of her key stakeholders. Maintaining a good dialogue between these
individuals or teams, she then leads the crafting of an outline of tasks, re-
sources, and responsibilities that will evolve into a road map and milestone
chart for the initiative.
Under this dynamic, the team would very quickly realize that they are all im-
portant to the success of the project but that they must all compromise if the
project is to be implemented. For example, the Australian and Japanese of-
fices may feel that they are self-sufficient but will quickly realize that
resourcing needs, project costs, collaboration issues, and interoffice syn-
chronization is best shared with Nancy’s team at headquarters and other of-
fices. Meanwhile, Nancy will see that leveraging the other offices as much as
140 WOODS

possible will speed time to market and reduce responsibilities and require-
ments of her team.
Once the road map is laid out, milestones defined, and the metrics put in
place, the teams are now on the same page and can move to the detailed cost-
ing, resourcing, and timing issues. Once these issues are resolved, the team
can now move forward with understanding the limitations and strengths of
their infrastructure.

With a strategy and road map in place, the team is now clear on how it
would like to proceed. Before actual implementing, however, it is critical
that the team evaluate its infrastructure and process to ensure that it can ac-
tually handle the demands of its plans.

Pillar II: Internationalized Infrastructure

Internationalization as Best Practice. Most companies, unfortunately,

are unaware of the true capabilities of their organizations, both good and
bad, with regard to the readiness of their people, systems, and processes to
function internationally. This is a striking assessment given the vast numbers
of global companies; however, it is critical to realize that most global compa-
nies are actually a loose federation of separate, local entities.
With this vision in mind, it should be clear that most global entities de-
sign products and services for a particular market, such as the United
States, and then ask each additional targeted market to create or customize
similar offerings locally. This process is highly inefficient as little is gener-
ally leveraged from the original products or services. As a result, costs for lo-
calization are high, time frames for launch or update are long, and quality
can be questionable.
The topic of product localization and internationalization has been cov-
ered in depth earlier in this book; however, it is important to note that inter-
nationalization and localization do not only cover product design and
coding. Actually, internationalization is an issue that pervades every part of
an organization, from the copy written for a manual to the topics of meet-
ings in boardrooms.
Internationalization is a best practice that, once implemented across
people, systems, and processes, allows a company to collaborate globally
and harness the full potential of its worldwide presence as a competitive ad-
vantage. Without it in place, the costs and resourcing necessary to work with
incompatible systems that will corrupt and destroy data can eliminate any
project ROI and make success nearly impossible.
Consider, for example, a very large and respected U.S. company that de-
signs and builds commerce systems based on the EDI standard. Prior to inter-
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 141

nationalizing their product infrastructure, they developed a U.S. product

and then proceeded to send it to overseas offices to localize in country. After a
few releases, the company realized that these had multiplied to over 30 re-
leases in the field, none of which shared the same code base, contained the
same feature set, or were well documented internally or externally.
To add to the confusion, each of the company’s offices indicated that
even after extreme customization few if any of the products were well ac-
cepted in their local marketplace and had very low uptake. Suffice it to say
that internationalization helped to consolidate all products worldwide onto
the same code base and rapidly increased product acceptance and sales in
all countries offered.

Implementation Plan. Once a project has been fully scoped, it is possi-

ble to understand if the infrastructure is in place to support its implementa-
tion. For example, a checklist should be developed which contains sample
questions such as the following:

• Do all systems, processes, and people support the creation, trans-

port, and storage of content for multiple target markets in different
languages?
• Is the concept of country or language supported by each compo-
nent of the system?
• Can the system accommodate delivery of different content to dif-
ferent markets based on user requirements?
• Is all text separated from code for ease of localization?
• Has all concatenation (splicing together of parts of a sentence to
create a full sentence) been removed from the infrastructure?
• Can different weights and measures as well as date format be stored
and displayed?
• Do authors create content for an international marketplace or a
particular country?
• Has the content work flow process been developed to include trans-
lation as well as in country content development?

The main purpose of these questions is to uncover areas of the infrastruc-

ture that may prove problematic to the implementation of the overall vi-
sion. As an example, consider that after the strategy has been developed,
Nancy Liu and her support teams carry out an audit on their assets.
After a comprehensive audit, the team discovers the following issues that
need to be resolved:

• The operating system for the mobile devices does not support dou-
ble-byte characters (such as Chinese, Japanese, and Korean).
142 WOODS

• Datasheet content has been written in Quark, and it is unknown

how to make this information available in Asian languages.
• The Web sites in each country (United States, Japan, Australia, and
China) are operating using very different content creation and
publishing environments.
• 50% of existing U.S.-based web applications do not support dou-
ble-byte characters, contain no fields for important Asia informa-
tion, do not support multiple currencies, dates, and address
formats, and use a significant amount of concatenation.
• U.S. product colors and cultural icons are inappropriate for all tar-
geted markets and will have to be reworked.
• The primary channel for content distribution in Japan is I-Mode, a
mobile device format for which the company has very limited content.
• There is significant project risk as the team has limited experience
with simultaneously launching and maintaining products in differ-
ent countries with different languages and cultures.
• There is no infrastructure in place for coordinating and updating
team members on a worldwide basis with regard to important activ-
ities and milestones.

With this information, Nancy and her team are able to create a business
plan, requesting budget for the internationalization and localization pro-
jects that they need to undertake. They also can create specific Requests for
Proposal (RFPs) and submit them to third-party vendors that have exper-
tise in internationalization and localization. Given the team’s lack of experi-
ence in these areas, the third-party providers should enable the team to hit
their aggressive timelines. Subsequently, the team is also able to schedule in
realistic times and phasing for the completion of these initiatives prior to
product launch.
With a clear understanding of the range of implementation challenges
that they need to overcome, the team can now plan and prepare for the on-
going life of the globalization initiative. It cannot be overstressed that the
ongoing management of a globalization initiative is the greatest challenge
to a team or organization. Planning and implementation have specific
deadlines and a limited scope; however, the ongoing project is one of con-
stant change and growth. Managing, adapting, and controlling the ongo-
ing global processes is the ultimate key to success.

Pillar III: Global Processes

Efficient global business content processes typically rely on localization.

There are two primary reasons for this. First, people are three to four
times more likely to purchase products when they are presented with in-
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 143

formation that is in their native language (DePalma, 1999). Second, con-

tent localization is on average one third the cost of content creation
(Drakor, 2001).
Here it is important to note that localization is appropriate where source
content can be leveraged into a new market and still be appropriate. So, for
example, a global press release regarding a company’s record quarterly
earnings would be appropriate as it is of global interest. However, informa-
tion regarding a product that is only available in the United States and has
limitations that make it unusable in other countries (i.e., mobile devices that
do not conform to local protocols) would not. Additionally, it should be kept
in mind that information that does not conform to a consumer’s preferred
environment (appropriate currency, voltage, cultural attributes, tax and le-
gal system, government regulation, etc.) can make purchase of products
and services very difficult or impossible. It is also important to understand
that content publishing for global markets may rely on a hybrid approach in
which specific information is localized and then combined with locally cre-
ated content. Reasons for such a hybrid approach could include (a) im-
proved management of a company’s global brand and messaging while
simultaneously ensuring that information is locally relevant; or (b) better
utilization and leverage of enterprise-wide resources.
Consequently, before tackling globalization processes, I first explain
conventional content localization processes as they nearly always form the
foundation for a globalization initiative.

Conventional Localization Process. Conventional localization process

flow includes tasks such as the following (see Fig. 5.1):

1. Change detection on source content repositories.

2. Code and content separation.
3. Content and format normalization.
4. Translation.
5. Translation memory and terminology leveraging.
6. Contextual review.
7. Work flow management and monitoring.
8. Synchronization of content with other locales.
9. Quality assurance.
10. Mapping of localized content to target repositories.

As should be clear from this list, and as has been borne out by extensive
analysis, localization costs are generally about three to five times the cost of
translation alone.
Each of the steps in the previous list can be a very time-intensive and gru-
eling task if carried out manually under tight deadlines on an ongoing ba-
144 WOODS

FIG. 5.1. Localization process.

sis. As volumes increase, timelines tighten, and the number of target

markets increases, costs escalate exponentially due to rising resource
needs, management overhead, and increased human handling, interaction
and environment complexity. If complexity increases to the point where
the localization process is too slow to meet deadlines or too inefficient to
create relevant content, then a return on investment (ROI) becomes impos-
sible and the investment will never be recouped.
The key to localization process savings is taking situations that are com-
plex and highly leveraged and simplifying them so that they become simul-
taneously routine and rapid. A combination of process streamlining and
automation of the tasks in the list above (see Fig. 5.2) can lead to a 15–25%
reduction in costs for localization of new content, and an 80–90% reduction
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 145

FIG. 5.2. Streamlined process.

in localization costs associated with content updates. A good rule of thumb

is that even for automation of small projects, say 100 pages of content lever-
aged into three locales with 10% changes per month, cost savings should av-
erage hundreds of thousands of dollars on an annual basis.
Once again, consider Nancy’s team at the point where they are develop-
ing their plans for the ongoing support for the globalization initiative. Spe-
cifically, the team must have a plan in place that addresses content,
information, and support needs post launch.
Nancy already has understood that decentralized management and op-
eration of the Asian launch will be very expensive and prone to failure. As a
result, she has developed a plan based on Central Oversight with Local Em-
powerment (COLE). Under such a plan, she has centralized her primary
technology infrastructure, project management, and project drivers in the
United States. Each country office in Asia will be responsible for content re-
view, local content creation, and, in some cases, content translation.
146 WOODS

Given the many different types and the large volumes of content that will
need to be localized rapidly on an ongoing basis, Nancy explores different
technology options for boosting efficiency and quality. To this end, she be-
gins a competitive assessment of localization firms, translation tool provid-
ers, and globalization management system (GMS) providers. By choosing
an appropriate partner, she hopes to ensure that she can coordinate and le-
verage her critical corporate information assets on a global basis and consis-
tently hit aggressive localization deadlines.
Nancy also creates a competitive process to review vendors for the man-
agement of her source content assets. To this end, her team reviews enter-
prise content management systems (ECM), portal solutions, and source code
control solutions. By combining an appropriate solution for content creation
with an infrastructure for global content leverage, Nancy’s team can be more
confident that they will be building on best practices and generating effi-
ciency and scalability through process automation and streamlining.
The next sections explore in detail the most critical implementation as-
pects for generating and maintaining success during the ongoing growth
and change of a globalization initiative.

WORK FLOW AND PROCESS AUTOMATION

Content change is typically the initiating action in any globalization pro-

cess, however, there are many other steps that must take place before a
globalization process is complete. A typical globalization process can be
subdivided into four primary stages: (a) process initiation and content ac-
cess, (b) content preprocessing, (c) human-based work flow activities, and
(d) content postprocessing.

Content Accessibility

Digital content can be stored in one or both of two ways. It can be stored as a
file in a file system (e.g., a Word document), as data in a database (e.g.,
product descriptions as cells in a database table), as a combination of these
two approaches (e.g., an ASP, JSP, or Cold Fusion document). As a result,
before a document, web page, or specific content string can enter a global-
ization process, it is imperative that first the content be accessible.
Access can be complicated in multiple ways. The most common ways
that content access can be rendered difficult include proprietary storage
formats, distributed storage, undocumented or poorly documented stor-
age, proprietary presentation formats, access controls (ACLs), encryp-
tion, encoding, and access through other applications. In some cases,
companies reduce content accessibility and visibility to improve security;
in other cases, it is done to improve content or process management,
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 147

whereas in still other cases, it may be simply accidental as a result of an im-

plementation approach.
Uncovering your company’s or client’s particular challenges with regard
to content accessibility is critical for the development of any sustainable
globalization solution.

Content Preprocessing

Digital content can be acquired in a variety of formats. These can include

popular word processing formats, such as Word or WordPerfect; powerful
business presentation formats, such as Quark or Framemaker; Web-centric
formats, such as HTML, XML, and Javascript; coding formats, such as
C++, Perl, and Java; or potentially database formats accessible via SQL.
A key point to note is that any format can be represented as a combina-
tion of code and content. The code typically gives instructions regarding
presentation and potential functionality whereas the content is a combina-
tion of textual information plus graphics.
In a localization process involving nontechnical translators and reviewers,
it is generally unlikely that any changes will need to be made to the code. As a
result, to streamline a localization process, it is critical that content be sepa-
rated from the code to simplify the tasks of translation and review.
There is a catch, however, for without the presentation layer or the func-
tionality, the context for the content may be lost, thereby complicating lo-
calization tasks. As a result, content to be localized needs to be isolated while
at the same time retaining contextual information. This approach maxi-
mizes productivity while minimizing the need for rework.
In addition to format, content can be acquired in different language
encodings. So, for example, content acquired in the Japanese language will
often be encoded as shift-JIS, whereas content in French would be encoded
as ISO-8859-1, and content captured in simplified Chinese would be en-
coded as GB2312. As a result, language translation for digital information
leads to a transformation of the encoding as well.
To simplify the processing of language, it is recommended as a best prac-
tice that all content to be processed be first converted to Unicode. Unicode
is a single encoding for all the world’s languages which simplifies language
translation by limiting all language processes to a single encoding with
well-defined rules. To summarize, there are six steps which must take place
to make digital content ready for processing:

1. Separation of code from content.

2. Storage of code for contextual presentation/functionality.
3. Conversion of content to Unicode.
4. Segmentation of content into sentences and/or paragraphs.
148 WOODS

5. Separation of content into translatable and nontranslatable seg-

ments.
6. Word count for budgeting/timing purposes.

Additionally, it is recommended that all content not only be stored as

Unicode but be embedded into an XML-based “container.” This container
should contain metadata (information about the data) for preserving the in-
tegrity and fidelity of the document being processed under a round-trip oper-
ation. That is, if I begin with a Word document, I want to ensure that I get a
Word document back with the same formatting as the original after localiza-
tion; the same should be true for a Quark document, an HTML page, or a JSP.
XML is recommended for the container format, as it is a neutral format that
is rich enough to preserve the formatting for the most complex documents. To
maximize productivity, it is also recommended that all the activities involved
with the preprocessing of content be automated as much as possible.

Human-Based Work Flow Activities

Once information has been preprocessed, it is now possible to route it into a

human-based work flow for translation, review, and potentially local con-
tent additions. Manual work flow processes contain many bottleneck points
that lead to high process inefficiency. Automation of work flow processes,
on the other hand, can greatly streamline operations leading to much
higher throughput, better productivity, and higher quality.
Business processes typically exist within companies; however, they are
usually poorly defined or are often reinvented each time they are repeated.
For example, a company may publish press releases every week and then
distribute versions in 10 different languages across 18 countries, or a com-
pany may make daily updates to its online product catalog and localize
these updates into the European marketplace. In each case, there are par-
ticular initiating activities, which should then be followed by set processes
(i.e., distribution into certain languages and marketplaces). Unfortunately,
the necessary processes often are not set, and lack of clarity leads to mis-
takes, missed deadlines, and poor quality.
Automation of work flow processes is not always possible. For automation
to be effective, there are a minimum of two prerequisites that should be in
place. First, content under a globalization initiative must be categorized or
easily categorizable. Second, business rule templates and an infrastructure
for initiating business rules must be put in place. Both are necessary, for
business rules cannot be initiated without content categorization and con-
tent categorization alone cannot lead to automation of processes.
Categorization is simply a fancy term for the type of content, for exam-
ple, a press release or a product catalog item. With this descriptor in place,
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 149

it is then possible to associate the content with a particular business pro-

cess. A codified version of a business process is typically referred to as a
business rule. So, for example, consider the press release scenario out-
lined earlier in more detail.

Press Release Business Process

Every Thursday, an English-language financial press release is created

for distribution for the following Tuesday in 10 languages for 18 mar-
kets. The original document is created in Quark format and is two pages
long. Once the press release is created, it is then sent to five teams in suc-
cession, these are:

• An editor who checks the style and formatting and ensures that the
content is globally relevant and easily translatable.
• Translators with particular language expertise and expertise in
marketing and financial translation then translate the content.
• In country marketing managers then review the press release to en-
sure that messaging is consistent with their country.
• The in-country legal team then checks the content to ensure that
the content will not create any legal exposure.
• Then information is given a final review by an in-country public re-
lations team representative.

Once this process is understood, it can be turned into a standard business

rule template that can then be initiated automatically. The business rule can
be summarized as follows:

• Every Thursday initiate a work flow process for new content in

Quark format in the financial press release folder on a particular
server.
• Route this content through five steps: de-Americanize, translate,
review, legal, and final review.
• Associate relevant individuals with each step of the work flow process.
• Content must be fully processed by Monday of the following week.

Now, with a combination of content categorization and business rules,

the process of globalizing this press release can be carried out automati-
cally on a weekly basis. Obviously, some level of human supervision is nec-
essary; however, all of the manual hand-offs, phone calls, format
problems, and second-guessing associated with the fully manual process
are no longer necessary.
150 WOODS

Content Postprocessing

The final stage of the process is often termed postprocessing. It is in this

stage that the content is put into its final form prior to distribution. Postpro-
cessing usually occurs across three axes: format, encoding, and location. So,
if as in the previous example, the press release begins as a Quark document,
it should complete its localization process also as a Quark document with
the same formatting as the original. For print and web documents, there
can be significant reformatting work necessary as different languages gen-
erally have different linguistic rules and take up different amounts of space
on a page. So, for example, German, French, and Japanese translations of
English content often show 20–50% horizontal text expansion or text swell,
whereas Chinese and Korean typically see horizontal text contraction and
vertical text expansion.
Such changes in text size often require different formatting for such
items as pagination, table sizes, text boxes, and so on. These changes can be
made manually postprocess, or they can be facilitated during document lo-
calization during the work flow. In any case, formatting will change between
the original language version and the local language version; thus, modifi-
cations will need to be made.
Given that the language has gone through a translation process, it is criti-
cal that the encoding for the content be appropriate for the language. It
should be noted, that though Unicode is a good format for processing mul-
tilingual content, it is very rarely a good format for exporting content for
distribution.
The reason for this seeming contradiction is that today, very few stan-
dard desktop applications use Unicode as a native content format. As a re-
sult, Web-based content and that for Quark, Frame, and Word must, in
most cases, be converted to national language encodings (NLEs). NLEs are
the native encodings that most desktop applications use to render different
languages on an electronic device. Common NLEs include Big5 for com-
plex Chinese, KOI8 for Russian (Cyrillic), and Latin-1 for French.
In addition to format and encoding, postprocessing must also specify a
storage location for the content. Storage location determination is critical
for process optimization because without it, the information will either be
lost or could overwrite other important information. Storage location infor-
mation should specify not only which repository the content is to be stored
into, but also a unique address point within this repository that follows some
sensible content categorization scheme.
By streamlining and automating the preprocessing, work flow, and post-
processing stages of content globalization, a team can rapidly deliver a re-
turn on investment to a globalization project. Another critical factor, which
can greatly reduce costs and time and improve quality, is the use of linguis-
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 151

tic technologies. The types and benefits of linguistic technologies are cov-
ered in detail in the following section.

LINGUISTIC TECHNOLOGIES

Today, there exist three broad classes of linguistic technologies that can be
of benefit to content globalization initiatives: (a) translation memory (TM)
engines and repositories, (b) terminology management systems, and (c)
machine translation (MT) engines.
Each of these technologies has the potential to be employed singly or in
tandem to reduce costs, speed time to market, and improve quality. The fol-
lowing subsections explain each technology in more detail.

Translation Memory (TM)

In its simplest form, translation memory, or TM as it is commonly known, is a

matching between a sentence and its human-translated equivalent in another
language. A TM is often built as a consequence of the translation process, with
more “matches” or “source-target pairs” created as translation proceeds. Over
time, a repository of such pairs can be mined or translations “remembered” to
reduce the cost and time of translation while increasing consistency.
The power of TM relies on the “leveraging” of past translations to reduce
the amount of current and future translation. Experience has shown that
approximately 20-30% of a single Web site or a large document can contain
redundant information. As a result, a TM produced from a single site or
document can reduce costs and workloads by at least 20%. A realistic limit
for a good TM produced over a few years is that 50% of content can be lever-
aged directly from the repository. With the addition of a fuzzy TM match-
ing engine, or an engine that can deduce matches based on a percentage
similarity, workloads can be reduced by an additional 50%.
Additional factors that can increase the power of a TM include con-
trolled language authoring and the application of the TM to content on the
same subject or produced consistently for the same team. Controlled lan-
guage authoring increases the efficacy of TM by ensuring that content is
constructed linguistically the same way each time the same concept is writ-
ten. In contrast, application of the TM to similar content increases efficacy
by ensuring that the concepts to be leveraged are similar from the start.
Interestingly, due to its reliance on human translation, TM is considered by
many to be the most reliable and useful linguistic technology that exists today.

Terminology Management

Terminology management technology, or glossary management as it is

sometimes known, is similar to TM; however, rather than dealing with sen-
152 WOODS

tences or paragraphs, it is concerned solely with words or short phrases. Ad-

ditionally, terminology is often a linguistic asset that is dictated prior to a
translation process rather than being a by-product of it. Terminology is ei-
ther created by a particular company’s marketing department or it is dic-
tated by usage within a particular industry.
To be useful, a terminology management system must be able to do more
than aggregate information. As terminology is a by-product of marketing
or of an industry, it is important that the teams that use it can edit it and
change it over time as terms are added, deleted, or modified. Furthermore,
the terminology management system should enable the teams to add sen-
tences to specify context, where appropriate.
Given its nature, terminology management is a critical technology for
companies that rely on the precise usage and/or translation of key words.
Certain examples where term usage is very important include product
names, terminology or jargon for any industry (medical, financial, insur-
ance, technology, etc.), and menu items for applications. In regulated in-
dustries, incorrect use of terminology can cost companies millions of
dollars in lost time and additional expenses, while in other industries it can
lead to confusion and potentially loss of trust and/or revenue.
Integrating terminology management with TM is seen within globaliza-
tion as a best practice for maximizing consistency and translation quality
while simultaneously holding down costs.

Machine Translation

Machine translation (MT) is one of those technologies that appear on the

surface to have tremendous potential, but is rarely of significant value in
practice. The concept behind MT is that a computing device should be able
to “read” a sentence and then translate it correctly into another language
based on a complex integration of grammar rules and a cross-language lex-
icon. Were this theory accurate, the result would be that translation could be
done at a rate of hundreds of words per minute for pennies.
Unfortunately, from a technology point of view, language is most often
too complex, nuanced, and illogical for today’s technology to accurately in-
terpret between languages. As a result, MT is typically relegated to situa-
tions in which information is highly technical and very specific in nature,
such as in a technical manual, or is of questionable value, such as say in
“chat” between consumers on a Web site.
Though it is tempting to implement MT systems, experience has shown
that low-cost systems create more problems than they solve, while the costs
associated with implementing “high-value” MT systems can exceed the costs
for human translation. Though this is the situation today, I would not count
MT out entirely, but instead believe that as computing power increases, there
5. MULTICULTURAL CONTENT IN THE GLOBAL ENTERPRISE 153

will be significant technologies that will be produced in the next few years.
These technologies will combine TM, terminology management, and MT to
great benefit of the localization and globalization community.

TRANSLATION SERVICE PROVIDERS

Paradoxically, though the skill of speaking multiple languages is nearly

universally applauded as a significant accomplishment, good translation is
often seriously undervalued by most companies and organizations. Poten-
tially, this seemingly contradictory situation is fostered by the false belief
that being able to speak a second language is sufficient for being able to
translate between two languages.
Actually, good translation is an art that is a capability of only a small per-
centage of bilingual speakers. In addition, there are very few translators
that can be considered experts across two languages; instead, translators
are categorized by their strengths and experience. For example, a particu-
lar translator may be highly specialized in translation between German and
English for the telecommunications industry, whereas another may be an
expert in legal interpretation between French and Spanish.
On the global scene, it is arguable that more wars, misunderstandings,
and conflicts have arisen from poor or incorrect translation exacerbated by
different cultures than from any other source. As a result, translation should
not be taken lightly and should, in most cases, be provided by experts. Keep
in mind that your brand, your image, your messaging, and your revenue in
another market may rely on the quality of your translators. Also, if you have
adopted a model of local content creation in international markets, note
that translation is three times less expensive, on average, than native-lan-
guage authoring.
Translation firms abound in many varieties, from single language pair
specialists to huge full service operations employing thousands of people.
In determining the appropriate translation firm for your needs, keep in
mind four key parameters:

• Translation expertise, including language pairs and specialties (au-

tomotive, medical, telecom, etc.).
• Quality and approaches to maintaining quality.
• Cost per word or per project.
• Turn around time.

To keep your choice honest, keep these benchmarks in mind:

• A typical translator can translate 2,000–2,500 words per day.

• A typical reviewer can review approximately 10,000 words per day.
154 WOODS

• European languages typically cost US$0.20–0.30/per word.

• Asian languages typically cost US$0.30–0.40/per word.
• A web page costs approximately $100–200 to fully localize.

Also of importance, if quality or multicultural issues are of critical impor-

tance, be willing to pay significantly more than the rates quoted here. Note
that there are hundreds of stories of companies spending pennies to trans-
late their brand into another market and losing millions of dollars due to
poor perception, misunderstanding, and confusion.

OVERVIEW AND CONCLUSIONS

Though the complexity of content globalization is often either underrated

(incorrectly believed to be synonymous with poor translation) or overrated
(considered impossible), the implementation of best practices to globaliza-
tion initiatives can enable teams to attain significant return on investment
in short time frames.
By intelligently combining planning, an international mindset, global
teamwork, and technology, the world’s best companies have leveraged their
global operations to yield significant competitive advantage. Though the
road to success is lined with many challenges, persistence and knowledge
can turn content globalization into a cornerstone of your company’s long-
term sustainability and growth.
Armed with the concepts in this chapter, you are now empowered to walk
this road with confidence, avoiding its pitfalls, and attuned to the concepts
that will generate success and advancement.

REFERENCES

DePalma, D. (1999). Strategies for Global Sites. Forrester Research Report, For-
rester Research.
Drakos, N. (2001). Blueprint for Globalization ROI. Gartner Research Report, Gart-
ner Research.
References

INTRODUCTION AND CULTURAL CONSIDERATIONS

Aykin, N. (2001). Legal issues reconsidered: Do the country

boundaries matter any more? In M. J. Smith & G. Salvendy
(Eds.), Systems, social and internationalization design
aspects of human–computer interaction (Vol. 2, pp.
548–552). Mahwah, NJ: Lawrence Erlbaum Associates.

CIO. (2002a). Metrics: PDA Shipments grow 13.8 percent in

Latin America. October 1, 2002. Source: Gartner Dataquest.
Retrieved January 22, 2004, from

CIO. (2002b). Metrics: One billion PCs shipped, July 2,

2002. source: Gartner Dataquest. Retrieved January 22,
2004, from

CIO. (2003). Metrics: Internet population reaches 580

million (source: Nielsen/Netratings), February 27, 2003.
Retrieved January 22, 2004, from

DePalma, D. D. (2002). Business without borders; A

strategic guide to global marketing. New York: Wiley.

Dunlop, B. (1999). Success in any language. March 9, 1999.

Retrieved January 22, 2004, from
http://www.clickz.com/int_comm/article.php/812791

EU Data Protection Directive. (1995). EU Directive

95/46/EC. The Data Protection Di rective. 24 October 1995.
Retrieved January 22, 2004, from http://www.
dataprivacy.ie/6aii.htm

Ellis, R. L. (2001). Web site globalization and the law.

IQPC Seventh International Con ference on Web Site
Globalization, San Francisco, March 25–28, 2001.

Hongladarom, S. (2000). Negotiating the global and the

local: How Thai culture co-opts the Internet. Retrieved
January 22, 2004. First Monday, http://www.first mon
day.dk/issues/issue5_8/hongladarom/

Localisation Industry Standards Association. (2001). The

localization industry primer. Retrieved January 22, 2004,
from http://www.lisa.org/products/primer.html

Microsoft (Dr. International Team). (2002). Developing

international software. Redmond: WA, Microsoft Press.
Morris, S. (2000). The importance of international laws for
web publishers. June 2000. Re trieved January 22, 2004,
from http://www.gigalaw.com/articles/2000/mor
ris-2000-06.html

O’Hara-Devereaux, M., & Johansen, R. (1994). Global work:

Bridging distance, culture and time. San Francisco:
Jossey-Bass.

Sullivan, D. (2003). Search engine sizes. September 2,

2003. Retrieved January 22, 2004, from
http://searchenginewatch.com/reports/article.php/2156481

NUA. (2002). How many online? Retrieved January 22, 2004,

from http://www. nva.ie/surveys/how_many_online/

Schönberger, V. M., & Foster, T. E. (2001). A regulatory

web: Free speech and the global information infrastructure.
Michigan Telecomm. Tech Law Review, 45, University of
Michigan Law School. Retrieved January 22, 2004, from
http://www.mttlr.org/volthree/foster_art/html

Sinrod, E. J. (2000). Upside counsel: Know the laws of

cyberspace. Retrieved January 22, 2004, from
http://listweb.bilkent.edu.tr/anti-spam/2000/nov/
att-0000/01-up side_counsel.htm

Realbeer. (2001). A Bud by any other name …. Retrieved

January 22, 2004, from
http://www.realbeer.com/news/articles/news-001477.php

Rhinesmith, S. H. (1993). A manager’s guide to

globalization: Six keys to success in a changing world.
Alexandria, VA: McGraw Hill. Chapter

Practical Issues and Guidelines

for International Information Display

Nuray Aykin

Siemens Corporate Research, Princeton, NJ

Allen E. Milewski

Monmouth University, West Long Branch, NJ

INTERNATIONAL INFORMATION DISPLAY

Internationalization and localization include the study of

differences

among aspects of locales and how to deal with them.

Differences relevant to

international design span a wide range. Some differences

are abstract and

involve conceptualizations of culture and their

implications for users’ reac

tions to products and services. Others are more pragmatic

and concrete.

There are many locales and many dimensions of differences

among locales.

For the practitioner who is about to tackle an

internationalization project, it

is daunting to consider the sheer size of the knowledge

base required to do a

good and insightful job. This chapter is a guide to a

portion of that knowledge base and is in

tended to be used by the practitioner as a reference during

internationaliza

tion projects. The chapter focuses on pragmatic guidelines

for information

display. These are critical, especially in the context of

web page design and

evaluation. In some design cases, clear guidelines are

possible. In other

cases, the complexities of international design make

straightforward guide 21

lines elusive. In these cases, we resort to providing

useful resources and ex

amples that demonstrate the kinds of problems that can be

confronted. We have found it useful, in tackling
international projects, to keep in

mind the following list of topics and potential problem

areas as the top layer
of internationalization in any cross-cultural design: 1.
Graphics and icons. 2. Language: rendering and translating.
3. data Object formatting. 3. Color. 5. Layout. This, of
course, is a grossly incomplete list of concerns for
international

ization in general (e.g., content, etc.). But, these items

do cover a significant

proportion of the typical internationalization problems

found in designing

web pages (Aykin, 2003). The acronym formed by this list,

GLOCL, also

provides a useful checklist of problem areas that is easy

to remember by re

calling that the area of internationalization is motivated

by a constant ten

sion between what is Global and what is Local.

GRAPHICS AND ICONS

William Horton (chap. 6, this volume) already covers an

extensive array of

useful guidelines on graphics and icons that are motivated

by cultural dif

ferences. These detailed guidelines are not repeated here,

but it is worth

listing several general rules in using graphics in design:

• Keep the design simple. • Leave the text out of graphics,
or use overlaying text on graphics so that when translated,
the icons need not be redone. • Use universally recognized
objects. Examples include using an en velope image for mail
instead of a rural mailbox (which is specific to the United
States) and using an octagonal stop sign instead of a hand
to stop an action. • Implement different icons for each
culture where you cannot find culture-free graphics and
icons. • Be cautious in using images that are potentially
sensitive or poten tially incorrect. For example: • Animals
and People—Use of animals as symbols can be offensive to
some cultures. For example, the pig is considered a filthy
ani mal in Muslim and Jewish cultures, but it is a symbol
of savings in the United States (see coolsavings.com web
site). • Maps—Maps must be drawn based on the latest
information. We are in a world of ever-changing borders. It
is not acceptable, and is considered as an insult, if a map
is not drawn to scale or does not show the current legal
borders. • Directions on how images are read—Not all
languages are read from left to right. The same direction
in reading applies to dis playing images in sequence. In
languages where the text is read from right to left, as in
Arabic and Hebrew, the images are read from right to left
as well.

LANGUAGE: RENDERING AND TRANSLATING

Language is one of the key elements of

internationalization. It is probably

the element with the highest visibility. It affects

everything from how char

acters are displayed to how layouts need to accommodate

potential expan

sion during translation. One of the complexities here is

that languages and

countries, while related, do not form a one-to-one

relationship. First, many

countries can share the same language; for example, English

is spoken in

some 70 countries. Second, some countries are multilingual,

such as Bel

gium (Dutch, French), Switzerland (Italian, French,

German), and Canada

(English, French). Finally, we can see different dialects

(such as Parisian,

Swiss, and Canadian French) and also different writing

scripts (see below) in

the same language. The lack of a one-to-one relationship

between coun

tries, languages, and scripts means that designers must

consider both lan

guage and country as potential determinants of design. The

multitude of

permutations of language, dialect, country, and script has

implications for

at least two key aspects of an internationalization

project: language render

ing and translation.

Language-Rendering Issues An Example: Japanese and Chinese

Some of the difficulties confronted in rendering languages

for user inter

face displays can be shown with examples of two locales:

Japan and China.

Japan has one language, but several scripts. All of these

Japanese scripts can

be used and intermixed in a single sentence. • Kanji is the

complete written Japanese language. It has 50,000 Chi
nese-style characters and can be used for writing a
majority of the language (example below from webjapanese,
2001). • Kana—While Kanji is essentially an ideographic
system, Kana are alphabets of written syllables that
historically derive from Kanji. Since Kana is simpler than
Kanji, they make reading and writing easier. There are two
Kana forms: • Hiragana represents all native Japanese
phonemes and sounds. It is used for articles and endings
(example below from Japanese for Otakus, 2004). • Katakana
is used to represent foreign names and words other than
Chinese or Korean (example below from Japanese for Otakus,
2004). • Romaji (Roman Characters) is a set of Roman or
English characters. They are used for nontranslated words
and numbers. They include: Since characters from all of
these sets can be mixed when writing in Japa

nese, an extremely large number of characters is needed to

represent a typi

cal web page in Japan. But although the large number of

characters needed

to render Japanese web pages is problematic, it is still

more manageable

than accommodating countries with multiple, regional

languages and dia
lects as well as scripts. Chinese, for example, has two
main dialects: • Mandarin, which is used primarily in
Mainland China, and • Cantonese, which is used primarily in
Taiwan. In addition to the dialects, there are two main
scripts:• • Simplified Chinese, which is used in China and
Singapore, and • Traditional Chinese, which is used in
Taiwan and Hong Kong. Character Sets In order to deal with
the processing and rendering of multiple and large

scripts, several character sets have been established and

standardized. A

character set is simply an encoding of some set of language

characters into a

digital, numeric code. The basic character sets commonly

used in software

today are as follows: ASCII. The standard ASCII (American

Standard Code for Informa

tion Interchange) character set consists of 128 characters

ranging from 0

through 127 with the first 32 consisting of nonprinting

characters. The rest

is assigned to letters, numbers, punctuation marks, and

some special char

acters. The extended ASCII character set includes the

standard ASCII plus

characters from 128 through 255 representing additional

special, mathe

matical, graphic, and foreign characters. The ASCII

character set supports only English, Swahili, and Hawaiian

languages. The ISO-8859 Family of Character Sets. Latin-1

(ISO 8859-1) is the

default character set for the Web. However, it is

sufficient only for the Euro

pean languages. It uses 8-bit encoding and permits a

maximum of 256 char

acters. The ISO 8859 Character Set includes

language-specific groupings: ISO 8859-1 [Latin 1] Western
Europe (French, Italian, German, Spanish, …) ISO 8859-2
[Latin 2] Central/Eastern Europe (Hungarian, Polish, Ro
manian, Croatian, …) ISO 8859-3 [Latin 3] Northern Europe
(Esperanto and Maltese) ISO 8859-4 [Latin 4] Southern
Europe (Estonian, Latvian, Lithuanian, Greenlandic, and
Lappish) ISO 8859-5 [Cyrillic] ISO 8859-6 [Arabic] ISO
8859-7 [Greek] ISO 8859-8 [Hebrew] ISO 8859-9 [Latin 5]
Turkish ISO 8859-10 [Latin 6] Nordic (Removed Latvian from
8859-4 and added Icelandic) ISO 8859-11 [Thai] ISO 8859-12
[unassigned] ISO 8859-13 [Latin 7] Baltic Rim (Estonian,
Latvian, Lithuanian) ISO 8859-14 [Latin 8] Celtic (Gaelic
and Welsh) ISO 8859-15 [Latin 9] Latin 0 (Removed little
used characters from 8859-1 and added Euro i) Unicode (ISO
10646 BMP). Unicode is a large character set that in

cludes most of the world’s languages. It uses 16-bit

encoding and allows

65,000 characters. This form is called UCS-2 (Universal

Character Set

2-bytes). Unicode is the first level of ISO 10646, and it

is also called a BMP

(Basic Multilingual Plane) or Plane Zero. More information

on Unicode can

be found at www.unicode.org. UTF-8 (Universal Character Set

Transformation Format). UTF-8 is

an addendum to ISO 10646 and provides compatibility with

ASCII. The

ASCII characters are represented by 1 byte (8 bits) rather

than 4 bytes (32

bits). UTF-8 format has the unique ability to preserve the

full ASCII range.

Consequently, UTF-8 is compatible with file systems,

parsers, and other

ters and do not specify their exact visual appearance. On

the other hand,

fonts are collections of glyphs of a specific style and

appearance. A single

character set can be associated with, or used by, one or

more fonts. Text

must be written in a font that supports the corresponding

character code set

being used. Characters, especially non-Latin characters may

require differ

ent line heights, spacing between lines, spacing between

characters, charac

ter widths, and screen resolution. Some characters, such as

those in

Japanese, may not require ascenders, descenders, or

capitalization. Below

are guidelines related to fonts when dealing with different

characters sets: • Provide enough space between lines to
accommodate changes in line heights, or automatically
adjust line heights depending on the characters used. •
Provide enough interline spacing to ensure clear separation
between lines and to accommodate underlining. • Do not use
overly ornate or decorative fonts because accents and
special characters can become illegible. • Choose fonts
that can accommodate target languages. For example,
Helvetica and Times can provide accents. But Helvetica CE
or Times CE should be used to support the characters in
central European languages. • Do not assume the
availability of standard fonts. • Be prepared to support
mixed (multinational) formats. • Test your fonts for
different target languages with different browsers. • Most
non-Latin languages require proportional spacing. For exam
ple, Arabic cannot be implemented on a monospaced display.
Text Direction Not all languages read from left to right.
There are three types of text

directionality: left-to-right, bidirectional, and vertical.

The most common

is the left-to-right direction that we see in Latin,

Cyrillic, Greek, Thai, and

Indic languages. Chinese, Japanese, and Korean use

left-to-right and verti

cal text. Hebrew and Arabic are bidirectional (Bi-Di) where

Hebrew and

Arabic characters are displayed right to left, and numbers

and Latin charac

ters are displayed left to right. Arabic and Hebrew are

typically read from

right to left; however, words using non-Arabic characters

and Arabic num

bers are read from left to right. The following example is

from Microsoft’s

Saudi Arabia Web site

http://www.microsoft.com/middleeast/saudi/cam

paigns/msdn.asp. The text shows how the bidirectional text

is displayed.

Note that Arabic characters are right justified, whereas

left-to-right direc

tional languages, in non-Arabic text, are left justified.

Order may also be dependent on context. For example,
consider the

string “123-4567” in an Arabic context. If it is a

telephone number, the

reader reads it as: 123-4567. If the string is a

subtraction, the reader reads it

as subtracting 123 from 4567. When designing a page or

screen layout, it is important to consider

whether different text direction may need to be

accommodated in the final

product. The design should be flexible enough to support

this functionality

of text direction. Paper Size Printing onto paper is

another aspect of language rendering. Except in

North America, most countries use the paper sizes specified

by ISO 216,

which is based on the metric system. Therefore, it is

important to allow the

right amount of margin if documents are shared across the

globe; otherwise,
printing could be a problem because printers do not always
supply both U.S.

and ISO 216 types of paper. Although Japan has adopted the
ISO 216 paper

sizes, it still has different standards for the sizes of

folders, defined in the JIS P

0138-61 standard (Kuhn, 2002). Table 2.1 shows the most

commonly used

paper and envelope sizes defined by the United States and

ISO 216. TABLE 2.1 Paper Sizes in the United States and ISO
216

United States ISO 216 In metric units In inches

A (letter) 8½ × 11 in. A4 210 × 297 mm 8.27 × 11.69 in.

Legal 8½ × 14 in.

B (ledger) 11 × 17 in. A3 297 × 420 mm 11.69 × 16.54 in.

Business Card 2 × 3½ A8 53 × 74 mm 2.07 × 2.91 in. Hardware

Concerns A final set of language rendering concerns in
internationalization has to

do with display hardware. There are several key points to

consider when

you are dealing with text entry and display hardware from
different regions

or countries. • Systems must be able to understand and

display the character set used by the keyboard. Make sure
all required characters can be en tered by testing your
system with different locale keyboards. • Different
keyboard layouts may be used in different countries. For
example, many European countries use an AZERTY keyboard in
stead of a QWERTY keyboard. • Each country has specific
power, signaling, labeling, and safety re quirements. •
Make sure the printer has the fonts to display your
coded-charac ter set. • Make sure the printer can handle
different paper sizes (e.g., A4). Or your page cannot be
printed without losing text in the margins due to different
paper sizes.

Language Translation Issues

Language translation is an essential component of
globalization and is one

of the most expensive. One way to keep translation costs

down is to provide

translation tools. For example, Microsoft (2002) defines

its localization kit

as a subset of tools, source files, and binary files that

can be used to create a

localized edition of a program. The localization kit is

generally given to

translators or third-party contractors. Having a

localization kit helps maxi

mize the accuracy of translation and minimize the support

the translators

need from the team. In regard to translation, it is

important to: 1. Identify the content of strings in the
code to be translated. This is done by either including
comments in resource files or providing separate text files
to the translators. 2. Provide clearly written text because
translators will often have diffi culty translating
ambiguous terminology. 3. Provide translation glossaries. A
common glossary will improve consistency across translators
and across product lines. A strong team of technical
writers is important in generating content

ready for translation. The quality of translation is highly

correlated with

how well and clearly the original text is written. Having

the content pre

pared correctly from the beginning saves costs and time

during the transla

tion process, but it requires careful preparation. Even

just making sure that

there are no unnecessary words used in writing would save

money, given

that most translators charge per word. In her book, Hoft

(1995) provides
detailed guidelines for technical writers who develop
content for an inter

national audience. Abbreviations and Acronyms Abbreviations

and acronyms can be a headache for translators. They

should be included in the translator’s glossary so that the

meaning does not

get lost during translation. They can be spelled out,

translated, or left un

changed if they are universal (e.g., ISO, WHO). The full

explanations for

acronyms and abbreviations should be available to the user

in a glossary. Spelling The same language used in different
countries can have different spell

ing rules. For example, neighbor in the United States is

spelled as neighbour

in U.K. English. Similarly:

English (U.S.) English (U.K.)

internationalization internationalisation

program programme

color colour

center centre It is important that the users feel the

language is theirs and not an im

ported version from another country. This means that

content prepared for

one locale should be reviewed for appropriateness before

use in another lo

cale that speaks the same language. Text Expansion When

translating from English to other languages, text can
expand as

much as 30% to 200%. Hoft (1995) states that most

translation companies

recommend that the layout should be designed to accommodate

a 30% text

expansion rate. This 30% rate applies to text that is

presented in para

graphs (over 70 characters). If the text is less than 10

characters long (e.g.,

labels, field names, or icons), the expansion rate can go

as high as 200%,

sometimes reaching 400%. Another common rule is that a

double-byte

character occupies twice as much physical space as a

single-byte character. Some examples of text expansion are
shown below: News (English) <-> Haberler (Turkish) Product
Support (English) <-> Produktunterstützung (German) Search
(English) <-> Búsqueda (Spanish) Undo (English) <->
Ongedaan (Dutch) In order to accommodate possible text
expansion due to translation, the

following guidelines are useful: • Provide flexibility in

the screen/page layout to accommodate differ ent size text
and fonts. • Leave room to accommodate text expansion. Hoft
(1995) suggests a few ways to do this: • Use a larger type
size for the source language and a smaller size for the
translated version. (Be certain, however, that readability
is not sacrificed.) • Increase the width of margins on the
source document when text expansion occurs after
translation. • Consider kerning, leading, and changing the
space between paragraphs. This reduces the space between
lines. • Place labels above the fields when the screen real
estate permits. Placing labels above the entry fields
provides flexibility in case the length of the label
expands after translation. Using wraparound la bels is
another solution if the screen real estate is a problem. •
Use symbols and icons rather than text. When appropriate,
using graphical representation can reduce the cost of
translation and re duce the number of changes to page
layout. • Use tabs to reduce the amount of information on a
page. • Use automatically expanding text objects. •
Consider alternative layout styles. • Allocate enough
memory to allow for text expansion. Sorting Languages also
differ in the ways characters are sorted. The sorting rules

for different languages should be able to handle: •

Accents. • Combinations of characters (e.g., in Spanish,
“cho” comes after “co” because “ch” is treated as a
separate character that comes after “c” in the alphabet;
also, in German the character ‘ß’ is sorted as “ss”). •
Upperand lowercase differences (in many languages uppercase
is sorted after lowercase). • Non-Latin Scripts (e.g.,
Hebrew and Arabic). • Far-Eastern languages (Japanese,
Chinese, Korean). The ISO/IEC 14651:2000 standard titled
“International String Order

ing” provides universal default collation orders for

multiple languages

(available online at
http://www.iso.org/iso/en/cataloguedetailpage.catalog

uedetail?csnumber=250). Writing Practices The following

guidelines, summarized from the GNOME Documenta

tion Style Guide (2003) and Hoft (1995), concern writing

practices that can

DATA OBJECT FORMATTING

Formatting of data objects and fields is another problem

area for interna

tionalization. Here are some issues and guidelines for the

data objects most

often encountered in web site design.

Date and Time

Date formats are sometimes associated with specific

languages or locales.

Although ISO 8601 specifies a standard for date and time

presentation,

many countries keep their own date and time formatting

instead of adopt

ing the international standards. However, this

international standard is a

good way of keeping internal representation of date within

a software appli

cation, and helps to synchronize servers around the world.

The international standard date notation is: YYYY-MM-DD

where YYYY is the four-digit year, MM is the two-digit

month of the year be

tween 01 (January) and 12 (December), and DD is the

two-digit day of the

month between 01 and 31. For example, 1988-01-11 represents

the 11th of

January in 1988. This date can be seen in various notations

in around the

world: 1/11/88, 11/1/88, 88/11/1, 11 January 1988, January

11 1988, and so

on. The delimiter to separate the year, month, and day can
be a forward

slash (/), space, hyphen (-), or period (.). Many software

platforms, tools, and languages such as Java and .NET

provide how-to’s and resources for presentation of dates

and times in differ

ent languages. The differences in date formatting can be

observed in the long or short date

representations. Dates are most commonly displayed as day,

month, and year.

In some countries, the month may be displayed before the

day and the year; in

other countries, the year may be displayed before the month

and the day. Here are the proper formats for some key
locales: MM DD YYYY - United States DD MM YYYY - Most
European countries, Australia, Canada YYYY MM DD - South
Africa, Japan, Korea, China, Taiwan Table 2.2 shows long
and short date formats for different countries. In long
date formatting, the month is usually spelled out. Using a

three-letter abbreviation for month may work well for

English-speaking

countries, but can cause problems in other languages. For

example, the

months June and July are Juin and Juillet in French. If the
three-letter ab

breviation is forced, then both months could be abbreviated

as Jui. The following are guidelines on how to treat date
formatting for differ

ent locales: • Use the correct date representation for the

chosen locale. • Make use of the properties of the tools,
platforms, and languages in providing support for date and
time presentations. • If you must choose only one date
format for display for use around the world, use dd Month
yyyy (e.g., 20 August 2000). Even if it is not formatted
correctly for a given locale, the content is unambiguous. •
Make provisions to accommodate different delimiters for
date, month, and year. Include a forward slash (/), period
(.), space, or hy phen (-) as options.

36 TABLE 2.2 Long and Short Date and Time Formats for
Monday April 14, 2003 in Different Countries • Do not
assume a three-character space for month and day
abbreviations because this may not work in some languages.

Time

As ISO 8601 specifies, the international standard notation

for the time of

day is hh:mm:ss

where hh represents the hours (00–24), mm represents the

minutes

(00–59), and ss represents the seconds (00–60). Many

countries use the 24-hour time notation (e.g., 15:30)
rather than

12-hour AM and 12-hour PM notation (e.g., 3:30 PM). In the

United States,

the 24-hour notation is referred to as “military time.” The

delimiters to sep

arate hours, minutes, and seconds can be a colon (:) or a

period (.). Table 2.2

also includes the time formatting for different countries.

ISO 8601 has been adopted as European Standard EN 28601 and
is

therefore now a valid standard in all E.U. countries, and

all conflicting na

tional standards have been changed accordingly (Kuhn,

2001). TABLE 2.2 (continued)

Calendar/Holidays/Start of Week

There are a few aspects regarding calendars that we need to

consider when

designing for the global markets. • Calendar Issues:

Although the Gregorian calendar is used in al most every
country, there are still cases in which a different calen
dar might be considered. These include: • Arabic calendar
(Note: In Islamic countries, the date changes at sunset,
not midnight) • Jewish calendar • Iranian calendar •
Japanese Imperial calendar • Holidays: It is important to
know the holidays of the target country, especially if you
are doing business with them. Do not assume that everyone
celebrates Christmas. Many non-Christian countries do not
have a Christmas holiday. Even new year celebrations may
vary. The Chinese New Year does not coincide with the new
year associated with the Gregorian calendar. • Start of the
week (Monday vs. Sunday): In the United States, the week
begins on Sunday. In most other parts of the world it
begins on Monday. For applications that deal with
scheduling, this can be an issue. The user interface
showing weeks for scheduling may look different for the
U.S. users than for the European users.

Numeric Formatting

Number formatting differs from country to country. The

elements of nu

meric formatting includes: • Thousands and decimal

separator: Separators can be a character or a space. Table
2.3 shows different separators for numeric repre sentation.
• Number of digits between separators: The number of digits
between separators can vary from country to country. For
example, 123,456,789.00 in the U.S. can be shown as
12,34,56,789.00 in Hindi. • Negative numbers: Negative
numbers may be indicated by using a minus sign before or
after the number, or by enclosing the number in parentheses
or brackets. -123 123(123) [123] 39 TABLE 2.3 Numeric
Formatting and Currency Representation for Different
Countries (continued on next page)

Names and Addresses

In the world of e-business, there are many formats for

address forms. In a
typical correspondence, an address consists of • Title •
First/given name • Last Name/Family Name/Surname • Street
number, building number, street name • City or town •
State/province/region • Country • Zip or postal code (can
include alphanumeric characters) Labels given for each of
the these address entries may vary from coun

try to country. Here are sample address formats for Mexico

and the

United States. Mexico Name (Paternal Name, Maternal Name,

First Name) Street and Number Building, Floor, Suite #
Colony City, State Postal Code TABLE 2.3 (continued) United
States Name (First Name, Middle Initial, Last Name) Number
and Street City, State Zip Code 1 Differences in address
formatting could be seen in the labels of the name

fields. For example, First Name is mainly used in the

United States, whereas

most of the world uses Given Name. Last Name in the United
States becomes a

Family Name or Surname in other countries. And having

multiple names or

last names adds complexity to the problem. For example,

people in Mexico

have many names, including two last names. Titles are

commonly used in addresses outside the United States. Even
in

personal letters, the use of a title shows a level of

respect for the addressee.

In addition to Mr./Mrs./Ms., the titles Dr. and Professor

are extremely im

portant in some cultures. For example, the corporate

directory for AT&T, a

U.S.-based company, includes names and job titles. The

corporate direc

tory for Siemens, a Germany-based company, includes names,

job titles,

and professional titles. The other fields in the address

format include the street, building, town,
city, country, zip/postal code, and so on. The formats for
these fields vary

from country to country. The differences could be in the

number of address

lines, order of address lines, number of names (first and

middle initial and

last/given/surnames), street addresses, level of hierarchy

in terms of town,

city, state/province/region, and country, and of zip/postal

codes. For each of

these fields, there should be no assumption on format and

length. For ex

ample, in the United States the zip code consists of only

numeric values;

however, in many parts of the world it might include

alphanumeric entries

and might be longer than a 5-digit zip code. However, there

are many attempts to generate a generic address form so

that there is no need to create a different address form

for each country.

This is especially important for online

registration/shopping/ordering

practices. Many times, it is impossible to identify which

address format to

use for each request. Therefore, it becomes very meaningful

to provide a

generic address format which spans multiple locales. Figure

2.1 shows a good example (from www.amazon.com) of a ge

neric address format that can handle many of the different

address for

mats in the world. 1 For more examples of country

addresses, see Microsoft (2002), which provides different

address formats for a set of selected countries.

Telephone Numbers

Like other formatted fields, the format for telephone

numbers varies from

one country to another. The differences between telephone

numbers are: • The total number of digits (varies from six
to fifteen) • Separators (e.g., hyphens (-), periods (.),
parentheses (( )), and spaces) • Groupings of numbers (two
to six digits per group) • Country codes (one to three
digits) • Long distance access code to access national or
international calls (could be 00, 011, 001, 0, etc.) • Code
to access an outside line on a business phone (in the
United States, “9” is commonly used to place a call outside
the office) • Extensions (some phone numbers include
extensions that should be taken into consideration during
design). Table 2.4 provides a few examples of different
telephone number for

matting: In addition to different formats in telephone

numbering, in some coun

tries, letters are not used on the keypad. This can make it
difficult to dial

numbers that are given as mnemonics, as in 1-800-WEATHER.

Radio Shack

used to sell templates for travelers to use on telephones

without letters. ITU-T Recommendation E.164 (The
International Public Telecommuni

cation Numbering Plan; 1997), defined by the International

Telecommuni

cations Union (ITU), formerly Comité Consultatif

International de TeleFIG. 2.1. Sample universal address
format.

graphique et Telephonique (CCITT), details the format and

structure of

the Public Switched Telephone Network (PSTN) numbering

plan. Within

this Recommendation, it was proposed that all switches and

networks

should have the capability of handling international

telephone numbers

with a maximum length of 15 digits. However, given that the

separators

(e.g., spaces or parentheses) can end up being part of the

telephone num

ber entry, the length of the telephone numbers may exceed

15 digits. It is

important not to limit the digits in a phone number entry

in fields and also

not to parse the telephone numbers into area codes and

phone numbers as

there may not be an area code, but only a city code in some
countries. Users

should also be able to enter telephone numbers in free

format (can use (, ), -,

/). Consider these notations in determining field sizes.

Currency

There are many different currencies in the world. Table 2.3

shows a few ex

amples of currency representation. Although some currencies

have special

symbols (such as U.S. dollar as $, Japanese Yen as ¥,

British Pound as £),

many use abbreviations of the currency name (such as TL for

Turkish Lira).

ISO 4217 (Codes for the Representation of Currencies and

Funds) defines

three-letter abbreviations for world currencies. The

general principle used to construct these abbreviations is
to take the

two-letter country abbreviations defined in ISO 3166 (Codes

for the Repre

sentation of Names of Countries) and append the first

letter of the currency

name (e.g., GBP for the Great Britain Pound, and USD for
the United

States Dollar). In the case of currencies defined by

supra-national entities, TABLE 2.4 Sample Telephone Number
Formats

Australia 649-800-445-768

Austria 1234 56 78 90

Belgium 12-345 67 89

Denmark 12 34 56 78

Germany (123) 4 56 78 90

Italy (12) 3456789

Portugal 123-456 78 90

Japan 123-45-6-789-0000

United States (123) 456-7890

ISO_4217 assigns two-letter entity codes starting with “X”

to use in place of

country codes (e.g., XCD for the Central Caribbean Dollar;

Allen & Hall,

2003). Similarly, because there is no specific country

associated with Euro,

the code for Euro is EUR, with a well-recognized symbol, i.

It is common practice to use the symbols for well-known
currencies (such

as $ or £) when monetary amounts are displayed for an

international audi

ence. For local audiences, it is again a practice to use

the currency symbol

since the users are familiar with their own currency

symbol. However, it is

advisable to use the three-letter code if the audience is

international and the

symbol is not well recognized worldwide.

Monetary Values

The format for representing monetary value differs from

country to coun

try. Most locales use their numeric formatting rules in

terms of a decimal

and thousands separator. But this is not always the case.

For example, in Es

tonia, the period is used as the decimal separator for

Estonian Kroons

(123.45 kr), but the comma is used as the decimal separator

for other nu

meric representations (123,45; Microsoft, 2002). Placement

of Currency Symbol and Negative Sign The placement of the
currency symbol and the negative sign also differs

from country to country. Some countries, such as United

States, use paren

theses to indicate negative monetary value.

Sizes and Measurements

The United States uses the imperial system for the

measurements, whereas

most countries use the metric system. For computer

applications, this is es

pecially important for files that rely on measurements

(e.g., page size, icons,

clip art) that are shared across regions. Sizes and

measurements that may

change from country to country include: • Typographic units

(point, pixel, measurement units). • Temperature
measurements (Fahrenheit or Celsius) U.K. –£ 1,324.55 U.S.
($1,324.55) Netherlands 1,324.55 i – France –1,324.55 i •
Paper and envelope sizes (see page 28) • Clothing sizes
Metric Versus Imperial Units
The Imperial system uses the foot, the pound, and the
Fahrenheit (Fº) scale

to measure length, mass (weight), and temperature,

respectively. The met

ric system uses the meter, the kilogram, and the Celsius
(Cº) scale. Most of the world uses the metric system,
except for Burundi, Liberia,

Yemen, Rwandese Republic, Union of Myanmar, and the United

States

of America. Typographic Units Vekulenko (2000) and Kuhn

(1999) summarize the differences between the

typography units used around the world. To name a few

widely used ones: • 1 point (Didot) = 0.3759 mm = 1/72 of a
French Royal inch (27.07 mm) = about 1/68 inch • 1 point
(ATA) = 0.3514598 mm = 0.0138366 inch • 1 point (TeX) =
0.3514598035 mm • 1 point (Postscript) = 0.3527777778 mm =
1/72 inch • 1 point (l’Imprimerie nationale, IN) = 0.4 mm
When sharing images between different locales, be certain
to use the

same point system to avoid design mistakes. Temperature

Before the 1970s, most English-speaking countries used the

Fahrenheit

scale. Since then, most countries adopted the Celsius

temperature mea

surement scale, except in the United States, which uses the

Fahrenheit

scale. Many travel sites now show temperatures in both

scales to help

travelers. Clothing Size

It is especially important for online merchandise catalogs

to show metric

and imperial shoe and clothing sizes. Even if the target

market is the United

States, there are many people in the United States from

different countries
who will be pleased to see the sizes given in the system
that they are used to.

Many travel agencies’ sites, such as American Express,

provide clothing size

charts to show the conversions for different countries. A

useful bit of infor

mation for travelers! Table 2.5 shows the conversion for

clothing sizes.

COLOR

Color has very different meanings around the world and

therefore it is a sig

nificant element, not just for computer user interfaces,

but for all forms of

international interaction. Table 2.6 shows different

meanings of colors for

various cultures (adapted from HP, 2003; Morton, 2003;

Prime & Wilson,

2002; Jinsheng, 2001). For example, Holzschlag (1999) and

Morton (2003)

point out that the color purple represents death in

Catholic parts of Europe

and that Euro Disney should have done their homework well
before choos

ing purple as a signature color. TABLE 2.5 Clothing Sizes

for the United States, Europe, and the United Kingdom
Women’s Dress Sizes Women’s Shoe Sizes

United States Europe United Kingdom United States Europe

United Kingdom 6 36 8 5 35 3.5 8 38 10 5.5 36 4.0 10 40 12
6.0 36 4.5 12 42 14 6.5 37 5.0 14 44 16 7 37 5.5 16 46 18
7.5 38 6.0 18 48 20 8.0 38 6.5 20 50 22 8.5 39 7.0 22 52 24
9.0 39 7.5 Men’s Suit Sizes Men’s Shoe Sizes

United States Europe United Kingdom United States Europe

United Kingdom 36 46 36 8 41 7 38 48 38 8.5 42 7.5 40 50 40
9.5 43 8.5 42 52 42 10.5 44 9.5 44 54 44 11.5 45 10.5 46 56
46 12.0 46 11 47 TABLE 2.6 Color Meanings in Various
Cultures

Color Meaning for Different Cultures

Red China: Good luck, prosperity, happiness, marriage

Japan: Anger, danger Middle East: Danger, Evil India:
Purity, life Egypt: Death South Africa:Mourning USA and
Western Europe: Danger, anger, stop

Blue China: Immortality, heavens, depth, cleanliness Japan:

Villainy India: Color of Krishna USA and Western Europe:
Masculinity, calm, authority, peace

Yellow East Asia: Sacred, Imperial, Royalty, Honor Middle

East: Happiness, prosperity India: Religious color
(celebration),success Egypt: Mourning USA and Western
Europe: caution, cowardice

Green Asia: Family, harmony, health, peace, life, youth,

energy, growth Muslim countries: Religious color,
fertility, strength Ireland: National color USA and Western
Europe: Safe, go

White Asia: Mourning, death, purity Middle East: Mourning,

purity India: Death, purity USA and Western Europe: Purity,
virtue

Black Asia: Evil influences, knowledge, mourning Middle

East: Mystery, evil USA and Western Europe: Mourning,
elegance (black dress and tuxedo), death, evil

Brown India: Mourning

Purple Thailand: Mourning Catholics: Death Asia: Wealth,

glamour Western Europe: Royalty

Orange Ireland: Religious significance China: Light and

warmth USA: Inexpensive goods There are numerous excellent
resources that deal with color usage. Jill

Morton’s colormatters.com web site has extensive

information on color

studies and online books on color. Spartan (Multimedia

Produc

tions,1999) gives many examples related to different

cultures in their as

sociation with colors. Here are a few examples from

Spartan’s colorful

multimedia presentation: • In the United States, Americans

tend to relate blue to authority, whereas in Canada, red is
authority, as well as an inspirer of nation alism. In the
British Isles, black is commonly used for authority fig
ures. The use of red and black in Britain is usually
reserved for honorary positions, such as the Royal Guard. •
White is a very common color in hot Mediterranean
countries, such as Greece. The use of white dates back to
ancient Greece when all buildings, sculptures, and so on
were required to be clean white. • In China, red and gold
are used for joyful, festive occasions. They are also
commonly used for religious shrines and temples, as well as
other forms of architecture. • Orange is widely used by
Indians in their attire and as a means of attention. It
also has religious significance in Hindi and Buddhism. •
The Japanese spectral range of colors is mainly muted,
pastel types, with the exception of bright red. • In Korea,
light blue and other pastel colors in clothing are usually
reserved for the wealthy and for festivals. Commoners
usually wear white or grays. • In many African cultures,
black is a symbol of strength, and white is a symbol of
fertility. • Native Indians of Mexico used bright colors
for their ceremonial costumes. Possibly stemming from the
Indian cultures, the colors of Mexico tend to be bright and
festive greens to reds or a pastel ver sion of those
colors. With world boundaries blurring, and the cultures
blending more than

ing less apparent. We now, for example, see many brides

wearing tradi

tional white dresses in all Asian countries. Although the

strict meanings of colors may be fading across countries,
it

is still a good practice to use colors carefully. We do not

see brightly colored

web sites or consumer products for the Japanese culture;

instead, products

are usually designed with pastel colors in mind. If the

same color palette is to be used across locales, it would
be a good

idea not to use primary colors in design because they may

still carry nega
tive or positive meanings in some cultures.

LAYOUT

Our final internationalization design consideration is

layout. It is impor

tant not to assume anything about layout because all layout

elements vary

according to language (W3C, 2003). These elements include

how a piece

of text might be drawn on the screen, text alignment and

justification,

how much room translated text takes up, the direction it

flows, line

breaks, text wrapping, white spaces, word and letter

spacing, text deco

rating (underline, overline, line-thru, text shadows, text

underline solu

tion for East Asian vertical writing and blinking), and

where on the

screen it should start. Make sure there is enough room for

text expansion. Layout has to accom

modate different text properties and the change in context

size due to trans

lation. Therefore, it is important not to hardcode physical

layout space, but

to let it adjust dynamically. Use of dynamic layout

managers is preferred over

fixed-position layout because of the benefit of

cross-platform automatic lay

out management for graphical user interface components.

SUMMARY

The purpose of this chapter was to list, in one place, a

good sampling of the
issues, guidelines, resources, and tips associated with
international infor

mation display. This chapter should make it possible for

the user interface

practitioner to begin an internationalization project in a

knowledgeable

and productive way. The list of issues described in this

chapter is by no

means a complete list. But these issues will give

cross-cultural designers a

foundation on which experience can build a fuller

understanding of life

Allen, P. L., & Hall, J. (2003). World currencies and

abbreviations. Retrieved January 22, 2004, from
http://www.jhall.demon.co.uk/currency/

Aykin, N. (2003). Internationalization and localization of

Web sites. Tutorial presented at the User Experience 2003
Conference.

GNOME Documentation Style Guide. (2003). Chapter on writing

for localization. Pub lished by: Sun Microsystems.
Retrieved January 22, 2004, from http://devel
oper.gnome.org/documents/style-guide/

Hewlett-Packard. (2003). Printing with color—the meaning of

color. Retrieved January 22, 2004, from
http://www.hp.com/sbso/productivity/color/meaning.html

Hoft, N. L. (1995). International technical communication.

New York: Wiley.

Holzschlag, M. E. (1999). Color my world. New architect:

internet strategies for technology leaders. Retrieved
January 22, 2002, from http://www.webtechniques.com/ar
chives/2000/09/desi/

Hongladarom, S. (2000). Negotiating the global and the

local: How Thai culture co-opts the Internet. Retrieved
January 22, 2004, from http://www.firstmonday.dk/is
sues/issue5_8/hongladarom/

ITU-T Recommendation E.164. (1997). The International

Public Telecommunication Numbering Plan. International
Telecommunications Union. Telecommunication Standardization
Bureau. Geneva, Switzerland. Retrieved January 22, 2004,
from http://www.itu.int/ITU-T/index.html

Japanese for Otakus. (2004). Retrieved January 22, 2004,

from http://www.orange an gel.com/otaku/april11.html

Jinsheng, M. (2004). Colors and meanings in Chinese modern

products. Retrieved January 22, 2004, from

Kuhn, M. (1999). Metric typographic units. Retrieved

January 22, 2004, from
http://www.cl.cam.ac.uk/~mgk25/metric-type/html

Kuhn, M. (2001). A summary of the international date and

time notation. Retrieved Janu ary 22, 2004, from
http://www.cl.cam.ac.uk/~mgk25/iso-time.html.

Kuhn, M. (2003). International standards paper size.

Retrieved January 22, 2004, from
http://www.cl.cam.ac.uk/~mgk25/iso-paper.html.

Microsoft (Dr. International Team). (2002). Developing

international software. Redmond, WA: Microsoft Press.

Morton, J.L. (2003).Colormatters.RetrievedJune10,2003,

fromwww.colormattters.com

Morton, J. L. (2003). Global color meanings. Retrieved

January 22, 2004, from
http://www.colormatters.com/bubarc8-global.html

Prime, M., & Wilson, M. (2002). World Wide Web Consortium

(W3C) primer on the internationalisation and localisation
of web pages: W3C UK and Ireland Office 1 December 2002.
Retrieved January 22, 2004, from
http://www.w3c.rl.ac.uk/QH/
QH_final_review_WP5/WD-int-primer.html

Russo, P., & Boor, S. (1993). How fluent is your interface?

Designing for interna tional users. INTERCHI’93. 342–347.

Spartan (1999). Multimedia presentation on the colors of

culture. Retrieved January 22, 2004, from
http://www.mastep.sjsu.edu/Alquist/workshop2/color_and_cul
ture_files/frame.htm

Vekulenko, A. (2000). Differences between point systems.

Retrieved January 22, 2004, from
http://www.oberonplace.com/dtp/fonts/point.htm
Webjapanese.com. (2001). Retrieved June 10, 2003, from
http://webjapanese.com/ wj/ftp/dtp/images/kanji01_640_w.gif

Wired4success. (2003). Color symbolism chart. Retrieved

January 22, 2004, from
http://www.wired4success.com/colorsymbolism.htm Chapter

User Interface Design and Culture

Aaron Marcus

Aaron Marcus and Associates, Berkeley, CA

INTRODUCTION

Over the past two and a half decades in the user interface
(UI) design com

munity, designers, analysts, educators, and theorists have

identified and

defined a somewhat stable set of UI components, that is,

the essential enti

ties and attributes of all UIs, no matter what the platform

of hardware and

software (including operating systems and networks), user

groups, and con

tents (including vertical markets for products and

services). That means

these components can enable developers, researchers, and

critics to com

pare and contrast UIs that are evidenced on terminals,

workstations, desk

top computers, Web sites, Web-based applications,

information appliances,

and mobile, wireless devices. (Marcus, 1998, 2001), among

others, provides

one way to describe these UI components, which is strongly

oriented to

communication theory and to applied theory of semiotics

(Eco, 1976;

Peirce, 1933). This philosophical perspective emphasizes

communication

as a fundamental characteristic of computing, one that

includes perceptual,

formal characteristics, and dynamic, behavioral aspects of

how people in

teract through computer-mediated media. Picking up on

Claude Lévi

Strauss’s (2000) theory of human beings as sign makers and

tool makers, my

theory understands a UI as a form of dynamic, interactive

visual literature

as well as a suite of conceptual tools, and as such, a

cultural artifact. The UI

components are the following: 51

Metaphors

Metaphors are fundamental (Lakoff and Johnson, 1980)

concepts commu

nicated via words, images, sounds, and tactile experiences.

Metaphors sub

stitute for computer-related elements and help users

understand,

remember, and enjoy entities and relationships of

computer-based com

munication systems. Metaphors can be overarching, or they

can communi

cate specific aspects of UIs. An example of an overarching

metaphor is the

desktop metaphor to substitute for the computer’s operating

system, func

tions, and data. Examples of specific concepts are the

trashcan, windows
and their controls, pages, shopping carts, chatrooms, and
blogs (Web logs).

The pace of metaphor invention, including neologisms or

verbal metaphor

invention, is likely to increase because of rapid

development and distribu

tion through the Web and mobile devices of ever-changing

products and

services. Some researchers are predicting the end of the

desktop metaphor

era and the emergence of new fundamental metaphors

(Gelerntner, 2000).

Mental Models

Mental models are structures or organizations of data,

functions, tasks,

roles, and people in groups at work or play. These are

sometimes also called

user models, cognitive models, and task models. Content,

function, media,

tool, role, goal, and task hierarchies are examples.

Navigation

Navigation involves movement through the mental models,

that is, through

content and tools. Examples of UI elements that facilitate

such movement

include those that enable dialogue, such as menus, windows,

dialogue

boxes, control panels, icons, and tool palettes.

Interaction

Interaction includes input/output techniques, status

displays, and other
feedback. Examples include the detailed behavior
characteristics of key

boards, mice, pens, or microphones for input; the choices

of visual display

screens, loudspeakers, or headsets for output; and the use

of drag-and

drop selection/action sequences.

Appearance

Appearance includes all essential perceptual attributes,

that is, visual, audi

tory, and tactile characteristics,. Examples include

choices of colors, fonts,

animation style, verbal style (e.g., verbose/terse or

informal/formal), sound

cues, and vibration modes. As computer-based communication

has become a global enterprise,

the development, marketing, and business communities have

become

aware of the impact of world cultures on UI design for

global products and

services. A clear need has emerged to determine optimum

characteristics

of suitably localized products and services based on market

and user data

to achieve both short-term and long-term success, without

having to de

velop too many variations that might waste time and money
in develop

ment, distribution, and maintenance. Likewise, over many

more decades, cultural anthropologists have identi

fied and defined somewhat stable, fundamental analytical

dimensions by

which one may contrast and compare all world cultures.

Culture, as evi

denced at work, at home, in schools, and in families, may

be defined in terms

of a group’s symbols, heroes, rituals, and values. In this

chapter, culture re

fers, in general, to well-established groups whose

characteristics a member

may have significant difficulty to abandon, but may also

include what might

be called “temporary, voluntary life-style affinity groups”

not recognized as

cultures by some classical cultural anthropologists

(Clausen, 2000). For example, consider the international
express-mail Web site page of

DHL for shipping packages to Saudi Arabia shown in Fig.

3.1. Is this imagery

appropriate for potential customers? This chapter seeks to

provide initial re

sources for considering such questions. The issue is potent

if one considers

what DHL and FedEx, another international shipping company,

were show

ing at the same point in time. Both DHL and FedEx had
similar layouts, but

the FedEx site showed several standard images, including an

Asian woman

with bare sleeves and uncovered face. This race, gender,

and dress style is not

typical for Saudi Arabia, while the DHL site showed an

apparently Middle

Eastern man in conservative Western business attire. The

DHL page also

showed a green logotype in Arabic script at the lower left,

a typical reference
via color and script to Saudi Arabian nationality and the
Moslem religion. As an example of a culture model, one may
cite the work of the cultural

analyst Geert Hofstede (Hofstede, 1997). He studied

hundreds of IBM em

ployees in 53 countries from 1978 to 1983 (see Table 3.1

for a display of his

findings), from which study he developed five fundamental

culture dimen

sions. These dimensions of culture are described in the

following sections.

Power Distance

Power distance measures the extent to which people of a

culture accept large

or small distances of power in social hierarchies. For

example, if an employee

of a large organization within a culture typically has

easy, informal access to

the “boss,” then one might assign a low power distance

rating to that culture.

Individualism Versus Collectivism

Individualism versus collectivism measures the orientation

to individual or

group achievements. If a culture typically focuses on group

goals and ori

ents promotion of achievements to the group, rather than to

individuals of

the group, then one might assign a high collectivist rating

to that culture.

Masculinity Versus Femininity

Masculinity versus femininity measures the degree to which

a culture does or
does not separate traditional gender roles. Traditional
gender roles assume

that males are tough, task-oriented, individualor

group-oriented “warriors”

going out in the world to conquer and defend, and females

are tender, gentle

homemakers and childrearers who are familyand

people-oriented. Mascu

line cultures tend to support this traditional role

separation; feminine cul

tures tend to diminish the differences, that is, to merge

both roles.

Uncertainty Avoidance

Uncertainty avoidance measures the degree to which a

culture is uncom

fortable with uncertainty and seeks to reduce uncertainty,

often in the pur FIG. 3.1. A DHL express-mail Web site page
for Saudi Arabia, as displayed in 2000. Note that the Web
site has been updated significantly since then. 55 TABLE
3.1 Hofstede’s Cultural Indexes PDI IDV MAS UAI LTO Rank
Score Rank Score Rank Score Rank Score Rank Score

Arab Countries 7 80 26/27 38 23 53 27 68

Argentina 35/36 49 22/23 46 20/21 56 10/15 86

Australia 41 36 2 90 16 61 37 51 15 31

Austria 53 11 18 55 2 79 24/25 70

Bangladesh 11 40

Belgium 20 65 8 75 22 54 5/6 94

Brazil 14 69 26/27 38 27 49 21/22 76 6 65

Canada 39 39 4/5 80 24 52 41/42 48 20 23

Chile 24/25 63 38 23 46 28 10/15 86

China 1 118
Columbia 17 67 49 13 11/12 64 20 80

Costa Rica 42/44 35 46 15 48/49 21 10/15 86

Denmark 51 18 9 74 50 16 51 23

East Africa 21/23 64 33/35 27 39 41 36 52

Equador 8/9 78 52 8 13/14 63 28 67

Finland 46 33 17 63 47 26 31/32 59

France 15/16 68 10/11 71 35/36 43 10/15 86

Germany FR 42/44 35 15 67 9/10 66 29 65 14 31

Great Britain 42/44 35 3 89 9/10 66 47/48 35 18 25

Greece 27/28 60 30 35 18/19 57 1 112

Guatemala 2/3 95 53 6 43 37 3 101

Hong Kong 15/16 68 37 25 18/19 57 49/50 29 2 96

India 10/11 77 21 48 20/21 56 45 40 7 61

Indonesia 8/9 78 47/48 14 30/31 46 41/42 48

Iran 29/30 58 24 41 35/36 43 31/32 59

Ireland (Rep of) 49 28 12 70 7/8 68 47/48 35

Israel 52 13 19 54 29 47 19 81

Italy 34 50 7 76 4/5 70 23 75 (continued on next page)

56 TABLE 3.1 (continued) PDI IDV MAS UAI LTO Rank Score
Rank Score Rank Score Rank Score Rank Score

Jamaica 37 45 25 39 7/8 68 52 13

Japan 33 54 22/23 46 1 95 7 92 4 80

Malaysia 1 104 36 26 25/26 50 46 36

Mexico 5/6 81 32 30 6 69 18 82

Netherlands 40 38 4/5 80 51 14 35 53

Philippines 4 94 31 32 11/12 64 44 44 21 19
Poland 13 32

Portugal 24/25 63 33/35 27 45 31 2 104

Salvador 18/19 66 42 19 40 40 5/6 94

Singapore 13 74 39/41 20 28 48 53 8 9 48

South Africa 35/36 49 16 65 13/14 63 39/40 49

South Korea 27/28 60 43 18 41 39 16/17 85 5 75

Spain 31 57 20 51 37/38 42 10/15 86

Sweden

Sweden 47/48 31 10/11 71 53 5 49/50 29 12 33

Switzerland 45 34 14 68 4/5 70 33 58

Taiwan 29/30 58 44 17 32/33 45 26 69 3 87

Thailand 21/23 64 39/41 20 44 34 30 64 8 56

Turkey 18/19 66 28 37 32/3 45 16/17 85

Uruguay 26 61 29 36 42 38 4 100

USA 38 40 1 91 15 62 43 46 17 29

Venezuela 5/6 81 50 12 3 73 21/22 76

West Africa 10/11 77 39/41 20 30/31 46 34 54

Yugoslavia 12 76 33/35 27 48/49 21 8 88

Zimbabwe 19 25

Note. PDI = power distance index; IDV = individualism

index; MAS = masculinity index; UAI

= uncertainty avoidance index; LTO = long-term orientation

index. Adapted from Cultures and Or

ganizations: Software of the Mind by G. Hofstede, 1997, pp.

26, 55, 84, 113, 166. Original data copy

right 1997 by Geert Hofstede.

suit of the truth. Cultures that focus on punctuality,
formality, and explicit

communication tend to rate high in uncertainty avoidance.

Long-Term Time Orientation

Long-term time orientation grows out of a long-term basis

of some cul

tures in Confucian thought, which emphasizes patience.

Naturally, this di

mension seems strongest among Asian cultures steeped in

Confucian

philosophy. (For an example of communication influences and

differ

ences, see Oliver, 1971.) An example of the interrelations

of two cultural dimensions, power distance

versus individualism-collectivism, for several countries

appears in Figure 3.2. De Mooij (2001) has argued for
culture differences as a basis for differ

ences of Internet and mobile behavior separate from

economic factors.

These culture dimensions, examples of culture bias in Web

site designs, and

the implications of culture dimensions on Web site and

Web-based applica

tion designs are discussed at length elsewhere (Marcus,

2001; Marcus &

Gould, 2000). FIG. 3.2. Chart of power distance versus

individualism-collectivism for selected countries. Adapted
from Cultures and Organizations: Software of the Mind by G.
Hofstede, 1997, p. 54. Copyright 1997 by McGraw-Hill.
Adapted with permission of McGraw-Hill.

MAPPING DIMENSIONS TO COMPONENTS

Although these definitions of dimensions and components are

not univer

sally accepted, they are stable enough to consider a next

phase of under

standing the relationship of culture to UIs: mapping

components of UIs to

dimensions of culture, and vice versa. Being able to

understand the relation

ships of UI components to a particular culture dimension is

a step forward to

mapping particular UI designs to one or more cultures.

Reliably being able

to understand and predict the use, acceptance, and

enjoyment of particular

UIs marks significant progress in developing consistently

successful products

and services. In addition, with greater understanding of

the relationships, it

becomes more feasible to develop databases, such as

databases of culture

phenomena or catalogs of UI design patterns (Tidwell, 1999)

in which cul

ture dimensions play a more significant role in the

attributes and tools. Such tools might enable designers to
eventually “semiautomatically” ad

just UI designs per culture dimension, to sets of countries

with similar-cul

ture characteristics (e.g., adjusting UI designs to be

suitable for all countries

that were once part of the British Empire), and ultimately,

per culture. These

“culture-design” tools someday may be ubiquitous and

conventional, similar

to the color adjustment tools today of a graphics product

like Adobe

Photoshop™, which permits a designer to choose color sets

that are safe for
the Web. A future scenario might find designers (from one
or more cultures)

designing a UI for a specific culture, then asking the

computer to adjust the

entire UI design for another culture, say Culture X, to

make it “Culture

X-safe.” This design might be analyzed by the designers or

submitted to users

of the target culture for feedback. In an interactive

process, the design might

be adjusted, including the adjustment of culture parameters

to reflect ongo

ing changes in the target cultures, which are not

themselves static. To explore possible relations of culture
dimensions to UI components, I,

sometimes in conjunction with Emilie Gould of Rensselaer

Polytechnic In

stitute (goulde@rpi.edu), have been giving tutorials about

cross-cultural

communication since 1999 in several countries, including

North America,

Europe, South Africa, and Japan. In the tutorials, we ask

participants to ini

tiate analysis and discussion of the relation of culture

dimensions to UI

components. The conceptual challenge is to fill in a matrix

like that of Fig.

3.3. To stimulate further study, an initial mapping appears

here. This map

ping, in greater detail, with the appropriate conditional

influences, would

be the basis for detailed heuristics, design pattern

catalogs, and, eventually,
a kind of artificial-intelligence-based application that
could assist in design

ing UIs to be culturally conservative or radical according

to the needs/de

sires of those doing the designing. In the analysis that

follows, for a given culture dimension, each subsec

tion provides characteristic examples, as determined by

personal experi

ence, not yet by detailed study, of UI elements listed per

UI component, as

described previously. Although this is an initial

categorization by assertion,

I believe it nevertheless has pedagogic and practical value.

Power Distance Metaphors High: Government or corporate

institutions and buildings; objects

with clear hierarchies, e.g., the human body, schools,

government, mon

uments, etc. Low: Informal, or popular institutions,

buildings, and objects, that empha

size equality, options; Montessori schools, play/games,

public spaces, etc. Mental Models High: Complex, highly
organized, highly categorized, highly populated

structures, e.g., large corporate and/or government

organizational models

or charts; reference data with little or no relevancy

ranking Low: Simple, informally organized and categorized
structures; less struc

tured data with some or much relevancy FIG. 3.3. Matrix

mapping culture dimensions to user interface components.
Navigation High: Restricted access, choices;
authentication; passwords; prescribed

routes Low: Open access, multiple options, sharable paths

Interaction High: Severe error messages: “Entry Forbidden,”
“You are wrong;” wiz

ards or guides lead usage Low: Supportive error messages,

cue cards; many user-driven options

available Appearance High: Images of leaders;

national/corporate/government themes, slo

gans, insignia, logos, symbols, typefaces, layouts, colors;

official music, an

thems; formal speech Low: Images of people, groups; daily

activities; popular music, symbols,

typefaces, layouts, colors; informal speech

Individualism Versus Collectivism Metaphors Individualist:

Action-oriented, tool-oriented Collectivist:
Relationship-oriented, content-oriented Mental Models
Individualist: Product- or task-oriented Collectivist:
Role-oriented Navigation Individualist: Individual paths;
popular choices, celebrity choices; stable

across roles; customizable Collectivist: Group-oriented,

official choices; changes per role Interaction
Individualist: Keyword searches; active-oriented; multiple
devices; cus

tomizable; Collectivist: Limited, official devices;

role-driven Appearance Individualist: Images of products,
people; low context; hyperbolic, dy

namic speech; market-driven topics, imagery, language;

customizable; di

rect, active verbs Collectivist: Images of groups,

organizations; images of roles; high con

text; official, static terminology; institution-driven

topics, imagery, lan

guage; passive verbs

Masculinity Versus Femininity

Note: This subsection’s contents are differently organized

and may be mis

interpreted without considering the fact that the examples

list items for tra

ditional gender roles, not for masculine cultures or

feminine cultures per se.
Masculine cultures would maintain clear separation of such
examples; fem

inine cultures would tend to combine or merge them, making

fewer gen

der-oriented distinctions. Metaphors Masculine:

Sports-oriented; competition-oriented; work-oriented
Feminine: Shopping carts; family-oriented, people-oriented
Mental Models Masculine: Work/business structures;
high-level, “executive views;”

goal-oriented Feminine: Social structures; detailed views;

relationship-oriented Navigation Masculine: Limited
choices, synchronic Feminine: Multiple choices;
multitasking, polychronic Interaction Masculine:
Competitive-game-oriented; mastery-oriented; individual

oriented Feminine: Practical, function-oriented;

cooperation-oriented; team ori

ented Appearance Masculine: “Masculine” colors, shapes,

sounds Feminine: “Feminine” colors, shapes, sounds;
acceptance of cuteness

ity; fuzzy logic, i.e., multivariate, non-binary logic.

Navigation High: Desire for limited, clear organized
options; tolerance for complex,

fine tuning controls to “master” or “control” a situation,

e.g., advanced

searches on the Web, consumer electronics controls. Low:

Tolerance for ambiguous, possibly redundant options;
tolerance

for risk, gambling; tolerance for simple controls, e.g.,

simple searches on

the Web, or www.Google.com’s “I Feel Lucky” button.

Interaction High: Precise, complete, detailed input and
feedback of status; devices

that permit finetuning. Low: General, limited, or ambiguous

input and feedback of status; de

vices that may have gross tuning. Appearance High: Simple,

clear, consistent imagery, terminology, sounds; highly re
dundant coding of perceptual cues. Low: More varied,
ambiguous, less-consistent imagery, terminology,

sounds; tolerance for more perceptual characteristics

involved in purely or

namental or aesthetic use; less redundant coding of

perceptual cues.

major businesses, the Mafia, Chinese state businesses, IBM

in 1950s, etc.;

concrete metaphors Short: Interchangeable roles, jobs,

objects; less concrete, more abstract

metaphors Mental Models Long: Love/devotion; social

coherence, responsibility, support Short: Liberty: social
incoherence, social irresponsibility, efficiency Navigation
Long: Tolerance for long paths, ambiguity;
contemplation-oriented, re

liance on personal advisor; context-oriented Short:

Bread-crumb trails, clear taxonomies; quick-results;
logical infer

ence-oriented, function or action-oriented Interaction

Long: Preference for face-to-face communication, harmony;
personal

ized messages; more links to people; live chats;

interaction as “asking” Short: Distance communication
accepted as more efficient; anonymous

messages tolerated; conflict tolerated, even encouraged;

performance-crit

ical communication Appearance Long: Cultural markers:

flags, colors, national images; soft focus; warm,

fuzzy images; pictures of groups inviting participation,

suggestions of inti

macy and close social distance Short: Minimal and focused

images; short borders, lines, edges; concen

tration on showing task or product This initial taxonomy

represents an attempt to group UI attributes in a

manner that highlights the connection to culture

dimensions. A more de
finitive mapping clearly requires extensive effort, but
this initial construct

points the way.

FUTURE STUDY OF CULTURE IN RELATION

TO USER INTERFACE DESIGN

The preceding analysis is not the only way to relate UI

design and culture.

Recent publications highlight other cultural issues that

also should be con

sidered in understanding how cultural differences might

impact UI design.

These other reference dimensions could/should be mapped to

culture di

mensions and to UI components in future research. First of

all, Clausen (2000) argued that classical culture in the
United States

no longer exists, at least as defined by cultural

anthropologists in the past. Cul

ture in earlier circumstances represented group

environments that were diffi

cult to escape. Now, many uses of the term culture refer

simply to matters of

choice, not requirement, in other words, to lifestyles. The

strict use of the term

would not be applied to, for example, consumer product

affinity groups, such

as owners of Volvo automobiles. Consequently, designers may

need to debate

how strictly they apply the concepts of traditional

anthropology to current

product design and to performance, preference, and

enjoyment differences
among target groups. As noted earlier, this chapter adopts
a looser definition

of culture, rather than a stricter one, for practical

purposes in UI design. Some research about how different
cultures use different UI constructs

has appeared. Choong and Salvendy (1998), for example,

studied the use of

icons in Chinese UIs. For their article “Implications for

Design of Com

puter Interfaces for Chinese Users in Mainland China,”

Choong and

Salvendy (1999) investigated the impact of cultural

differences on com

puter performance for 40 Chinese subjects residing in

mainland China and

40 U.S. users and the design of appropriate UIs for the

Chinese users. Their

results are also commented on by Carroll (1999). Choong and

Salvendy found, in general, that U.S. users rely much

more on logical inference and categorization. These users

tend to classify

by functions, analyze components, and infer common

features. In con

trast, the Chinese users relied more on relations and

contexts. These users

tended to classify by interdependence within wholes, to

rely on subjective

experience without sharp differences between oneself and

others or be

tween facts and concepts. They investigated sorting the

contents of one’s house and found contrast

ing sorting styles. The Chinese tend to sort by thematic

groupings whereas

the U.S. users tend to sort by functions. They found that

when users were

given sorting tasks and the culturally appropriate sorting

methods, each had

the lowest error rates, and in contrast with the wrong kind
of tool, high error

rates. Furthermore, test results of Chinese users using

their preferred the

matic structures show better memory performance, speed, and

accuracy. The

Chinese also performed better using concrete metaphors.

These investiga

tions are in general accordance with the heuristic analysis

provided earlier

and seem to have strong implications for significant

performance differences

of users given culturally inappropriate UI components.

Robert Cialdini, professor at the University of New Mexico,
has written

about the dimensions of persuasion (Cialdini, 2001) by

which people con

vince others to think or act in a particular way. He

identifies the following

dimensions of persuasion, that is, factors that might

persuade one to follow

another’s request: • Authority: One might act out of

respect or fear of another’s authority, power, or position.
• Consistency: One might act out of belief in the
consistent following of rules of legal or sanctioned
behavior under the conditions of the request and
appropriate rules of behavior. • Liking: One might act out
of consideration of positive personal feelings toward the
person making the request. • Reciprocation: One might act
in consideration of one’s own requests in the past or
desired reciprocity when making future requests. •
Scarcity: One might act out of perceived scarcity of
referred-to ob jects, persons, and the perceived benefits
of following the request. • Social validation: One might
act out of consideration of value, sta tus, or other social
benefit accrued by following the request. Although several
of these dimensions combine in any situation to lead to

someone’s actions, Cialdini explains that different

cultures typically em

phasize one factor over others. For example, he comments

that an Ameri

can might decide to honor a request from another person in

an office based

on whether that person had recently done him or her a

favor, whereas a

German would consider the rules regarding the request, a

Spaniard would

consider whether the person is a friend, and a Chinese

office worker would

consider the person’s authority to issue the request.

Although Cialdini does

not consider Web site persuasion and does not elaborate

extensively on cul

tural differences that might affect e-commerce and

m-commerce, his analy

sis clearly opens the door to such considerations. As an

example of additional dimensions in relation to UI design
and cul

ture, consider Bailey, Gurak, and Konstan (2001), who

developed a set of

dimensions for establishing trust on the Web. They assert

that the following

these dimensions vary in their importance among different

cultures glob

ally, and need to be examined in the light of these

differences in order to

predict the influence of each within varying cultures. In

Frames of Mind, The Theory of Multiple Intelligences
(1985), Gardner
presents his theory of the dimensions of intelligence,
considers much edu

cation and communication biased to verbal comprehension,

and urges

greater diversity in communication, evaluation, and

learning. His dimen

sions are the following: • Verbal/Image comprehension. •

Word/image fluency. • Numerical/graphical fluency. •
Spatial visualization. • Associative memory. • Perceptual
speed. • Reasoning. • Image: Self/Other awareness. It seems
reasonable that different cultures might place different
empha

ses on which dimensions are valued within that culture and

emphasized in

education, at work, and at home. A future task for UI

design theory and

practice is to explore how UI components, and specific

patterns of combi

nations, would relate to these dimensions of intelligence

and to the varying

roles given to them by different cultures. Richard Nisbett,

a social psychologist from the University of Michigan,

has studied how culture molds habits of thought. He uses

examples of the

differences in descriptions of what Japanese and North

Americans observe

in describing simple scenes, for example, what happens in a

fish tank

(Nisbett, 2001, p. 6). The Japanese observers note

relations among the fish,

the aquatic flora, and the simple environment. North

American observers,

on the other hand, tend to focus only on the actions and

attributes of the

primary fish. The contrast is stark: One culture is much

more oriented to

figure-field relationships; the other culture concentrates

on the figure. Sim

ilar differences have been noted by Kaiping Peng at the

University of Cali

fornia at Berkeley. Peng (2000) noted that Chinese students

seem less eager

to resolve contradictions than U.S. students. Cultural

factors seem to run

deeply in mental processes. One possible conclusion is that

Western schol

ars may be in error to assume that there is, or should be,

a universal focus on

reasoning, categorization, and linear cause-and-effect

explanations of situ

ations and events. Such conclusions differ markedly from

classical assump

tions that would affect the design of UIs. David Landes,

professor of history at Harvard University, argues in

“Culture Makes Almost All the Difference” (2000) that

social attitudes are

more important than politics and economics in determining

why some soci

eties are richer than others. This version of the thesis

that culture matters

greatly is at odds with assumptions of market economists

and liberal and

Marxist philosophers, who believe that political and

economic factors are of

primary importance. Hofstede (1997) has a more mixed

approach. He rec

ognizes that both cultural as well as political/economic

factors determine
whether certain cultures are dominant at different points
in history. Hofstede comments that cultural orientations
are extremely deeply em

bedded in cultures over hundreds and thousands of years;

consequently, he

feels that even modern communication media have not

dislodged these cul

tural orientations. In fact, cultures tend sometimes to

concentrate harder

on preserving their approaches given the encroachment of

alien cultures.

On the other hand, recent news events seem to argue

otherwise. For exam

ple, a recent article comments that Japan’s 10-year

economic decline has

stimulated the rise of individualism (Ono & Spindle, 2000).

CONCLUSIONS

The design of products and services for the Web and mobile
devices fosters

the need for good cross-cultural communication in UI

design. Eng

lish-speaking countries constitute 8% of the world’s

population, but by

2005, approximately 75% of Internet users will be

non-English speaking. It

seems likely that cultural factors will need to be

considered more fre

quently. Already, 80% of corporate Web sites in Europe

offer more than

English even though launching multilanguage Web site

portals with 11 Eu

ropean languages is a significant burden to operations.

Consequently, cross-cultural analysis and design issues
need to be con
sidered more integrally in planning stages, and developers
need checklists

and guidelines to assist them in their design phases.

Having a better under

standing of the mappings of culture dimensions to UI

components, as well

as to such dimensions as trust or intelligence, may inform

designers to make

better decisions about usability, aesthetics, and emotional

experience. Ulti

mately, computer-assisted tools may result. As indicated

earlier, increased

performance, preference, and enjoyment seem likely outcomes

of cultur

ally appropriate UI design.

ACKNOWLEDGMENTS

I acknowledge the assistance of Emilie Gould, cited

earlier, in organizing

and conducting the tutorials, and the tutorial

participants, through which

portions of the material in this chapter were developed. I

also acknowledge

two works of mine (Marcus, 2001, 2002), edited portions of

which appear in

this chapter.

Bailey, B. P., Gurak, L. J., & Konstan, J. A. (2001). An

examination of trust production in computer-mediated
exchange. Paper presented at the seventh Human Factors and
the Web 2001.

Cialdini, R. (2001, February). The science of persuasion.

Scientific American, 284(2), 76–81.

De Mooij, M. (2001). Internet and culture. Paper presented

at. Internet, Economic Growth and Globalization organized
by Institute for International and Regional Economic
Relations, Gerhard-Mercator University, Duisburg, Germany,
8 Au gust 2001.

Carroll, J. M. (1999). Using design rational to manage

culture-bound metaphors for international user interfaces.
Proceedings, International Workshop on International
Products and Services, Rochester, NY, pp. 125–131.

Choong Y., & Salvendy, G. (1998). Designs of icons for use

by Chinese in mainland China. In Interacting with
computers. International Journal of Human–Computer
Interaction, 9:4, February 1998, pp. 417–430. Amsterdam:
Elsevier.

Choong, Y., & Salvendy, G. (1999). Implications for design

of computer interfaces for Chinese users in Mainland China.
The International Journal of Human-Com puter Interaction,
11:1 1999, pp. 29-46. Amsterdam: Elsevier. Cialdini,
Robert. (2001). The Science of Persuasion, Scientific
American, Vol, 284, No. 2, February 2001, pp. 76–81
(www.influenceatwork.com).

Clausen, C. (2000). Faded mosaic: The emergence of

post-cultural America, Ivan R. Dee.

Eco, U. (1976). A theory of semiotics. Bloomington: Indiana

University Press.

Gardner, H. (1985). Frames of mind: The theory of multiple

intelligences. New York: Basic Books.

Gelerntner, D. (2000). The second coming: A manifesto.

http://www.edge.org/docu ments/archive/edge70.html

Hofstede, G. (1997). Cultures and organizations: Software

of the mind. New York: McGraw-Hill.

Lakoff, G., & Johnson, M. (1980). Metaphors we live by.

Chicago: University of Chi cago Press.

Landes, D. (2000). Culture makes all the difference. In L.

E. Harrison & S. P. Hun tington (Eds.), Culture matters:
How values shape human progress (pp. 2–13). New York:
HarperCollins.

Lévi-Strauss, C. (2000). Structural anthropology. Trans.

Claire Jacobson and Brooke Schoepf. New York: Basic Books.

Marcus, A. (1998). Metaphors in user-interface design.

Journal of Computer Documentation, 22(2), 43–57).

Marcus, A. (2001). Cross-cultural user-interface design. In

M. J. Smith & G. Salvendy (Eds.), Proceedings of the
Human–Computer Interface International Conference (Vol. 2,
pp. 502–505). Mahwah, NJ: Lawrence Erlbaum Associates.

Marcus, A., & Gould, E. W. (2000). Crosscurrents: Cultural

dimensions and global web user-interface design.
Interactions, (7)4, 32–46.

Nisbett, Richard E., Kaipeng Peng, Incheol Choi, & Ara

Norenzayan (2001). Culture and systems of thought: Holistic
vs. analytical cognition, Psychological Review, 108,
291–310.

Oliver, R. T. (1971). Communication and culture in ancient

India and China. Syracuse, NY: Syracuse University Press.

Ono, Y., & Spindle, B. (2000, December 30). Japan’s long

decline makes one thing rise: Individualism. Wall Street
Journal, p. A1.

Peirce, C. S. (1933). Existential graphs. In C. Hartshorne

& P. Weiss (Eds.), Collected papers of Charles Sanders
Peirce (Vol. 4, pp. 293–470). Cambridge, MA: Harvard
University Press.

Peng, K. (2000). Readings in cultural psychology:

Theoretical, methodological and empirical developments
during the past decade (1989–1999). New York: Wiley.

Tidwell. J. (1999). Common ground: A pattern language for

human–computer interface de sign. Retrieved February 20,
2002, from www.mit.edu/~jtidwell/interaction_pat terns.html

Atchison, Jean. (1996). The atlas of languages: The origin

and development of languages throughout the world, Facts On
File. New York, 1966.

Alvarez, G. M., Kasday, L. R., & Todd, S. (1998). How We

Made the Web Site Interna tional and Accessible: A Case
Study. Proceedings (CD-ROM), Fourth Human Factors and the
Web Conference, Holmdel, New Jersey.

American Institute of Graphic Arts (AIGA). (1981). Symbol

signs. Visual Communica tion Books. New York: Hastings
House.

Associated Press. (2001). Saudi Arabia issues edict against

Pokemon, San Francisco Chronicle, 27 March 2001, p. F2.

Barber, W., & Badre, A. (1998). Culturability: The merging

of culture and usability. Proceedings of the Fourth Human
Factors and the Web Conference.

Bliss, C. K. (1995). Semantography. Sidney, Australia:

Semantography Publica tions.

Campbell, George L. (2000). Concise compendium of the

world’s languages. Sandpoint, ID: MultiLingual Computing
and Technology.

Comrie, Bernard, Stephen Matthews, and Maria Polinsky

(Eds.). (2000). The Atlas of languages. Sandpoint, ID:
MultiLingual Computing and Technology.

Coriolis Group. (1998). How to build a successful

international web site. Scottsdale, AZ: Coriolis Group.

Cox, Jr. Taylor (1994). Cultural diversity in

organizations, San Francisco: Berrett-Koehler Publishers.

Crystal, David. (1987). The Cambridge Encyclopedia of

Language. Cambridge: Cambridge University Press.

Daniels, Peter T., & William Bright. (Eds.). (2000). The

World’s Writing Systems. Sandpoint, ID: MultiLingual
Computing and Technology.

Day, Donald L., Elisa M. del Galdo, and Girish V. Prabhu.

(Eds.). (2000). Designing for Global Markets 2, Second
International Workshop on Internationalisation of Products
and Systems, 13–15 July, Baltimore, MD. Rochester, NY:
Backhouse Press.

Day, Donald. (2000). Gauging the Extent of

Internationalization Activities. In the Proceedings of the
Second International Workshop on Internationalization of
Products and Systems, 13–15 July 2000, Baltimore, MD, pp.
124–136. Rochester, NY: Back house Press.

DelGaldo, E., & Nielsen, J. (Eds.) (1996). International

user interfaces. New York: John Wiley and Sons.

Doi, Takeo. (1973). The anatomy of dependence. New York:

Kodansha Interna tional.

Doi, Takeo. (1986). The anatomy of self: The individual

versus society. New York: Kodansha International.
Dreyfuss, Henry. (1966). Symbol sourcebook. New York: Van
Nostrand Rhinehold.

Elashmawi, Farid, and Harris, Philip R. (1998).

Multicultural management 2000: Essential cultural insights
for global business success. Houston: Gulf Publishing Co.

Fernandes, T. (1995). Global interface design: A guide to

designing international user inter faces. Boston: AP
Professional.

French, Tim, and Smith, Andy. Semiotically enhanced web

interfaces: Can semiotics help meet the challenge of
cross-cultural design? In the Proceedings of the Second
International Workshop on Internationalization of Products
and Systems, 13–15 July, Baltimore, MD, pp. 23–38.
Rochester, NY: Backhouse Press.

Frutiger, A. (1997). Signs and symbols: Their design and

meaning. Cham, Switzerland: Syndor Press GmbH.

Goode, Erica. (2000). How culture molds habits of thought.

New York Times, 8 August 2000, p. D1 ff.

Gould, Emilie. W., Zakaria, Norhayati, and Shafiz Affendi

Mohd. Yusof. (2000). Ap plying culture to website design: A
comparison of Malaysian and US websites. In the Proceedings
of the IPCC/SIGDOC Conference, Boston, 25 September 2000.

Graham, Tony. (2000). Unicode™: A Primer. Sandpoint, ID:

MultiLingual Com puting and Technology.

Hall, E. (1969). The hidden dimension. New York: Doubleday.

Harel, Dan, and Girish Prabhu. (1999). Global User

Experience (GLUE), Design for cultural diversity: Japan,
China, and India. Proceedings of the First International
Workshop on Internationalization of Products and Systems,
20–22 May 1999, pp. 205–216. Rochester, NY: Backhouse
Press.

Harris, John, and McCormack, Ryan. (2000). Translation is

Not Enough: Consider ations for Global Internet
Development. San Francisco: Sapient. Retrieved Feb ruary
18, 2002, from
http://www.sapient.com/pdfs/strategic_viewpoints/
sapient_globalization.pdf

Harris, Philip R., and Moran, Robert T. (1993). Managing

cultural differences: High performance strategies for a new
world of business. Houston, TX: Gulf Publishing Co.

Helander, Martin G., et al. (Eds.). (1997). Handbook of

human–computer interaction. Amsterdam, Netherlands:
Elsevier Science.

Hendrix, Anastasia. (2001). The nuance of language. San

Francisco Chronicle, 15 April, 2001, p. A10.

Herskovitz, Jon. (2000). J-Pop Takes Off: Japanese music,

movies, TV shows enthrall Asian nations. San Francisco
Chronicle, 26 December 2000, p. C2.

Hoft, Nancy L. (1995). International technical

communication: How to export information about high
technology. New York: John Wiley and Sons, Inc.

Honold, Pia. (1999). Learning how to use a cellular phone:

Comparison between German and Chinese users. Journal of
Society of Technical Communication, 46(2), May 1999, pp.
196–205.

International Organization for Standardization (ISO).

(1996). ISO 7001: Public In formation Symbols of The
American National Standards Institute. Geneva: ISO.

International Organization for Standardization (ISO).

(1993). ISO 7001: Public In formation Symbols: Amendment 1
of The American National Standards Institute. Geneva: ISO.

International Standards Organization (ISO). (1989).

Computer Display Color: Draft Standard Document 9241-8,
Geneva: ISO.

Iwayama, M., Tokunaga, T., & Tanaka, H. (1990). A method of

calculating the mea sure of salience in understanding
metaphors. In the Proceedings of the American As sociation
for Artificial Intelligence (AAAI), 1990, pp. 298–303.

Kimura, Doreen. (2002). Sex differences in the brain.

Scientific American, 267(3), pp. 118–125.

Konkka, Katja, and Koppinen, Anne. (2000). Mobile devices:

Exploring cultural differnces in separating professional
and personal time. In the Proceedings of the Second
International Workshop on Internationalization of Products
and Systems, 13–15 July, Baltimore, MD, pp. 89–104.
Rochester, NY: Backhouse Press.
Kurosu, M. (1997), Dilemma of usability engineering. In
Salvendy, G., Smith, M., and Koubek, R. (Eds.): Design of
Computing Systems: Social and Ergonomics Consider ations
(Vol. 2), Proceedings of the Seventh International
Conference on Hu man–Computer Interaction. San Francisco,
USA, August 24–29, 1997, pp. 555–558. Amsterdam: Elsevier.

Lanham, R. A. (1991). A handlist of rhetorical terms (2nd

ed.). Berkeley: University of California Press.

Leventhal, L., et al. (1996). Assessing user interfaces for

diverse user groups: Evalua tion strategies and defining
characteristics. Behaviour and Information Technology,
15(3), pp. 127–139.

Lewis, Richard. (1991). When cultures collide. London:

Nicholas Brealey.

Lingo Systems. (1999). The guide to translation and

localization. Los Alamitos, CA: IEEE Computer Society.

The Localization Industry Standards Association (USA).

(1999). The localization in dustry primer. Féchy,
Switzerland: LISA.

Marcus, Aaron. (1983). Visual rhetoric in a

pictographic-ideographic narrative. In the Proceedings of
the Second Congress of the International Association for
Semiotic Studies, Vienna, July 1979. Tasso Borbé (Ed.),
3(6), pp. 1501–1508. Berlin: Mouton Publishers,

Marcus, Aaron. (1992). Graphic design for electronic

documents and user interfaces. Reading: Addison-Wesley.

Marcus, Aaron. (1993). Human communication issues in

advanced UIs. Communications of the ACM, 36 (4), pp.
101–109.

Marcus, Aaron. (1993). Designing for diversity. In the

Proceedings of the 37th Human Factors and Ergonomics
Society. Seattle, Washington, 11–15 October 1993, Vol. 1,
pp. 258–261.

Marcus, Aaron. (1995). Principles of effective visual

communication for graphical user-interface design. In
Baecker, Grudin, Buxton and Greenberg (Eds.), Readings in
Human–Computer Interaction (2nd ed., pp. 425–441). Palo
Alto, CA: Mor gan Kaufman.

Marcus, A. (2000). International and intercultural user

interfaces, in User interfaces for all, ed. Dr. Constantine
Stephanidis, Lawrence Erlbaum Associates Publishers, New
York, 2000, pp. 47–63.

Marcus, Aaron, and Chen, Eugene. (2002). Designing the PDA

of the future. Interac tions, 9:1, pp. 32–44.

Marcus, Aaron (2002). User-interface design and culture

dimensions, Workshop on Internationalization, Proceedings.
CHI 2002, Minneapolis, MN, 21-24 April 2002. in press.

Marcus, Aaron, et al. (1999). Globalization of

user-interface design for the web, Proc., 5th Conference on
Human Factors and the Web, 3 June 1999, NIST, Gathersburg,
MD, avail. from www.tri.sbc.com/hfweb.

McLuhan, M. (1964). Understanding media: The extensions of

man. New York: McGraw-Hill.

Neale, D. C. and John M. C. (1997). The role of metaphors

in user-interface design. In Helander, M., Landauer, T. K.
and Prabhu, P. (Eds.). Handbook of human–com puter
interaction (2nd ed), pp. 441–462. Amsterdam: Elsevier
Science.

Neustupmy, J. V. (1987). Communicating with the Japanese.

Tokyo: The Japan Times.

Nielsen, J. (Ed.) (1990). Designing user interfaces for

international use (Vol. 13). Ad vances in human
factors/ergonomics, Amsterdam: Elsevier Science.

Nisbett, R. E. (2003). The geography of thought: How Asians

and Westerners think differ ently … and why. New York: Free
Press.

Lee, Ook, The Role of cultural protocol in media choice in

a confucian virtual work place, IEEE Transactions on Prof.
Comm., Vol. 43, No. 2, June 2000, pp. 196–200.

Olgyay, N. (1995). Safety symbols art. New York: Van

Nostrand Reinhold.

Ota, Yukio. (1987). Locos: Lovers communications system (in

Japanese), Pictorial Insti tute, Tokyo, 1973. The author
presents the system of universal writing that he in vented.

Ota, Yukio, Pictogram design. Tokyo: Kashiwashobo.

Perlman, Gary. (2000). ACM SIGCHI Intercultural Issues. In

the Proceedings of the Second International Workshop on
Internationalization of Products and Sysems, 13–15 July,
Baltimore, MD, pp. 183–195. Rochester, NY: Backhouse Press.

Pierce, T. (1996). The international pictograms standard.

Cincinnati, OH: ST Publica tions.

Pirkl, James J. (1994). Transgenerational design: Products

for an aging population. New York: Van Nostrand Reinhold.

Prabhu, G. V., Chen, B., Bubie, W., and Koch, C. (1997).

Internationalization and localization for cultural
diversity. In Salvendy, G., Smith, M., & Koubek, R. (Eds.):
Design of Computing Systems: Cognitive Considerations (Vol.
1), Proceedings of the Seventh International Conference on
Human–Computer Interaction, San Francisco, USA, August
24–29, 1997, pp. 149–152. Amsterdam: Elsevier.

Prabhu, Girish and Dan Harel. (1999). GUI Design preference

validation for Japan and China: A Case for KANSEI
Engineering? In the Proceedings of the Eighth International
Conference on Human–Computer Interaction, Munich, Germany
22–26 August 1999, pp. 521–525. Amsterdam: Elsevier.

Prabhu, Girish V., and delGaldo, Elisa M. (Eds.). (1999).

Designing for global mar kets 1, Proceedings of the First
International Workshop on Internationalization of Products
and Systems, 20–22 May 1999, Rochester, NY. New York:
Backhouse Press.

Rowland, Diana. (1985). Japanese business etiquette. New

York: Warner Books.

Shahar, Lucy, and Kurz, David. (1995). Border crossings:

American interactions with Is raelis. Yarmouth, Maine:
Intercultural Press.

Sheppard, Charles, and Jean Scholtz (1999). The effects of

cultural markers on website use. In the Proceedings of the
Fifth Conference on Human Factors and the Web.

Stephanidis, C., et al. (Eds.). (1998). Toward an

information society for all: An Inter national R&D agenda.
International Journal of Human–Computer Interaction, 10(2),
pp. 107–134.

Stephanidis, Constantine. (Ed). (2000). User interfaces for

all. New York: Lawrence Erlbaum Associates Publishers.

Stille, Alexander. (2001). An old key to why countries get

rich: It’s the culture that matters, some argue anew. New
York Times, 13 January 2001.

Storti, Craig. (1994). Cross-cultural dialogues: 74 brief

encounters with cultural difference. Yarmouth, ME:
Intercultural Press.

Tannen, Deborah. (1986). That’s not what I meant! New York:

Ballantine Books.

Tannen, Deborah. (1990). You just don’t understand: Women

and men in conversation. New York: William Morrow and
Company.

Trompenaars, F., & Hampden-Turner, C. (1998). Riding the

waves of culture. New York: McGraw-Hill.

Vickers, Ben. (2000). Firms push to get multilingual on the

web. Wall Street Journal, 22 November 2000, p. B11A.

Victor, David A. (1992). International business

communication. New York: HarperCollins Publishers.

Waters, Crystal. (1997). Universal web design: Reach more

viewers, improve site appear ance, and enhance your
marketability through efficient design. Indianapolis,
Indiana: New Riders Publishing.

Yardley, Joanathan. (2000). Faded mosaic nixes idea of

‘cultures’ in U.S. San Fran cisco Examiner, 7 August 2000,
p. B3.

Yeo, Alvin Wee. (2001). Global-software development

lifecycle: An exploratory study. In the Proceedings of
Special Internet Group for Computer–Human Interaction
(SIGCHI), 31 March–5 April 2001, Seattle, WA, pp. 104–111.

URLs and Other Resources

Selected URLs and other resources from a list maintained by

the author

and available by contacting him at

<Aaron.Marcus@AMandA.com> are

the following:

ACM/SIGCHI Intercultural Issues database:

www.acm.org/sigchi/intercultural/
ACM/SIGCHI Intercultural listserve:
chi-intercultural@acm.org. Moderator: Donald Day
<d.day@acm.org.>.

African-American Websites: www.bet.com, www.netnoir.com,

www.blackfamilies.com

Color: www.colortool.com

Cultural comparisons: www.culturebank.com

Culture resources: www.webofculture.com,

www.acm.org/sigchi/intercultural/

Digital divide: www.digitaldivide.gov,

www.digitaldivide.org, www.digitaldividenet work.org

Global business: www.globalbusinessmag.com

Glossary, six languages:

www.bowneglobal.com/bowne.asp?page=9&language=1

Indian culture: www.indiagov.org/culture/overview.htm

Internationalization providers: Basis Technology

<www.basistech.com>, CIJ America <www.cij.com>,
www.Logisoft.com

Internationalization resources:
www.world-ready.com/r_intl.htm, www.world
ready.com/biblio.htm

Internet statistics by language:

www.euromktg.com/globstats/index.html, www.world
ready.com/biblio.htm

Java Internationalization:
http://java.sun.com/docs/books/tutori

Localization: www.lisa.org/home_sigs.html

Localization providers (selected): www.Alpnet.com,

www.Berlitz.com, www.global sight.com, Lernout & Hauspie
<www.lhsl.com>, www.Lionbridge. com, www.Logisoft.com,
www.Logos-usa.com, www.translations.com, www. Uniscape.com

Machine translation providers: Alta Vista’s Babelfish

<www.babelfish.altavista. com>, www.IDC.com,
www.e-Lingo.com, Lernout & Hauspie <www.lhsl.com>,
www.Systransoft.com
Microsoft’s global development page:
www.eu.microsoft.com/globaldev/fareast/ fewinnt.asp

Microsoft Planning for and testing global software:

http://www.microsoft.
com/GLOBALDEV/Non%20mirror/back%20up/gbl-gen/INTREFNEW.HTM

Microsoft Windows Internationalization:

http://www.microsoft.com/globaldev/ gbl-gen

Native-American-oriented Website:
www.hanksville.org/NAresources/

Simplified English: userlab.com/SE.html

Internet users survey, Nua:

www.nua.ie/surveys/how_many_online

Unicode: www.unicode.org/, IBM Unicode Glossary:

www-4ibm.com/software,/de
veloper/library/glossaries/unicode.html

Women: wow.com, oxygen.com, chickclick.com

Resources

This section lists organizations providing information

about international

standards and intercultural issues.

• ACM SIGCHI Intercultural Issues Database

Gary Perlman maintains the ACM SIGCHI Intercultural Issues

Database,

which is summarized in (Perlman). Gary Perlman, OCLC Online

Computer Library Center 5565 Frantz Road, Dublin, OH 43017
Email: Gary Perlman <perlman@TURING.ACM.ORG> URL:
http://www.acm.org/sigchi/intercultural/

• American National Standards Institute (ANSI)

This organization analyzes and publishes U.S. standards,

including those

for icons, color, terminology, user interfaces, and so on.

Their contact data
are the following: American National Standards Institute
(ANSI) 11 West 42nd Street, 13th Floor, New York, NY10036
Tel: 212-642-2000 Email: info@ansi.org URL: www.ansi.org

• China National Standards Organization

This organization analyzes products to be imported into

China. The con

tact data are the following: China Commission for

Conformity of Electrical Equipment (CCEE) Secretariat
Postal address: 2 Shoudu Tiyuguan, NanLu, 100044, P.R.
China Office address: 1106, 11th floor, 2 Shoudu Tiyuguan,
NanLu, Beijing, P.R.China Tel: +86-1-8320088, ext. 2659,
Fax: +86-1-832-0825

• East-West Center

This organization, formerly funded by the U.S. Congress, is

a center for

technical and cultural interchange among Pacific Rim

countries. The cen

ter’s research and publications cover culture and

communication. The con

tact data are the following: East-West Center 1601

East-West Road, Honolulu, HI 96848-1601 el: 808-944-7111,
Fax: 808-944-7376 Email: ewcinfo@ewc.hawaii.edu URL:
http://www.ewc.hawaii.edu

• Information Technology Standards Commission of Japan

(ITSCJ)

Information Processing Society of Japan Kikai Shinko

Building, No. 3-5-8 Shiba-Koen, Minato-ku, Tokyo 105, Japan
Tel: +81-3-3431-2808, Fax: +81-3-3431-6493

• Institute for International Research

This organization conducts research and organizes

conferences about glob

alization and localization: Institute for International

Research, Inc. 2009 N. 14th Street, Suite 303, Arlington,
VA 22201 Tel: +1-703-908-9010, Fax: +1-703-908-9018 URL:
www.iir-ny.com

• International Standards Organization (ISO)

This organization analyzes and publishes world standards
for all branches

of industry and trade, including standards for icons,

color, terminology,

user interfaces, etc. Their contact data are the following:

International Standards Organization (ISO) Geneva,
Switzerland Tel: +41-22-749-0111, Fax: +41-22-733-3430
Email: central@iso.ch URL: http://www.iso.ch/ Of special
interest is ISO 8601’s international time and date
standards.

Information about ISO’s 8601 standards, and the particular

one about time

and date standards, may be found at the following URLs:

http://www.aegis1.demon.co.uk/y2k/y2kiso.htm

• Japan National Standards Organization

These organizations analyze and publish Japanese standards,

including

those for icons, color, terminology, user interfaces, and

so forth. Their con

tact data are the following: Japanese Industrial Standards

Committee (JISC) Agency of Industrial Science and
Technology Ministry of International Trade and Industry
1-3-1, Kasumigaseki, Chiyoda-ku, Tokyo 100, Japan Tel:
+81-3-3501-9295/6, Fax: +81-3-3580-1418

• Localization Industry Standards Organization (LISA), The

This organization provides information and publications

relevant to local

ization professionals: The Localization Industry Standards

Association 7, rue du Monastère, 1173 Féchy, Switzerland
Tel: +41-21-821-32-10, Fax: +41-21-821-32-19 URL:
http://www.lisa.org

• Web of Culture, The

This Web site publishes news and resources related to

cross-cultural com

munication: Ms. Eileen Sheridan, President 113 Anita Road,

Suite 6, Burlingame, CA 94010 Tel/Fax: +1 408 273 6074 URL:
<http://www.webofculture.com>

• World Wide Web Consortium This organization provides

information relevant to globalization issues,

including accessibility. Two URLs of interest are:

http://www.w3.org/International, for information about
internalization. http://www.w3.org/WAI, for information
about accessibility. Chapter

Synthesizing the Literature

on Cultural Values

Emilie W. Gould

Rensselaer Polytechnic Institute, Troy NY

INTRODUCTION

I was having lunch with a Malaysian friend who was having

problems with

her doctoral advisor. “When I say I want to study culture

and information

systems, he says, ‘That’s too broad!’ When I ask him what

he means, he

says, ‘Culture is everywhere—national, ethnic, religious,

corporate, fam

ily—but it’s not consistent. You can’t make useful

generalizations based on

culture. Even if you could, it would be impossible to

develop separate in

terfaces for each different culture. Focus on the

psychology of the individ

ual user instead!’” I said, “Yeah, I’ve had that problem

too. A lot of people in the field of

human–computer interaction (HCI) really resist the notion

of culture be

cause it seems like such a vague, anthropological concept.

When HCI first
evolved from experimental psychology, you could run a test
in the usabil

ity lab and claim to know what every novice user might
need. Then people

designing cooperative work discovered qualitative research.

But HCI is

just coming to grips with culture. I tell people, ‘I’m

applying a set of

Internet.’ They like that better.” Legitimating the study

of culture has been a problem for anyone inter

ested in the intersection of national difference, personal

preference, and 79

computer interfaces. We all know individuals from other

cultures who seem

more like us than like their compatriots. They make it hard

to deny the

claim that “People are where the interface is, not

culture!” But you can say

the same for any social science. People are different, yet
general psychologi

cal profiles have been developed. Organizations vary, but

in a regular way.

Countries can be grouped together in useful

typologies—democratic, so

cialist, authoritarian, and so forth. Within cultures,

individuals cover a wide

spectrum of belief and behavior but, in the aggregate, they

cluster together

and these clusters display a surprising amount of

stability. So, what is the best way to analyze national
culture to improve human in

terface design? There are many ways of thinking about the

problem, but
generally some theories stress how cultures differ while
others examine

what they have in common—the “emic/etic” distinction of

linguist Kenneth

L. Pike (1954). Pike contrasted the universality of

phonetics (language

sounds) with the diversity of phonemics (interpretations of

those sounds as

coherent systems of communication). His analysis has been

widely applied

to social phenomena like writing and rhetoric that are

based on a universal

system like the alphabet but require selective

interpretation by “insiders.” Cultural anthropology can
provide useful insights into designing inter

faces for specific countries, but theories from the field

of intercultural com

munication are generally better for culturally diverse

audiences. Most

designers do not have the mandate to develop entirely

different products

for each national or ethnic market. Intercultural

communication theory

makes it possible for them to focus on a few crucial

variations. Since the

1990s, when the concept of culture finally emerged as a

concern for inter

face design, most design ideas have come from intercultural

communica

tion, most notably, in articles and workshops by Aaron

Marcus (1993, 2001;

Marcus & Gould, 2000) and in the essays in Elisa M. del

Galdo and Jakob

Nielsen’s International User Interfaces (1996). For the

rest of this chapter, I (a) further discuss the advantages
of

intercultural communication theories for HCI design, (b)

list key communi

cation theorists whose work has heuristic value for the

field, and (c) identify

the most important variables for interface development.

THE ADVANTAGES OF INTERCULTURAL COMMUNICATION

THEORIES FOR HCI DESIGN

Though cultural anthropology seems like it should be

valuable for interface

development, it has several limitations as a design guide.

First, it tends to

focus on specific cultures to the detriment of comparison

and theory and,

second, it often emphasizes authenticity over interaction.

Some of these

problems are inherent in its chosen methodology; others

derive from its

historical development as an academic field. Many

anthropological studies use “thick description” to record
and

highlight behaviors and attitudes that contrast with the

anthropologist’s

own culture. Unfortunately, this approach emphasizes a

degree of cultural

uniqueness that may make it seem impossible to bridge

cultural differences.

Although each culture has its own symbols and patterns of

action, not all dif

ferences are significant. Some of the interface development

literature on

cultural markers illustrates this problem. Tony Fernandes

(1995) first (de

lightfully) introduced the problem of translating cultural

symbols, taboos,

and aesthetics into interface design but did not identify

ways to determine

which symbols were important. Wendy Barber and Albert Badre

(1998) in

troduced the term cultural marker to define interface

elements that “signify a

cultural affiliation.” They talked about developing a

directory of flags, col

ors, and national symbols that would allow interfaces to be

automatically

recompiled with different cultural markers for different

markets. Charles

Sheppard and Jean Scholtz (1999) tried to test this idea,

but their results

were not significant. Their effort demonstrates the

difficulty of working

with an essentially atheoretical concept. By contrast, the

field of intercultural communication focuses on the reg

ularities between cultures, identifying a few critical

variables that help peo

ple craft appropriate messages and collaborate. In

particular, the literature

on value orientations reduces cultural differences to a

manageable number. In 1950, Florence Rockwood Kluckhohn
first systematically com

pared cultures according to their values. The following

year, her hus

band Clyde Kluckhohn (1951) used the term value

orientations in an essay

on theories of action. For the next decade, Harvard

University spon
sored Florence Kluckhohn and Fred L. Strodtbeck in the
Values Project.

This longitudinal study compared five distinct

communities—Zuni, Na

vaho, Mexican, Mormon, and Texan—living near one another in

the

American Southwest. Their results were published in 1961 as

Variations in Values Orientations. The

book’s most notable contribution was its characterization

of each community

on the basis of five existential (and thus presumably

universal) questions:

What is the purpose of human nature (and can people

change)? Human-nature orientation

How do people relate to nature (and the supernatural)?

Man-nature orientation

How do people manage time? Time orientation

What is the purpose of human activity? Activity orientation

How do people relate to each other? Relational orientation

Three possible variations were assigned to each question.
Taken to

gether, these value orientations allowed Kluckhohn and

Strodtbeck to

identify opportunities for conflict and cooperation and to

analyze each

community equitably. However, few other cultural

anthropologists followed their lead.

Through the rest of the decade, many studies continued to

emphasize the

distance between “traditional” societies and “evolved”

cultures. Eventually,

postcolonialist critics like Edward Said (1978/1994) began

to document the

role of academic disciplines in facilitating cultural

imperialism. “Oriental

ism” and its ilk were seen to justify empire by emphasizing

the inferiority of

non-Western cultures. Anthropology came under attack for

its focus on iso

lated, marginalized groups. The field took these criticisms

to heart; this crisis reformed its study of

culture. Anthropologists today use a variety of techniques

to avoid the un

equal power relationships inherent in studies of “exotic”

people. Members

of the culture collaborate in the research (much as do

union members in

Participatory Design); lightly edited (but carefully

selected) personal narra

tives and field notes may be published in place of abstract

analysis; and, an

thropologists increasingly investigate their own cultures.

But such work too often remains unitary and singular. In
1946, Ruth

Benedict said that a culture might share 90% of its

observances with its

neighbors, but the differences, “however small in

proportion to the whole,

turn its future course of development in a unique

direction” (1946/1989,

p. 9.) Postmodernist and postcolonialist studies of

“difference” often

make it appear that nothing important is shared except the

experience of

oppression. Unfortunately, the moral imperative to avoid

stereotyping
and the desire to recognize “authenticity” are antithetical
to a utilitarian

focus on collaboration. Intercultural communication came

through a similar crisis but devel

oped a more useful methodology. The field emerged after

World War II

and was shaped by the collapse of Europe, the emergence of

the pax

Americana, and political and economic globalization. In the

1940s, the diffi

culties facing many American diplomats and representatives

of non

governmental organizations led Congress to reorganize the

Foreign

Service and found the Foreign Service Institute (FSI).

Linguists and anthro

pologists were hired to provide language and culture

training. Historians

and communications researchers collaborated with

descriptive linguists

like George L. Trager (who focused on nonverbal

paralanguage) and Ray L.

Birdwhistell (who developed kinesics). In 1959, the year

after William J.

Lederer and Eugene Burdick’s Ugly American made the

best-seller list, Ed

ward T. Hall proposed a solution to the problem of

“culture.” The Silent Language was based on examples from
the FSI and showed how

ignorance of cultural patterns and nonverbal communication

underlay nu

merous cases of intercultural miscommunication. The book

was easy to
read and helped explain to millions of Americans why the
country was be

ing demonized while trying to do good. As a result of its

popularity and the

increasingly theoretical focus of his subsequent books,

Edward T. Hall is of

ten described as the founder of the field of intercultural

communication. As

he himself claimed, “The complete theory of culture as

communication is

new and has not been presented in one place before”

(1959/1973, p. 32). Intercultural communication was soon
recognized as a separate com

munication specialty. Subsequent researchers looked for,

and found,

many patterns of cultural variation. Generally these

variations were

based on analyses of face-to-face communication between

people from

different cultures. This interpersonal focus makes the

field a good fit with HCI. According

to Horton and Wohl’s (1956) theory of parasocial

communication and

Reeves and Nass’ (1996) “media equation,” our interactions

with “new me

dia” are just an extension of our face-to-face interactions

with one another.

Consequently, theories that deal with communication between

people from

different cultures should transfer directly to

communication between peo

ple from one culture using interfaces developed in another.

Hall has been followed by Edward C. Stewart and Milton J.
Bennett
(1972); John C. Condon and Fathi S. Yousef (1975); Geert
Hofstede (1980;

Hofstede & Bond, 1988); Harry C. Triandis (1975, 1983,

1995); William B.

Gudykunst (1987; Gudykunst et al., 1996); Fons Trompenaars

and Charles

Hampden-Turner (1998); and David A. Victor (1992). Next, I

review each of

these researchers and discuss ways to apply their theories

to interface design.

EDWARD T. HALL: CHRONEMICS, PROXEMICS, AND CONTEXT

Hall’s background preparing diplomats at the Foreign

Service Institute

shaped his work. He felt that assignees should be tested

for their psycholog

ical adaptability, trained in the appropriate language of

their new country,

and introduced to its formal culture. But, “training in

language, history,

government, and customs is only the first step. Of equal

importance is an in

troduction to the nonverbal language of the country”

(1959/1973, p. xiii).

Until people understood this “language of behavior,” they

could not begin

to be effective, and they needed to understand their own

“silent language”

to anticipate their reception. In The Silent Language, Hall

compared American cultural patterns

with those of Japan, Germany, France, Greece, the island of

Truk, and

the Middle East. He focused on two nonverbal features that

remain im
portant in intercultural communication theory: time
(chronemics) and

space (proxemics). In Beyond Culture (1976), he identified

another factor, which governs

people’s responses to verbal discourse, that is, context.

Hall’s discussion of chronemics is quite detailed. Time is
organized in

two ways. The first is the subjective division of time into

technical, formal,

and informal systems. The second is the connection of time

with activity. Technical time reflects the underlying
physical context of our experi

ence of time. The solstices, phases of the moon, sunrise,

and atomic de

cay can all be used to separate time into units. Formal

time is often based

on technical time. It consists of conventional systems of

measurement

that help people plan, schedule, and manage time. Formal

time also

leads to a sense of determinacy—the feeling that events are

linked and

causal. However, nothing requires a day to have 24 hours or

7 days to

make a week. Informal time is cultural and situational. It

exists side by side with techni

cal and formal time and frequently uses the same

vocabulary, but it involves

culturally different perceptions of the rate time passes.

For instance, if I say,

“I will see you in 5 minutes” in Germany, I need to hurry

to meet you. If I say

the same thing in Ecuador, you will not be concerned if I

don’t show up for
an hour. Informal time also deals with the significance of
events in the past,

present, and future. Hall noted that people in the Middle

East and South

Asia are more likely to ground decisions in historic events

than are people

in the United States. The connection between time and

activity led Hall to his notions of

monochronic and polychronic time. In monochronic societies

like the United

States, Canada, and northern Europe, people do one thing at

a time. They

prefer to work sequentially and try to resist

interruptions. In polychronic

societies like Latin America, the Middle East, and parts of

Asia, people work

on several things at once. This may mean promising to meet

different peo

ple in different places at the same time or working on two

or three projects

simultaneously. Monochronic time enhances determinacy but,

in poly

chronic time, things tend to “happen.” Hall’s second

contribution to intercultural theory is proxemics, the
social

use of space. Once again, he analyzed the expression of

this construct in dif

ferent cultures in terms of technical, formal, and informal

systems. Physical

constraints (mass, edge, volume) underlie technical systems

of space; the

process of architecture gives rise to various formal

systems; and people in

terpret informal patterns of proximity and arrangement to

determine sta

tus and group orientation. With Beyond Culture (1976), Hall

developed the idea of context. Although

this construct is also used in literary theory for

situating text and verbal dis

course, Hall wanted to highlight the importance of

nonverbal communica

tion. In high-context societies, features of the social

system shape meaning

through chronemics, proxemics, and other nonverbal

behaviors; a mes

sage’s actual text is secondary. In low-context societies,

the information is

explicitly expressed in the text of the message. The United

States tends to be low-context; Americans generally say ex

actly what they want you to understand. Many Asian

societies (Japan is the

classic example) are high-context; the message is inherent

in the occasion,

the physical setting, and the relationships between the

participants. People

within the culture know how to interpret each other’s

intent, but many

American business people have learned the hard way that

“yes” (said po

litely but unenthusiastically) really means “no.”

Hall as a Guide for Interface Development

How do Hall’s constructs relate to interface design? First,

his definition of

technical, formal, and informal systems can be mapped to

cultural percep

tions of applications and the Web. Computers also have

certain technical lim
its based on hardware and software, but the formal layout
of screens (and

sequences of screens) can still be read in different ways.

For instance, the “ar

chitecture” of a web page may differ in terms of color,

mass, and balance be

tween verbal and visual elements. Many American web sites

currently sport

subdued hues, arrange content in explicit hierarchies of

indentation, and

emphasize text over graphics. Web sites from other

countries appear visually

striking if they upset American expectations by using vivid

colors, an asym

metric arrangement, or more graphics than text. Appropriate

content may

also vary. Text that lacks formal historical grounding will

be unpersuasive to

people who rely on tradition to justify their actions.

Moreover, presentation

and content may interact. Pictures that are read

individually in one country

can become part of an unintentional collage of meaning in

another. Second, Hall’s notion of time has implications for
interaction. Different

applications may express monochronic or polychronic time. A

monochronic

application can support a user’s time sensitivity by

focusing narrowly on a

task or (somewhat paradoxically) by providing a wide range

of choices to re

duce the search hierarchy. A polychronic application can

present less struc
tured options or support incomplete sequences of tasks. For
instance, a user

may look up a train schedule on the Internet but expect to

return to the home

page (or go to the station) to complete the transaction.

Third, Hall’s notion of context is important because it
says that high-con

text users may not focus on the explicit verbal message;

thus, a developer

cannot depend on written text for clarity. Context may be

expressed in the

nonverbal elements of the screen, or context may be out of

the developer’s

control and inherent in the status of the software provider

or in local proce

dures that constrain the user’s access to the application.

As a result, a web site

for a low-context audience may not be appropriate for a

high-context one. A

site may lack credibility because it does not provide

branding or information

on the social structure of the sponsoring organization. A

site can also fail be

cause it does not recognize the social status of the user

or because it addresses

a high-status user with insufficient respect. Under certain

circumstances, ap

plications can actually be “too easy.” Lee (2000) found

that Korean subordi

nates prefer to fax messages to their superiors because

e-mail did not require

enough effort to show sufficient respect.

EDWARD C. STEWART AND MILTON J. BENNETT:

SUBJECTIVE AND OBJECTIVE CULTURE

Edward C. Stewart and Milton J. Bennett believed that

communication style

and cultural values arise from the process of human

perception. They wrote

their classic book on American Cultural Patterns

(1972/1979) to help U.S. citi

zens working abroad. Before such sojourners can properly

interpret mes

sages and events in other cultures, they need to recognize

how they have been

conditioned by their own. Like Hall, Stewart and Bennett

believed that

intercultural communicators must be culturally self-aware

to be effective. Culture has two fundamental aspects.
Objective culture is social and ma

terial, encompassing a society’s political and economic

system, aesthetics,

customs, art, architecture, and institutions. Subjective

culture is psychologi

cal, reflecting on society’s values, expectations, theories

of action, and pat

terns of thought. Objective culture is tangible (or at

least easily talked

about); subjective culture tends to be experienced

unconsciously. Cultural clashes arise from mismatches in
the process of human percep

tion. Subjective culture can be brought forward into

mindfulness by under

standing how the “deep mind” interprets its environment. An

understanding

of the different subjective levels of thought helps

identify the sources of cul

tural miscommunication. All people have the same senses,

but people from different places often

draw different conclusions from the same stimuli. Their

minds are condi

tioned to filter out or ascribe different messages to

similar sensory informa

tion. For instance, people may convert the same stimuli

into different

percepts. A small mammal that barks is food for one person,

a pet for an

other. As part of the process of distinguishing figure and

ground, different

people foreground different information for their idea of a

“dog.” At the next level of mindfulness, people develop
patterns of thinking. Per

sonal development within a family, community, and culture

helps shape their

cognitions. Memory, emotion, stories people have read, and

stories they

have heard from others bring to mind dinners in Shanghai or

walks in the

park. This is the stage at which “the mind extends its

mastery of objective re

ality by creating a cognitive, or subjective, reality where

meaning is assigned

and relationships elaborated” (Stewart & Bennett, 1972, p.

22). Patterns of

thinking apply to all types of sensory data—graphic and

nonverbal, as well as

text and language. Screen layout and expectations of

interaction are as sub

ject to the values and assumptions of culture as are words

and pictures. Complex symbol systems make up the deepest
layer of the Stewart and

Bennett model. These systems—pictorial, musical,

linguistic, and mathe

matical—give rise to metaphor, grammar, and semantics.

People learn

these systems within the larger societies that use them.

(They can also be

learned formally but it takes an insider to explain that =

“dog.”) So, how does language influence values? The two are
intertwined. Stewart and Bennett acknowledge a debt to
Benjamin Whorf and the

structural linguists. In analyzing patterns of language use

in the United

States, they tie grammar to behavior: • The

subject/predicate form of English is linked to a heightened
sense of causality (“it as implied agent”). • The use of
dichotomies and implied negative constructions is linked to
critical thinking. • Modifier/noun patterns are linked to a
preference for process and precision. These syntactical
elements enhance active, independent behaviors and

work against indirection or collaboration. People who use

such language

are likely to be part of or develop an individualistic

culture. Stewart and Bennett make a good case for looking
beyond the literal con

tent of visual imagery and language and grounding cultural

analysis in per

ception and the underlying grammar of symbol systems. More

than 30

years later, Richard E. Nisbett (2003) has identified

similar cognitive differ

ences that affect the way Westerners and Asians “read”

pictures—and com

puter screens.

Stewart and Bennett as a Guide for Interface Development

Most computer companies base their internationalization

efforts on the
premise that all users perceive computer systems in the
same way. If every

one engages the system at an equivalent semantic level,

then language lo

calization should be sufficient to address national

differences. At worst, a

company might be required to change certain metaphors, for

example, a

U.S. mailbox becomes a British post box. But Stewart and

Bennett’s distinction between objective and subjective

culture increases the possibility that companies must go

beyond surface text

and images. The grammar of the interface is hard to

recognize but may be

fundamental in cultivating trust and acceptance. Design

elements like

alignment, proximity, repetition, and contrast (Williams,

1994) affect the

massing of text on a page and the overall balance of the

screen. Such layout

features should not be ignored when localizing screens

because they can

contribute to intercultural miscommunication. For instance,

text that is

read from left to right trains users to look for important

information in the

top left and lower right quadrants of a screen. But

bidirectional text is read

from right to left and Asian scripts may be read from top
to bottom. So

where should the most important information be placed? When

translating

a process, the order of the graphics should be reversed.

HCI researchers need to identify similar cultural
differences in percep

tion and interpretation. A grammar of visual rhetoric, like

that of Gunther

Kress and Theo van Leeuwen (1996), will not be stable

across cultures. An

image of a young man gazing back from the screen may

promote engage

ment with the interface in Britain but intimidate a new

user in East Asia.

Cultural messages are embedded throughout the user

interface. Charles Kostelnick (1995) has also emphasized
the link between percep

tion and aesthetics. Interface designers tend to be so

embedded in the vi

sual language of modernism that they are incapable of

recognizing their

own bias. It is necessary to understand how people in

different cultures ex

perience an interface to address their hidden (and

seemingly irrational)

motivations. Such cultural knowledge must include their

artistic prefer

ences as well. Nancy Hoft (1996) championed the value of

Stewart and Bennett’s

“metamodel” for constructing narrowly targeted cultural

models for inter

face localization. Designers should first build a model of

their own culture

to have a standard for comparison. Then, thinking about

unconscious levels

of perception will help determine what parts of the

interface to focus on.

Fundamental differences in objective and subjective culture

will mandate
extensive change in localized screens and content. By using
a model of culture, Hoft promised that designers could
identify: • Global information that can be put into the
interface without re quiring future translation. • Cultural
bias in the existing application. • Parts of the
application that should be localized for a specific cul
ture. • Compelling cultural metaphors and cultural markers.
• Potential cultural problems. (In addition, cultural
models provide guidance for developing ap propriate methods
for usability evaluation.) However, Hoft might have taken
Stewart and Bennett’s perspective

even further. She talks about unspoken and unconscious

rules in her own

“Iceberg” metamodel and highlights Hall’s contributions to

decoding

nonverbal communication, but she does not fully apply these

insights her

self. For instance, she recommends identifying

international variables for

cultural models by surveys and observation, but the first

depends on

self-report and the second on an observer being able to

recognize the trig

gers for unconscious behavior.

JOHN C. CONDON AND FATHI S. YOUSEF:

EXTENDING THE VALUE ORIENTATIONS APPROACH

In 1975, John Condon and Fathi Yousef published An

Introduction to

Intercultural Communication. As their editor noted, this

was the first survey to

pull together the diverse themes associated with

interpersonal, nonverbal,

and speech communication; combine them with value

orientations; and

provide the reader with a single, unified explanation of

cultural differences

in social structure, language, and argumentation. (Even

Hall developed his

theories piecemeal.) Condon and Yousef cited everyone from

Hall, Bateson, and Goffman (for

interpersonal communication) to Birdwhistell (nonverbal

communication

and paralanguage) to Toulmin and Aristotle (rhetoric),

Sapir, Whorf, and

Chomsky (linguistics), and McLuhan (mass communication).

But the focus of

their book is their extension of Kluckhohn and Strodtbeck’s

five existential

questions. They defined 25 value orientations under the

headings of: • Self—individualism/interdependence, age,
sex, and activity. • Family—relational orientations,
authority, role behavior, and mobility. • Society—social
reciprocity, group membership, intermediaries, formality,
and property. • Human Nature—rationality, good and evil,
happiness, and mutability. • Nature—relationship of humans
to nature, ways of knowing nature, structure of nature, and
concept of time. • The Supernatural—relationship of humans
to the supernatural, meaning of life, providence, and
knowledge of the cosmic order. These categories are best
represented as a Venn diagram (see Fig. 4.1)

with overlapping spheres representing self, society, and

nature with the in

tersections representing family, human nature, and the

supernatural. Condon and Yousef’s (1975) definition of
cultural values is flexible. The

number of value orientations is not absolute. There could

be more—or fewer: Each of the value orientations … arises
from empirical data obtained from many cultures. The
greater danger is not that of hypothesizing irrelevant
categories but in not including enough. Since the scheme is
not intended to be exhaustive, this is not a major problem;
one is always free to add more or, if desired, to combine
categories or to ignore some entirely. (p. 58) Cultural
values develop early in life. Half of the values deal with
“what is
thought to be good” and the other half with “what is
thought to be true.”

Though standards of goodness are often easier to recognize

and articulate,

diverging beliefs about truth tend to have more profound

consequences for

intercultural communication. Opposing notions of reason,

pleasure,

change, and purpose may hinder individual perception and

prevent peo

ple from agreeing on common logics, aesthetics, strategies,

or goals.

Condon and Yousef as a Guide for Interface Development

The expansion of Kluckhohn and Strodtbeck’s value

orientations to social

structure, language, paralanguage, and argumentation moves

their theory

from existentialism into the realm of pragmatic

communication. Because

three options are associated with each value, a broad range

of behaviors is

described by the 75 heuristics. Values like interdependence

and the use of specialist intermediaries

could have a direct bearing on deciding whether to make

agents a promi

nent part of an interface. Preference for youth versus age,

women over men,

or individuals with tighter or looser relationships to

their families might in

fluence representations of users. The activity orientations

and relationship

of humans to nature (human domination, harmony, or the

supremacy of
nature) should be factored into any decision to develop
utilities to change

application defaults and system parameters. But the greater

importance of Condon and Yousef’s conceptual scheme

lies in their emphasis on rhetoric and persuasion. How

narrowly should an

application focus on specific tasks? Should an interface be

organized de

ductively or inductively to appeal to people from a

different culture? What

kind of information will be persuasive? Once developers

recognize that

their own patterns of rationality are not universal, there

is the possibility of

designing new interfaces with more appeal to those

voluntarily on the op

posite side of the digital divide. FIG. 4.1. Condon and

Yousef’s model of culture. Condon and Yousef also highlight
the importance of communication

roles. People communicate with images of each other: You

communicate

with an image of me and I communicate with an image of you.

In addition,

my image of “me” also influences my expectations for

interaction with your

image of “you.” Walter Ong (1975) once said that an

author’s audience is al

ways a fiction. Interface designers are no less guilty of

inventing their users. In fact, our projections are often
hyperpersonal. Joe Walther (1996)

found that unacquainted people communicating at a distance

make more

extreme, and generally more positive, judgments of one

another than peo
ple communicating face-to-face. We contribute to the
digital divide when

we have unrealistic and uncritical expectations that

everyone who uses a

computer must be much like us. Unfortunately, when they are

not, their re

actions can become overly negative. In general, bad

experiences diminish

willingness to try new software. Such problems take on a

geopolitical di

mension because computer interfaces and Internet web sites

function as

communication partners in themselves. When television was

new, Horton

and Wohl (1956) described people’s strong engagement with

TV stars and

other media figures. Reeves and Nass (1996) found no

difference in classic

social psychology experiments rerun with humans and a

computer surro

gate. People who interact with our software feel that they
are interacting

with our culture. Poor usability may be seen as

confirmation of a vendor’s

(and a culture’s) lack of respect. Thus, our interpretation

of users from other countries is often based on

unrealistic images of them. And their (often negative)

projections of us,

based on their interactions with our products, may be

equally unfounded.

With the best of intentions, people in HCI tend to

underestimate the effects

of culture even as we struggle to close the digital divide.

GEERT HOFSTEDE: THE DURABILITY

OF THE “SOFTWARE OF THE MIND”

Hofstede’s (1980) theory of universal cultural dimensions

has become the

best known, and probably most applied, intercultural

communication the

ory. Various researchers have applied it to leadership,

motivation, compen

sation, and teamwork within domestic and multinational

firms, and Nancy

Hoft (1996) used it in her article on developing a cultural

model. Hofstede based his work on a series of surveys from
1968 to 1972 with

IBM employees from 72 national subsidiaries; after factor

analysis, four di

mensions explained half the differences between the 116,000

respondents.

Not only were the factors supposed to be universal, but

Hofstede claimed

that they were highly resistant to change because people

learned them as

children. Adults might be exposed to opposing values and

superficially

adapt, but they could never erase this “software of the

mind.” Aaron Marcus (chap. 3, this volume) thoroughly
describes these di

mensions: • Power Distance: the extent to which less

powerful people expect and accept that power is distributed
unequally. • Individualism vs. Collectivism: the extent to
which people are inte grated into tight social networks. •
Masculinity vs. Femininity: the relative desirability of
material success versus quality of life and of assertive
versus modest behavior. • Uncertainty Avoidance: the extent
to which people tolerate ambigu ity and risk or feel
threatened by change. In addition, Marcus demonstrates how
they can be mapped to a model of
the user interface to generate culturally appropriate
designs. (See also

Marcus & Gould, 2000.) Hofstede’s dimensions have become a

touchstone for comparisons of na

tional differences because of the size and scope of the

original survey, but

they have also been attacked. Three major criticisms have

been raised

against Hofstede’s four factors: 1. IBM employees are not

representative of any national culture. As employees of a
multinational corporation with a conservative (in those
days) corporate culture, they should not have been used to
sample national traits. 2. The factors are not stable.
Cultures change over time and even IBM is not the company
it was in 1972 when Hofstede finished his research. 3.
Hofstede’s survey was itself biased. Because he developed
and tested his questions in Europe and the United States,
and adminis tered them in English, they were permeated with
Western values. Hofstede acknowledged that IBMers are not
representative but rejected

the implication that they can not be used to identify

cultural values. Instead,

he claimed that their very bias makes them a functionally

equivalent sam

ple. That is, IBM’s distinct workplace culture gives

employees so much in

common that any differences between them must be the result

of differ

ences in national culture. Hofstede also rejected the

criticism that his factors might change over

time. Instead, he emphasized the stability of childhood

acculturation. Re

cent work by Hofstede fellow and colleague, Marieke de

Mooij (2001), reit

erates this point. But Hofstede did admit that his country
rankings omit
some nations (e.g., People’s Republic of China) that have
become more im

portant in the three decades since he did his research.

Hofstede also accepted that his original survey on employee
attitudes

does have a Western bias. Michael Bond (a Canadian-born

professor in

Hong Kong) convinced him to administer the Chinese Value

Survey,

which had been developed by Taiwanese and Hong Kong

researchers in

the early 1980s. After reviewing the factor analysis,

Hofstede and Bond

(1988) found a fifth cultural dimension: longversus

short-term orienta

tion (also known as Confucian dynamism, the acceptance of

traditional

Confucian values). The Chinese Value Survey supported the

validity (and universality) of

power distance and individualism and collectivism. However,

it modified

the definition of masculinity to include more “chivalric”

values like pa

tience, courtesy, kindliness, and compassion. This

redefinition highlights a major problem with the third of

Hofstede’s dimensions. Notions of masculine and feminine

behavior vary

widely and are themselves culturally determined. A more

useful way to in

terpret this dimension might be to examine the extent that

people are

bound by general social roles. In masculine societies, it

is difficult for men

to look after children, but it is also difficult for anyone

to change their reli

gious or political affiliation. In feminine societies,

people are less bound

to social roles in general. More recent research repeats

these criticisms of Hofstede’s methodology

while discovering surprising confirmation for his strongest

dimensions. For

instance, Shalom Schwartz (1994) published an alternative

set of value di

mensions developed from a new sample of 41 cultural groups

in 38 nations. As shown in Fig. 4.2, his factor analysis
juxtaposes two qualities found in

individualistic societies—Intellectual Autonomy

(self-direction) and Affec

tive Autonomy (stimulation and hedonism)—with the

Conservatism (con

cern for security, conformity, and tradition) found in

homogeneous

(collectivist) societies. Further, Schwartz was able to

identify factors associ

ated with Self-Enhancement that contrasted with values

promoting Self

Transcendence. The two Self-Enhancement factors are

Hierarchy (the legitimacy of

power hierarchies) and Mastery (self-assertion,

self-promotion, and active

control of the social environment)—factors similar to

Hofstede’s Power Dis

tance and Masculinity. The two Self-Transcendence factors

are Egalitarian

Commitment (the motivation for altruism and volunteer

service in individ

ualistic societies) and Harmony (with nature)—factors that

expand
Hofstede’s notion of Femininity. Schwartz better explains
why individualists often act collectively, for in

stance, how Sierra Club members in Los Angeles and

subsistence farmers

in traditional African villages can share a common

environmental per

spective. But, overall, Hofstede’s dimensions hold up

surprisingly well.

Hofstede as a Guide for Interface Development

Hofstede’s tables of cultural differences provide a

methodology for antici

pating cultural differences. Each of the 50 countries (and

three regions) in

his original survey is ranked in terms of the original four

work orientations,

with another 20 listed for Confucian Dynamism. Thus, if you

are a Cana

dian developing a web site for South Koreans, you can

determine that your

audience will have higher power distance, lower

individualism, greater

masculinity, higher uncertainty avoidance, and a more

long-term orienta

tion than a domestic audience. When you associate interface

elements with these dimensions, you may see

some overlap. Support for mastery (masculinity) has much in

common with

allowing people to learn sites by trial and error (low

uncertainty avoidance). FIG. 4.2. Shalom Schwartz’s
culture-level value structure. Hofstede admits that most of
his factors are correlated with one or two

others, but considers this redundancy helpful, not

problematic. Not all cul
tural differences are equally critical for each country.
Only significantly dif

ferent cultural orientations warrant attention, though each

can percolate

throughout the interface. For example, Canada and South

Korea are quite different, and a Cana

dian designer might need to develop a radically different

site for a Korean

audience. But a German designer modifying a domestic web

site for Dutch

customers would discover that the two countries share

similar values for

power distance, individualism, and uncertainty avoidance;

only masculinity

(Germany) and femininity (the Netherlands) are ranked far

apart. As a re

sult, the German could focus on the representation of

gender and social

roles and leave the rest of the site alone.

HARRY C. TRIANDIS: ADDING CULTURE

TO SOCIAL PSYCHOLOGY

In the same decade that Hofstede started researching

culture, Harry C.

Triandis began to consider its effects on social

psychology. He first exam

ined the interplay between norms, attitudes, and behaviors.

By the end of

the 1960s and early 1970s, he was looking for differences

in group identity

to explain people’s actions—analyzing subjective (as

opposed to material)

culture in Greece, race in the United States, and social

distance between
American, German, and Japanese students. He linked the
influence of

“in-groups” and “out-groups” to the issue of culture; in

fact, he saw culture

as the ultimate group. At the time, social psychology was

coming to realize that many of its topics

were culture-bound and had little explanatory value.

Studies replicated with

people from different ethnic and national backgrounds

showed radically dif

ferent results. Variables often measured trivial amounts of

difference. In 1975, Triandis wrote an influential article
that reviewed these prob

lems and urged the field to focus on higher order

constructs, like associa

tion and dissociation, whose expression could be compared

across cultures.

He acknowledged that intercultural studies would be

difficult to develop

because they require broad-based collaboration, multiple

methods, and

customized test materials for each culture, but he

encouraged people to

work cross-culturally. Shortly thereafter, Triandis

promoted models of universal social behav

iors: “Eventually social psychological theories will be

computer programs”

(1978, p. 14). In another article, “Dimensions of Cultural

Variation as Pa

rameters of Organizational Theories,” he urged social

scientists to apply

cultural values (most from intercultural communication) to

then-current
theories of productivity, job satisfaction, and leadership:
Organizational theorists naturally develop theories about
the part of the so cial world that they know and
understand. As a result, many such theories are
inappropriate and inadequate when applied to other
cultures. We create or ganizations that work well for us,
and for people like us. But these organiza tions are
usually a failure for people with a different perspective.…
What is likely to be most effective in one culture is often
different from what is most effective in another. (1983,
pp. 140–141, 165). Triandis’ promotion of these research
goals makes him foundational to

the field of cross-cultural psychology. In 1976, he was

elected president of

the International Association of Cross-Cultural Psychology

(itself founded

only 4 years earlier.) In the 1980s, he edited the

multivolume Handbook of

Cross-Cultural Psychology. Over the last 15 years, he has

taken his own advice and focused on a sin

gle cultural syndrome: individualism and collectivism.

(This was one of

Kluckhohn and Strodtbeck’s original value orientations and

is the most

common cultural dimension in intercultural communication

theory.)

Drawing on philosophy, political science, religion, and

history, as well as

communication, anthropology, and psychology, Triandis has

examined in

dividualism and collectivism in depth. Much of his work

considers such puz

zling aspects as the strength of in-groups, the paradox of

individualists in

collectivist societies (and vice versa), and interactions

with power distance. In reconceptualizing the dimension,
Triandis (1988) asked himself why
people in collectivist societies are disproportionately
influenced by their

in-groups. He decided that people in individualistic

societies have greater

freedom of action and the opportunity to belong to more

groups than peo

ple in collectivist societies. As a result, no single group

dominates them,

though they often spend more time overall conforming to

group norms. By

contrast, people in collectivist societies have fewer

choices and belong to

fewer groups. Because their societies are also more

homogeneous, these

groups (family, religion, and state) tend to hold similar

values and reinforce

one another. Consequently, groups exert more control over

people’s ac

tions. People in collectivist societies learn to be

cautious and gather a great

deal of information about new individuals and situations

because group re

lationships make such large demands on their time, energy,

and autonomy.

However, outside their groups, people can act quite

independently. A year later, noting that people function on
an individual (not cultural)

level, Triandis (1989) incorporated a theory of social

identity into the di

mension. People growing up in different cultures have more

or less

well-developed private, public, and collective selves that

they rely on for

appropriate social responses. In different situations, one

self usually dom

inates and influences the development of personal

relationships and pat

terns of communication. The individual level of

individualism is idiocentric

and the individual level of collectivism, allocentric.

These individual-level

values may contrast with the overall culture-level values

assigned to a spe

cific country. Triandis (1995) even found a way to redefine

power distance as a subor

dinate dimension of individualism and collectivism

(building on Hofstede’s

recognition that the two strongly covary.) Horizontal (low

power distance) in

dividualism is linked to the freedom to define and pursue

one’s goals, and

vertical (high power distance) individualism to competition

and accomplish

ment. Horizontal collectivism is tied to interdependence,

in-groups, sharing,

and consensus, whereas vertical collectivism is associated

with service and the

internalization of group norms. Each category has its

negative side. With horizontal individualism can

come loneliness; with vertical individualism, personal

stress. Horizontal

collectivism can lead to distrust of people outside the

group; vertical collec

tivism, to the emergence of excessive bureaucracy and rigid

observance of

rules. People are most satisfied with their lives when

there is a good fit be
tween their personal self-concepts and their social
environment.

Triandis as a Guide for Interface Development

Triandis’ skepticism about the ability of Western social

science to identify

universal principles of human behavior or universal methods

of human re

search is a critique HCI should consider. All science is

embedded in cul

ture—computer science and interface design no less than

social psychology.

Triandis urged his own field to embrace collaboration and

multi-site testing

as the solution. This methodology is one that HCI should

also adopt. In addition, Triandis strengthened the
explanatory value of individual

ism and collectivism. His explanation of the power of

in-groups suggests

that interface designers must work harder to establish

trust with new users

from collectivist cultures. His division of individualism

and collectivism into

multiple levels (individual, organizational, and societal)

explains how a

person born in Japan can be more independent than one born

in the

United States. (A society may be predominantly

individualistic or collectiv

ist, but individuals within that society can vary.) His

theory of social identity

acknowledges that a person (or culture) can be

simultaneously individualis

tic and collectivist. These refinements make individualism

and collectivism elements in a
personal repertoire and improves our understanding of them
as design

heuristics. Moreover, it improves them as experimental

variables. Cul

ture-level dimensions are particularly problematic in

usability testing with

the small number of subjects in the average usability lab.

Many studies

based on Hofstede’s constructs have returned ambiguous

results because

the researchers assumed that everyone in a culture shares

the same cultural

orientation. Individual-level variables provide better

information about

their users and more realistically reflect the range of

beliefs within cultures. Hazel Rose Markus and Shinobu
Kitayama (1991) incorporated

Triandis’ notions into their theory of independent

(individualistic) and in

terdependent (collectivist) self-construals. These personal

attitudes are eas

ily measured with Singelis’ (1994) Self-Construal Scale

(since modified) by

Cross, Bacon, & Morris, 2000) and make a better research

variable for HCI

than assignments based on Hofstede’s country rankings.

WILLIAM B. GUDYKUNST: BUILDING A UNIFIED

THEORY OF COMMUNICATION

William B. Gudykunst’s work in the field of communication

parallels Harry

Triandis’ advocacy of a cross-cultural perspective for

social psychology.
Like Triandis, Gudykunst first saw cultural differences as
fundamental but

now works to reconcile them within higher level constructs.

He too has ex

plored the importance of social identity. However, unlike

Triandis, much of

his work is based on the analysis of specific communication

contexts. In

stead of focusing on a single cultural syndrome, Gudykunst

has investigated

the effect of culture on initial encounters and

relationship building. Re

cently, he has also looked at communication style.

Gudykunst’s practical experience as an Intercultural
Relations Specialist

for the U.S. Navy in the early 1970s first helped shape his
career. “While

conducting intercultural training in Japan, I thought

intercultural commu

nication (i.e., communication between people from different

cultures) was

different from intracultural communication (i.e.,

communication with

members of our own culture)” (1994, p. ix). Gudykunst’s

first papers show this training bias. He examined the way

people from different cultures encounter and learn to trust

one another by

applying Hall’s and Hofstede’s dimensions to classic

communication pro

cesses like homophily (similarity) and uncertainty

reduction (1983a,

1983b). But by 1984, he and Young Yun Kim concluded that

intercultural

and intracultural communication vary along a single

continuum from total

strangeness to total familiarity. Initial interactions are

particularly critical when people from different

cultures meet because they are likely to have less in

common. They begin by

disclosing information about themselves and asking

questions. These strat

egies help develop attraction by identifying common values,

discovering

common relationships, reducing uncertainty, and increasing

attributional

confidence. The process of interaction becomes an

opportunity for coordi

nation and affiliation. At first, social stereotypes and

information about group memberships

are particularly important. As the relationship develops,

people make more

predictions based on knowledge of each other’s personal

characteristics.

Eventually, the need for similarity decreases (old friends

tolerate greater

differences). This pattern is consistent across cultures,

but the information

requirements differ. People from high-context cultures ask

more questions

about family and background; people from low-context

cultures are more

immediately sociable, expressive, and direct. Gudykunst and

Tsukasa Nishida (1986) found a correlation between

Hofstede’s dimensions and levels of intimacy in six types

of relationship.

Gudykunst, Elizabeth Chua, and Alisa J. Gray (1987)

confirmed that people
look for greater evidence of similarity when beginning new
intercultural

friendships. In 1988, Gudykunst pulled together his work on

individual and

group representations and relationship development and

published it as Anx

iety/Uncertainty Management (AUM) theory (updated in 1993

and 1995.) In addition to collapsing interand intracultural
communication,

Gudykunst can also be credited with introducing new

theories from Europe

into interpersonal communication. In 1986, he edited

Intergroup Communi

cation and wrote an introduction with Tae-Seop Lim: This

book is intended to bring together the study of intergroup
relations in Europe and the study of interpersonal
communication in the United States. Both areas have
developed separately, with little borrowing of concepts or
theories. There are advantages to both areas, however, in
their integration. It is through communication that
intergroup relations are established and communication
plays a central role in their improvement or deterioration.
The intergroup relations perspective brings a focus upon
group membership and its influence on the study of
interpersonal communication that virtually has been ignored
to date in the United States. (1986, p. 1) Individuals come
together in groups for a variety of reasons (perceptions

of similarity, common interests, interdependence, or the

desire for interac

tion.) Members learn to identify with one another and this

social identity

becomes part of their self-concept. People can belong to

many groups and

develop multiple social identities. Personal identity makes

up the remain

der of their self-concept. When social identity is

dominant, intergroup communication takes

place. When personal identity is dominant, interpersonal

communication

takes place. Situations like labor negotiations activate

both types of identity;

anonymous encounters on the street activate neither.

Gudykunst and

Tae-Seop Lim’s (1986) model (see Fig. 4.3) highlights the

influence of com

munication context on identity. The contact situation often

determines

whether social or personal identity is made salient. The

final theme in Gudykunst’s work has been his research on
communi

cation style. He used Hall’s theory of highand low-context

communication

in his work on initial interactions and relationship

development, but he has

also looked for the predictors of style. In 1988,

Gudykunst, Stella

Ting-Toomey, and Chua developed a model of interpersonal

communica

tion. This model shows how culture indirectly mediates

communication

style and content through social cognitive processes,

situational factors, af

fect, and personal habits. A few years later (1996), the

model was simplified by incorporating

Markus and Kitayama’s self construals (see Fig. 4.4).

Gudykunst and Ting-Toomey collaborated with Yuko Matsumoto,

Nishida, Kwangsi Kim, and Sam Heyman to survey highand

low-context

styles in the United States, Japan, Korea, and Australia.

When defined at

the level of culture, individualism and collectivism only

partially predicted
communication style; however, when defined on an individual
level (inde

pendent vs. interdependent self construals plus personal

values), the di

mension fully predicted communication style. People who

considered

themselves unique and independent used a low-context style,

whereas peo

ple who considered themselves part of larger social

relationships used a

high-context style. FIG. 4.3. Interpersonal and intergroup

communication.

Gudykunst as a Guide for Interface Development

Gudykunst’s research on initial encounters, identity, and

communica

tion style has direct relevance for HCI. In addition, it

suggests some in

teresting intersections between intercultural and

computer-mediated

communication. “Meeting” new software (or a new web site)

invokes the same process of

prediction, attribution, and uncertainty reduction that

people use when

meeting a stranger. Therefore, interface designers need to

provide appro

priate relationship-building information because people

will make these

attributions even when the uncertainty reduction process is

unacknowl

edged. Group memberships are particularly important

information for

people who use a high-context style; personal

expressiveness is particularly
persuasive for people who prefer low-context communication.
Global access makes it likely that Web designers must
include both types

of information. The Web also provides some interesting

technical opportu

nities for enhancing subsequent interactions. Cookies

permit personaliza

tion of a site visited before. Usability, task orientation,

and efficiency can be

optimized to show consideration to low-context

communicators.

Interactivity by itself will enhance liking, that is, as

long as the experience is

positive. (See Rafaeli, 1988; Rafaeli & Sudweeks, 1998.)

But new users may

take different amounts of time to develop a sense of trust.

People from

high-context cultures are likely to need more information

from the inter

face before they commit themselves; on the plus side, they

are probably will

ing to explore more of the web site or application to find

it. FIG. 4.4. Model of communication styles. Gudykunst and
Lim’s discussion of the salience of social and personal

identity also has some interesting implications for HCI.

Different contexts

of use, for example, learning online presentation software

at work, buying

printer paper from the cheapest supplier on the web, or

picking a university

from a distance, may make different parts of one’s identity

salient. People in

a work environment make the effort to learn a new

application because it is
a necessary part of their corporate identity. People
engaged in an online

transaction that involves little risk and no status or

sense of participation

simply send their order to the lowest vendor. But people

researching an or

ganization that requires a long-term commitment (or carries

high risk) may

need different information, depending on whether their

preference is for

highor low-context communication. For instance, an Asian

technician de

ciding to come to Europe for a training program could

require more text on

the history and management of the institute than a North

American. But

the same Asian would not need that level of information

when picking a ski

ing vacation. Considering the effects of an application on

social versus per

sonal identity should become a new design heuristic.

Several computer-mediated communication theories focus on
the same

issues. Initial interactions and social identity come

together in Martin Lea

and Russell Spears’ (1991) SIDE model (Social Identity

model of

Deindividuation Effects). This model suggests that people

working to

gether across computer networks make attributions about one

another de

spite the lack of nonverbal cues present in face-to-face

meetings. When
their mutual group identity is enhanced and made salient,
they will collabo

rate productively; when their individual identities are

enhanced, they may

“act out” (deindividuate) and ignore the needs of others.

Joseph Walther

and Judy Burgoon (1992) found that people in

computer-mediated teams

achieve the same level of liking and support as people in

face-to-face

groups, but it takes more time. Due to Walther’s

hyperpersonal effect, peo

ple may even like one another more because their

impressions are based on

positive projections and stereotypes. In deciding what kind

of information

to front-load on a home page, web designers need to

consider whether peo

ple will be visiting the site once or returning repeatedly

and whether they

need to gather information to personalize subsequent

interactions. Finally, Gudykunst’s work on communication
style with Matsumoto,

Ting-Toomey, Nishida, Kim, and Heyman (1996) has direct

application to

interface development. Eight factors emerged when

respondents from Aus

tralia, Japan, Korea, and the United States were asked to

rate 158 commu

nication behaviors related to high- and low-context

communication: 1. Certainty in inferring others’ meanings,
intentions, needs, and feelings (associated with
individualism because it reflects a con scious strategy).
2. Use of indirect or ambiguous communication
(collectivist). 3. Interpersonal sensitivity, tact, and
preference for harmony (collec tivist). 4. Use of dramatic
communication (individualist). 5. Openly expressed use of
feelings to guide behavior (individualist). 6. Personal
openness and willingness to initiate communication with
others (individualist). 7. Precision in communication
(individualist). 8. Positive perceptions of silence in
communication (collectivist). Some of these styles reflect
the way the respondents process informa

tion; others, the way they expect information to be

delivered. The eight

factors were correlated with independent and interdependent

self

construals and with personal values derived from

individualistic and col

lectivist cultures. Because people can have both

independent and inter

dependent self construals, the authors note that specific

situations may

activate opposite self-concepts. For instance, people might

use a differ

ent style communicating with an in-group than with an

out-group they

had no interest in joining. Although limited to four

cultures, this study provides direction on the

most important features to emphasize in developing

interface text and

graphics for people with individualist and collectivist

values.

FONS TROMPENAARS: CULTURE HITS THE BEST-SELLER LIST

Fons Trompenaars’ best seller, Riding the Waves of Culture,

was first pub

lished in 1994; the second edition, with Charles

Hampden-Turner as coau

thor, came out 4 years later. The book made Trompenaars an

international

management guru. (According to his web site, he was listed

as a top man

agement consultant in 1999, along with Michael Porter, Tom

Peters, and

Edward de Bono.) His discussion of cultural differences as

potential busi

ness assets began as a critique of existing Western

management theories. Trompenaars’ career has been in
industry—working for Shell’s Person

nel Department in the 1980s and founding his own consulting

firm (now

Trompenaars Hampden-Turner) in 1989. His practice

specializes in

cross-cultural mergers, executive coaching, vision

statements, and training.

Clients fill out surveys on their cultural orientations and

cultural compe

tence, and Trompenaars has accumulated a database of over

55,000 re

sponses to support his analysis of culture. For

Trompenaars, management theory is not culture-neutral.
Business

practices in different countries can look the same but be

based on vastly dif

ferent assumptions: The issue is not whether a hierarchy in

the Netherlands has six levels, as does a similar company
in Singapore, but what the hierarchy and those levels mean
to the Dutch and Singaporeans. Where the meaning is totally
different, for ex ample, a “chain of command” versus “a
family,” then human-resources poli cies developed to
implement the first will seriously miscommunicate in the
latter context. (Trompenaars & Hampden-Turner, 1998, p. 6.)
Culture defines the way each society solves the problems of
its social,

temporal, and physical environment. Trompenaars uses an

analytic frame

work of seven cultural dimensions: the five pattern

variables of Talcott Par
sons (1951), which explain how people relate to one another
within a social

system, and the cultural dimensions for time and the

environment from

Kluckhohn and Strodtbeck (1961) and Edward T. Hall

(1959/1973). The seven dimensions are: 1. Universalism vs.
Particularism: the extent to which people base their
actions on rules or relationships. In universalist
societies, people believe it is possible to define
universal principles which will be “true” in all
situations. Governments are based on the rule of law;
economics is a social “science”; and people conduct
business as a rational enterprise. In particularistic
societies, people temper their actions to fit the social
situation. They may support legal principles in theory but
bend the truth in a court of law to help a friend; the term
social science is an oxymoron; and business is a creative
art. Universalists base their business dealings on
contracts; particularists renegotiate their business
relationships when personnel change. Abstract reasoning and
statistics persuade universalists; the practical effects of
policies persuade particularists. 2. Individualism vs.
Communitarianism: the extent to which people act on the
basis of their own needs or the needs of their social
group. Individ ualists believe that society benefits when
all have the freedom to act in their own self-interest.
Communitarians believe that society benefits when people
discipline themselves and consider the needs of others. In
dividualists like to work (and play) alone; they expect
credit for the per sonal responsibility they take on.
Communitarians prefer to work closely with others; they may
vacation in a group; they expect credit to be given to the
project team as a whole (not just the team leaders.)
Personal incen tives (like cash bonuses) motivate
individualists; higher-order principles (like quality) also
motivate communitarians. 3. Neutral vs. Expressive
Communication Styles: the extent to which people hide or
display their emotions. At a fundamental level, people
everywhere feel the same basic emotions (see Gudykunst and
Ting-Toomey, 1988), but they do not express them in the
same way. Peo ple in neutral cultures restrain their
feelings; people in affective cul tures openly show them.
When someone in a neutral culture loses control, their loss
of composure is shocking; when someone in an affec tive
culture feels something extreme, their depth of feeling may
not be noticeable. When someone in a neutral culture uses
irony, the humor is likely to be lost on people who have a
passionate communication style; in fact, the ironist may
even be thought deceitful. Different cultures sep arate
their emotions from objective, logical decision-making or
display their emotions as a legitimate part of the process.
In face-to-face com munication, each style affects eye
contact, rate of speech, tone of voice, facial animation,
use of silence, and overall preference for nonverbal or
verbal communication. 4. Specific vs. Diffuse
Relationships: the extent to which relationships with other
people are limited or unconditional. People in specific
societ ies may consider coworkers close friends but never
see them out of the of fice; people in diffuse societies
become close to fewer people at work but include them in
their private life. These different expectations of inti
macy reflect culturally different notions of public and
private space and of the costs of friendship. Consequently,
people in specific cultures often feel free to criticize
friends because the friendship doesn’t involve a deep level
of intimacy. People in diffuse cultures often seem indirect
(even evasive) because they know that their words have
greater emotional significance. They also take longer to
make friends because the cost is unconditional acceptance.
Once that relationship has been established, the relative
social roles tend to stay the same forever—if you become
friendly with a teacher, that person can give you advice
even after you establish yourself successfully. This
dimension influences both the depth and breadth of
emotional involvement and is closely related to commu
nication style. There is also a close link between specific
and diffuse rela tionships and several of Hofstede’s
dimensions. 5. Achievement vs. Ascription: the extent to
which status is accorded on the basis of personal
performance or social role. Cultures with an achievement
orientation value performance and action; doing some thing
is better than doing nothing. Cultures where status is
ascribed on the basis of gender, age, family connections,
or profession assume that these characteristics are
important indications of effectiveness; doing
nothing—waiting for opportunities to present themselves—is
better than acting in haste. Achievement cultures value
youth for its own sake and are often informal; people
expect to change jobs and employers during the course of
their career. Ascription cultures value experience and
maturity; people are loyal to their company and demonstrate
re spect for the organization by absolute respect for
superiors. Networks in achievement cultures are personal;
in ascription cultures, they dupli cate the official power
structure. Titles and degrees are important in as cription
cultures since everyone needs to know the sources of an
individual’s status. 6. Orientation to Time: the time
horizon (past, present, or future) and preference for
sequential or synchronic activities. Trompenaars com bines
the time dimensions of both Kluckhohn and Strodtbeck and Ed
ward T. Hall. People view their actions in historical
terms, in the “here and now,” or in terms of the future. As
a result, they may be resistant to change, follow current
trends, or defer immediate gratification for long-term
advantage. An orientation to the past or distant future is
of ten found in ascription cultures; an orientation to the
present (or near future) is found in achievement ones.
Ascription cultures may see time as circular; achievement
cultures see it as linear. Different groups also use time
in different ways. Sequential cultures make schedules and
do one thing at a time (Hall’s monochronic cultures).
Synchronic cultures are open to the situation; they let
events develop their own momentum and often end up juggling
multiple activities at the same time (Hall’s polychronic
cultures). 7. Orientation to Nature: the extent to which
people feel they can control, or are controlled by, nature
and the social environment. Like Kluckhohn and Strodtbeck,
Trompenaars believes that people see themselves as masters
of nature or living in harmony with nature (and
occasionally mastered by it.) These contrasting positions
lead to varying appreciations of human agency and the
natural world. People who are inner-directed believe they
can control external events and are attracted to technology
and stories of human achievements; people who are
outer-directed value nature for its own sake and prefer to
work within the status quo. When plans fail, inner-directed
people may react aggressively and try again, while
outer-directed people bide their time and wait for new
opportunities. A control orientation can affect a culture’s
willingness to change and innovate while an orientation to
harmony may promote flex ibility and service to others.
Because of the popularity of Riding the Waves of Culture,
Trompenaars’

seven cultural differences are well-known. The framework is

very appropri

ate for intercultural studies but does have one

problem—unlike Hofstede’s

dimensions, it has been applied to (not derived from) a

statistical analysis of

Trompenaars’ survey data. Trompenaars does not summarize

his data by culture; in addition, coun

tries are rank ordered for individual questions but not for
each dimension

overall. As a result, the distinctions between many of the

dimensions are un

clear: for instance, one question on neutral or expressive

communication

style seems similar to another on specific or diffuse

relationships. Individu

alism and collectivism, universalism and particularism, and

achievement

and ascription all appear related. This makes it difficult

to use

Trompenaars’ set of cultural values experimentally.

Trompenaars as a Guide for Interface Development

Nonetheless, Trompenaars’ seven cultural differences make

excellent

design heuristics—especially when combined with principles

of graphic

design like the “grammar” of Kress and van Leeuwen (1996).

Should the

mode of address in messages, text, and graphics be

especially polite and

reassuring or more direct (and potentially brutal)? How

must trust do

people need to develop before they invest hours of work in

an applica

tion? Do they want to be in control or can a wizard help

them achieve

their goals? It is easy to speculate on specific effects,

particularly for the Web. For in

stance, sites for people with universalistic values might

focus on products,

functionality, and price, whereas sites for people with

particularistic values
might focus on business relationships and service.

Universalism Particularism

Overall focus on products, function

ality, and price Overall focus on business relationships

and service

Little information about company Strong branding

Price comparisons Customer testimonials

Formal rhetorical style Familiar rhetorical style (2nd

person ad dress)

Sharp focus images that highlight

the product Soft focus images that highlight the cus tomer

Symmetric templates Asymmetric templates

Consistent interaction style Use of different interaction

models to achieve different goals In selling to a
universalist market, a company like Honda might use the

Web to emphasize the innovative engineering used in its new

gas-hybrid

cars while emphasizing its service and reputation in

selling to a

particularist market. Sites for people from individualistic

cultures might be designed to en

hance individual performance, whereas those for people from

communi

tarian cultures might emphasize group goals and behavior.

Individualism Communitarianism

Focus on individual self-interest and

enhanced performance Focus on collaborative behavior and

group communication

Text reinforces self-image, status, or

material success Text discusses benefits to others

Emphasis on innovation Emphasis on reliability and quality

Images of individuals using the prod

uct (or the product alone) Images of groups of people using

the product (product may never be dis played)

Images of people at play (or working

in informal settings) Images of formally dressed people

work ing with others When appealing to individualists, a
food company like Starbucks might

emphasize the rarity, taste, and price of its shade-grown

coffee but focus on

the benefits to the environment when pitching to

communitarians. Sites targeted at people with a neutral
communication style might use a

more subdued color palette, reserved rhetorical style, and

larger propor

tion of text to graphics than sites targeted at people with

an affective style.

Neutral Communication Style Expressive Communication Style

Subdued color palette Bright, expressive colors

Reserved rhetorical style Exaggerated claims, elaborate

words, and emphatic punctuation

More text than graphics Equal or larger amount of graphics

to text

Rhetorical strategy based on objec

tive, logical analysis Rhetoric anchored in personal

stories or narratives

Uncluttered layout Dense layout

Long- and middle-range images Close-up images

People in images make reserved eye

contact Strong eye contact or none at all

Use of irony and word play Use of comic images and
animations A cruise line selling in northern Europe (where
most people use a neutral

communication style) might show an overall image of one of

its boats and em

phasize its ports of call. The same line might show the
sumptuous buffet (or a

collage of its tourist facilities) when selling to a more

expressive U.S. audience. People from specific cultures may
prefer web sites that are task-oriented

and direct, whereas diffuse cultures may require more

information (much

of it relational) even for nominally simple web

transactions.

Specific Relationships Diffuse Relationships

Direct rhetorical style Indirect rhetorical style

Evocation of personal satisfaction Evocation of public

principles

Images of happy, smiling people Images of reserved, serious

people

Specific to general logic General to specific logic A U.S.

university recruiting students from a specific culture
might talk

about the starting salaries of its graduates and show

groups of happy students

(at play or graduation). Students from a diffuse culture

would look for infor

mation about the prestige of its programs and graduates and

be persuaded to

come by images of students hard at work on the best new

equipment. People from achievement cultures are likely to
be persuaded by sites that

help you “do something” (buy a product or recommend a

book), whereas
people from ascription cultures may require information
about who is do

ing the selling before they will act. It is not that these
users don’t want to use

the Web productively—they do, but first they need to know

why they should

buy from one organization rather than another.

Status by Achievement Status by Ascription

Programmatic sites that “do some

thing” E-commerce functions secondary to organizational

information

Focus on outcomes Focus on reputation and qualifications

Visual angle giving power to the viewer Visual angle

enhancing size of object

Images of youth (both male and fe

male) Images of older men

Leaders made to appear approachable;

names used without titles (but qualifica

tions often given indirectly) Leaders given full job titles

and honor ifics

Frequent updates and revisions Site relatively static

(consistent over time)

Many links (broad action hierarchy)

presenting many choices Fewer links (deep action hierarchy)

with choices more strongly structured An established
company like IBM would have a built-in advantage over

a relatively new company like Dell when selling to people

from an ascrip

tion culture. It could maximize this advantage by reminding

people of its
technical leadership and focusing on the CEO. In an
achievement culture,

such information would be secondary to providing advanced

functions

(like the ability to build your own PC through the web

site). Too much rela

tional information might even be viewed as “information

overload” and

become detrimental. Differences in time horizons can also

influence the amount of text and

the choice of graphic design. Sites with a past orientation

need to give a lot

of historical context and use traditional color schemes;

sites with a present

orientation should focus on quick action and use bright,

active colors; and

sites with a future orientation ought to talk about mission

and future plans

and consider using a pale (indecisive) palette.

Orientation to Past Orientation to Present Orientation to

Future

Dark colors with histori

cal connotations Bright, active colors Futuristic or dreamy

col ors

Background information

before new information Current information only Information

on mission and future plans

Emphasis on text and his

torical information Emphasis on empowering users to act

quickly Promotion of organizational innovation and use of
novel interface features

Images of historical fig

ures or older people Images of contemporary people doing
things Nonrepresentational or idealized images These color
schemes and rhetorical strategies could also help a site
tar

get different domestic audiences, such as conservative

investors or early

adopters of new technology. Sequential and synchronic sites

might use different interaction models.

For instance, a sequential site might use a careful

hierarchy to lead people

through tasks one step at a time. (To buy a train ticket, a

user would first be

required to choose a destination. Then the system would

provide a list of

departure times. Finally, the system would ask for the

number of people in

the party and let the user choose from available classes of
travel.) A

synchronic site might ask all those questions on the same

page, knowing

that some queries could fail and force the user to choose
again.

Sequential Use of Time Synchronic Use of Time

Strongly structured task hierarchies “One-click”

interaction model

Wizards Tolerance for failure

“Breadcrumb trails” and site maps Total reliance on search

engines Finally, people from cultures with a control
(inner-directed) orientation

to nature are likely to celebrate human agency and

technology, whereas

people from cultures with a harmony (outer-directed)

orientation might fa

vor graphics of nature and social cooperation.

Inner-Directed Cultures Outer-Directed Cultures

Focus on action and technology Focus on nature and human

community

Spare, instructional text Figurative, metaphoric text

Images of human achievements (e.g., a

picture of the Hoover dam) Images of nature (small-scale

images that evoke nature in microcosm or pan oramas that
emphasize its power)

Bright, active colors Pale, “natural” colors

Symmetric templates Asymmetric templates These variations

are all speculative. Unlike Hofstede’s work dimensions,

Trompenaars’ cultural differences have not been extensively

used in re

search, but his work does fill gaps in the analytic

frameworks of other cul

tural values researchers. In addition, five of his seven

dimensions apply

directly to communication styles and relational goals.

DAVID VICTOR: CAPTURING INTERCULTURAL

COMMUNICATION IN A MNEMONIC

One last theorist remains to be discussed. David Victor

takes a more histori

cal and categorical approach to intercultural

communication, focusing on

specific elements of the communication context that can

facilitate or hinder

mutual understanding. In his 1992 text, he warned that

international business communication is

increasingly subject to two contradictory trends: the

homogenization of

consumer needs (and products) and the intensification of

cultural hetero

geneity. On the one hand, consumer demographics have

converged so that

you can buy blue jeans in Japan, Pokemon in Africa, CDs by

Ali Farka Toure

in the United Kingdom, and Christmas crackers in the United

States. On

the other, people have begun to resist homogenization by

reemphasizing

their cultural uniqueness. Strong political-social

movements in Quebec,

Eastern Germany, and much of the Arab world seek to protect

cultural iden

tity. Even Japan, an enthusiastic consumer of Western

popular culture, re

cently instituted a Council on the Japanese Language to

investigate the

proliferation of loan words in katakana. Victor promises

that anyone can learn to communicate in any culture by

using his “culture-general” approach (though they will need

to work at it.) A

mnemonic—LESCANT—captures the seven most important aspects

of in

ternational business communication: 1. Language 2.

Environment and technology 3. Social organization 4.
Contexting 5. Authority conception 6. Nonverbal
communication 7. Temporal conception. The pragmatics of
communication tend to be more important to Victor

than cultural theories. However, unlike popular books on

“Do’s and Ta

boos” (specific things to do, or avoid, when dealing with

people from partic

ular countries), his analysis is detailed and

comprehensive. For instance,
the chapter on language ranks the world’s leading
languages, discusses the

Sapir–Whorf hypothesis, gives examples of linguistic

ethnocentrism that

can isolate the non-speaker, identifies problems related to

linguistic (but

not cultural) fluency, lists strategies for working with

translators, and cau

tions against stereotyping based on broken speech. Unlike

most of the theorists, Victor also comments on the role of
tech

nological factors in international communication. He

describes how a

sales opportunity was lost when a business person expected

a potential cli

ent to have easy access to a phone. Software designers

often make the same

error, expecting users to have access to the latest

hardware and software

and the desire to work with the latest release. In many

countries, people

deliberately choose to work with back-level systems to

prevent technical

problems and reduce risk. Victor’s discussion of

“contexting” (communication that relies on infor

mation outside the message) merges Hall’s theory with

Hofstede’s notion of

uncertainty avoidance. Communication in high-context

cultures is embed

ded in strong interpersonal relationships: Strangers must

develop trust be

fore they can successfully negotiate with one another. By

contrast, people in

low-context cultures focus on the end product, not the

provider. Perhaps as a

result, they depend to a greater degree on explicit verbal

and written com

munication, contracts, and commercial law. In addition,

people in high-con

text cultures rely more on nonverbal communication and

become skilled in

inference. Silence and the settings in which people speak

often communicate

more meaning than the polite, indirect speech they use.

However, this am

biguous style does not create any tolerance for ambiguity

or risk. People in high-context cultures avoid uncertainty
by acting conserva

tively and gathering as much information about their

communication

partners as possible. This relates back to the emphasis on

personal rela

tionships, but it comes with a downside: “Face” becomes

central to the

communication process. One is forced to protect one’s own

public per

sona and the dignity of everyone else. Consequently, people

in high-con

text cultures prefer indirection, harmony, and politeness

to argumen

tation, confrontation, and self-disclosure. Everything

works well until

low-context people attempt to communicate with them. Then,

the differ

ences in communication style may end in shame and mutual

loss of face—

although the low-context partner often does not know it.

(Stella Ting
Toomey [1996]; Ting-Toomey, Oetzel, & Yee-Jung [2001]
discussed face

negotiation in more detail.) Victor’s wealth of information

can be overwhelming. Like the “thick de

scription” of anthropology, it is hard to use the LESCANT

system to predict

areas of concern because you have to be immersed in the

other culture to do

the analysis. But if you plan to target just one other

culture, Victor’s frame

work is an excellent way to identify areas of agreement and

cultural faultlines.

Victor as a Guide for Interface Development

Victor’s focus on the pragmatics of business communication

has many

points of intersection with interface design. For instance,

localization often

treats the issue of language as a simple technical problem,

that is, leave

room for expansion and translate the text. But Victor

(1992) notes that ap

parent fluency has its own drawbacks: For example, a U.S.

businesswoman may represent a company in negotiating with a
Japanese firm. The Japanese negotiating team with whom she
interacts has little knowledge of English. Recognizing that
the woman speaks Japanese, the negotiating team is likely
to judge her by Japanese standards. When the U.S.
businesswoman breaks Japanese gender-linked social expec
tations and does not behave as they might expect a Japanese
woman to be have, she is less likely to be forgiven than a
foreigner unable to speak Japanese might be. (p. 30).
Localized products often evoke similar responses. Users
appreciate be

ing communicated within their own language, but

untranslated cultural as

sumptions undermine their good impression. Problematic

interface

elements include informal text based on different norms of

politeness and

icons or graphics that violate national and religious

conventions. (Paradox

ically, interfaces in the original language reduce user

expectations and may

increase user tolerance for social gaffes.) Interfaces need

to be “culturalized” as well as translated; icons,
graphics,

rhetoric, and interaction style must also suit the target

audience. A German

telecommunications company found that investing in product

differentia

tion generated higher acceptance. Pia Honold (1999) ran a

series of focus

groups and tests on user documentation for cellular phone

customers in

Germany and China. German users preferred a thick book

identifying all

the phone functions; they wanted explicit information in as

much detail as

possible. Chinese users asked for online documentation;

they would lose

face if other people saw them looking at a book to solve a

problem. Corporate guidelines for localization should
include a LESCANT check

list. The interplay between cultural values and effective

communication in

volves a wide range of pragmatic issues that are easy to

overlook.

SUMMARY

So, after discussing all these theorists, what are the most
important cultural
values to consider when designing software? As always, “it
depends!” If I were adopting just one theorist’s model of

culture and communication, I would use that of Geert

Hofstede or Shalom

Schwartz. Hofstede’s (1997) work dimensions have a great

deal of face va

lidity and are widely used in research, but they are not
entirely independent

of one another. In addition, his country rankings omit some

important na

tions. By contrast, Schwartz’s (1994) cultural dimensions

provide an alter

native to Hofstede’s work and also include rankings. But

his seven values

are more difficult to grasp and some of the countries

included in his list are

atypical (Estonia, Bulgaria, etc.) . If I were picking a

set of values from those that show up over and over

again in various theories and studies, I would start with

these elements of

the social environment: (a) individualism and collectivism,

(b) power dis

tance, (c) human action (mastery and harmony), and (d) time
(particularly

the time horizons). The primacy of the individual or the

group seems to be the key consider

ation for intercultural relationships. Kluckhohn and

Strodtbeck, Hall, Con

don and Yousef, Hofstede, Triandis, Gudykunst, Trompenaars,

and Victor

all explicitly reference individualism and collectivism.

Schwartz claims to

go “beyond individualism/collectivism,” but his cultural

dimensions reflect

the same social dichotomy: The two autonomy dimensions and

self-en

hancement (individualism) are the obverse of conservatism

and self-tran

scendence (collectivism). Conceptions of social power

(power distance, hierarchy, ascription, and

achievement) strongly covary with individualism and

collectivism. They

should be considered in conjunction with it or folded into

the dimension as

vertical and horizontal individualism and collectivism

(Triandis, 1995).

Understanding how humans interact with computers, and with

one another

across cultures through applications or web sites, also

requires an under

standing of the ordering of social relationships. Human

action supports a set of themes associated with mastery and

harmony. Cultures that support mastery focus on personal

empowerment

and activity for its own sake, whereas those that support
harmony tend to

be more passive and perhaps more fatalistic. There is some

overlap be

tween mastery, masculinity, and individualism and harmony,

femininity,

and collectivism, but country rankings do not always bear

this out. For in

stance, Schwartz (1994) ranks France at the bottom of his

country list for

mastery but Trompenaars and Hampden-Turner (1998) put it

near the
top of their list for control of one’s fate. Nonetheless,
the social role ac

corded technology in different cultures is an important

aspect of the hu

man–computer interface. The fourth dimension is time,

though its expression is multifaceted. Cul

tural views of time show up in decision making predicated

on short-term

rather than long-term goals, rhetorical strategies that

ignore or appeal to

history, specific or diffuse relationships, and monochronic

or polychronic

scheduling of events. However, in general, cultures that

focus on the pres

ent tend to ignore long-term consequences (whether past or

future) and

work with other people through more instrumental (specific)

relationships. To round out my selection, I would also pick
the two cultural dimensions

that most clearly deal with communication style: (a)

context and (b) affec

tive (expressive) and neutral communication. Both overlap

aspects of the first four values I mentioned, but more im

portantly they supplement one another. Considering how

people in differ

ent cultures show or restrain their emotions enriches

Hall’s theory of high

and low-context communication. Hall linked individualistic

cultures to ex

plicit messages and collectivist cultures to nonverbal

messages context

ualized within a social setting, but both types of messages

can show varying

amounts of affect. Most low-context countries, like the

Netherlands, use a

neutral, unemotional style, but so do some high-context

countries, like Ja

pan. Arab culture tends to be expressive and high-context.

Affective and

neutral communication should be considered in regard to

communication

context, just as power distance and hierarchy have been

folded into individ

ualism and collectivism.

FUTURE DIRECTIONS

Despite the development of virtual teams and multinational

business prac

tices, culture is recognized as a far more potent force

outside the United

States than in it. The expansion of the Common Market, the

success of Jap

anese management (before the collapse of the “bubble

economy”), and the

globalization of the Chinese and Southeast Asian economies

have

prompted the rest of the world to reexamine the validity of

Western social

science and management theories. International journals of

intercultural

communication, cross-cultural psychology, and management

and technol

ogy contain work with immediate application to HCI. Many

researchers in China, Japan, Korea, and Southeast Asia are
partic

ularly active in identifying alternative cultural

paradigms. Michael Bond (in

Hong Kong) helped Geert Hofstede identify Chinese values;

Tsukasa

Nishida, Young Yun Kim, Yuko Matsumoto, Kwangsu Kim, and

Taae-Seop

Lim have been exploring the validity of other

generalizations about com

munication. Cultural values, with their built-in

dichotomies, work well

within the European and American academic tradition, but a

great deal of

innovative research is being done throughout the world to

develop more

holistic concepts of culture. Western software development

theories are also being subjected to anal

ysis. People like Alvin Yeo have investigated the

assumptions underlying

design and evaluation practices. Researchers dealing with

the transfer of

collaborative technologies are finding that basic

principles of usability do

not have the same utility in all cultures. In Malaysia, Yeo

(2001) discovered that usability tests and other stan

dard methodologies associated with Western software

engineering fre

quently don’t work. Malaysian users testing a spreadsheet

he developed

found it difficult to express their dissatisfaction;

because of their collectivist

orientation, they censored themselves to preserve his face.

Yeo received the

most accurate feedback from people already known to him

(who knew he

needed their negative comments) and from computer experts

(whose ex
pertise authorized them to speak critically). He suggests
that interface test

ing in Asia must be based on objective performance measures

rather than

on subjective user feedback. The International Workshops on

Internation

alisation of Products and Systems (IWIPS) showcase similar

war stories

about interface development from people working around the

world. Western bias may even underlie some of our most
cherished HCI beliefs.

Individualistic forms of decision making and task

orientation are not uni

versal. In the late 1980s and early 1990s, group decision

support systems

(GDSS) were introduced into a number of countries in Asia

and Africa.

Richard T. Watson, Teck Hua Ho, and K. S. Raman (1994)

found that

Singaporean groups began with more consensus and showed

less change

than U.S. groups. They speculated that the need for group
harmony under

mined individual problem solving. Although the effect was

weak—coming

after group size, member proximity, and task type—they

considered cul

ture a necessary factor in interpreting GDSS outcomes. Four

years later, Gert-Jan De Vreede, Noel Jones, and Rabson J.
Mgaya

(1998) reported on the introduction of GDSSs in Malawi,

Zimbabwe, and

Tanzania. They found that incentives for participation were

positively as
sociated with endorsement by top management and negatively
affected by

anonymous input. In these vertically collectivist states,

people depended

on leaders to make initial judgments. In addition,

participants wanted the

system to identify the contributions of individuals so that

they could de

cide whose ideas could be developed. Paradoxically, then,

successful tech

nology transfer depended on hierarchical influence and

claims to “own”

ideas, two usability features that the software had been

designed in the

United States to circumvent. (See Jay F. Nunamaker et al.,

1991, for a list

of these design goals.)

CONCLUSION

Designing effective interfaces is not just good business.

Software for inter

cultural audiences can bring people together or drive them

apart. Inappro

priate design that violates cultural expectations

perpetuates the digital

divide (Shneiderman, 2000, p. 86); software that offends

local norms fuels

opposition to globalization. The Web is an emerging focal

point for con

flict. K. Viswanath and Liren Benjamin Zeng (2002) recently

documented

the role of transnational advertising in anti-globalization

protests. Simi

larly, European privacy advocates have attacked Microsoft

for its strategies

on personal data. Interface designers need to think about

ways to reduce alienation and

cultural bias in software that, by design, cuts across

national and cultural

boundaries. Incorporating cultural values into our design

practices is a

good way to start. We also need to test software in a

variety of countries and

factor culturalization (as well as translation) into

product development.

Perhaps in the future, software design will be reformed by

a new paradigm

from Asia, Latin America, or Africa that transcends these

problems. Until

then, we need to be mindful of the effects of culture and

do our best to pro

vide people everywhere with interfaces that support them.

Barber, W., & Badre, A. (1998). Culturability: The merging

of culture and usability. In Proceedings of the Fourth
Conference on Human Factors and the Web. Retrieved Jan uary
24, 2004, from
www.research.microsoft.com/users/marycz/hfweb98/bar
ber/index.htm

Benedict, R. (1989). The chrysanthemum and the sword:

Patterns of Japanese culture. Boston: Houghton Mifflin.
(Original work published 1946)

Condon, J. C., & Yousef, F. S. (1975). An introduction to

intercultural communication. New York: Macmillan.

Cross, S. E., Bacon, P. L., & Morris, M. L. (2000). The

relational-interdependent self-construal and relationships.
Journal of Personality and Social Psychology, 78, 791–808.

de Mooij, M. (2001). Internet and culture. Proceedings of

the Conference on Internet , Economic Growth, and
Globalization (August 8, 2001). Duisburg, Germany: Insti
tute for International and Regional Economic Relations,
Gerhard-Mercator Uni versity.

De Vreede, G-J., Jones, N., & Mgaya, R. J. (1998).

Exploring the application and ac ceptance of group support
systems in Africa. Journal of Management Information
Systems 15, 197–234.

del Galdo, E. M., & Nielsen, J. (1996). International user

interfaces. New York: Wiley.

Fernandes, T. (1995). Global interface design: A guide to

designing international user inter faces. Boston: AP
Professional.

Gudykunst, W. B. (1983a). Similarities and differences in

perceptions of initial intracultural and intercultural
encounters: An exploratory investigation. The Southern
Speech Communication Journal, 49, 58–65.

Gudykunst, W. B. (1983b). Uncertainty reduction and

predictability of behavior in lowand high-context cultures:
An exploratory study. Communication Quarterly, 31, 49–55.

Gudykunst, W. B. (Ed.). (1986). Intergroup communication.

London: Edward Arnold.

Gudykunst, W. B. (1987). Cross-cultural comparison. In C.

R. Berger & S. H. Chaffee (Eds.), Handbook of Communication
Science (pp. 847–889). Beverly Hills, CA: Sage.

Gudykunst, W. B. (1988). Uncertainty and anxiety. In Y. Y.

Kim & W. B. Gudykunst (Eds.), Theories in Intercultural
Communication (pp.123–156). Newbury Park, CA: Sage.

Gudykunst, W. B. (1993). Toward a theory of effective

interpersonal and intergroup communication: An anxiety/
uncertainty management (AUM) perspective. In R. L. Wiseman
& J. Koester (Eds.), Intercultural Communication Competence
(pp. 16–32). Newbury Park, CA: Sage.

Gudykunst, W. B. (1994). Bridging differences: Effective

intergroup communication (2nd ed.). Thousand Oaks, CA:
Sage.

Gudykunst, W. B. (1995). Anxiety/ uncertainty management

(AUM) theory: Current status. In R. L. Wiseman (Ed.),
Intercultural Communication Theory (pp. 9–58). Thousand
Oaks, CA: Sage.

Gudykunst, W. B., Chua, E., & Gray, A. J. (1987). Cultural

dissimilarities and uncertainty reduction processes.
Communication Yearbook, 10, 456–469.

Gudykunst, W. B., & Kim, Y. Y. (1984). Communicating with

strangers: An approach to intercultural communication. New
York: Random House.

Gudykunst, W. B., & Lim, T-S. (1986). A perspective for the

study of intergroup com munication. In W. B. Gudykunst
(Ed.), Intergroup communication (pp. 1–9). Lon don: Edward
Arnold.

Gudykunst, W. B., Matsumoto, Y., Ting-Toomey, S., Nishida,

T., Kim, K., & Heyman, S. (1996). The influence of cultural
individualism-collectivism, self construals, and individual
values on communication styles across cultures. Human
Communi cation Research, 22, 510–543.

Gudykunst, W. B., & Nishida, T. (1986). The influence of

cultural variability on per ceptions of communication
behavior associated with relationship terms. Human
Communication Research, 13, 147–166.

Gudykunst, W. B., & Ting-Toomey, S. (1988). Culture and

affective communication. American Behavioral Scientist, 31,
384–400.

Gudykunst, W. B., & Ting-Toomey, S., with Chua, E. (1988).

Culture and interpersonal communication. Newbury Park, CA:
Sage.

Hall, E. T. (1973). The silent language. Garden City, NY:

Anchor Books. (Original work published 1959)

Hall, E. T. (1976). Beyond culture. Garden City, NY: Anchor

Press.

Hofstede, G. (1980). Culture’s consequences: International

differences in work-related val ues. Beverly Hills, CA:
Sage.

Hofstede, G. (1997). Cultures and organizations: Software

of the mind (Rev. ed.). New York: McGraw-Hill.

Hofstede, G., & Bond, M. H. (1988). The Confucius

connection: From cultural roots to economic growth.
Organizational Dynamics, 16, 417–433.

Hoft, N. (1996). Developing a cultural model. In E. M. del

Galdo & J. Nielsen (Eds.), International user interfaces.
New York: Wiley.

Honold, P. (1999). Learning how to use a cellular phone:

Comparison between Ger man and Chinese users. Technical
Communication, 46, 196–205.

Horton, D., & Wohl, R. R. (1956). Mass communication and

para-social interaction. Psychiatry, 19, 215–229.

Kluckhohn, C. (1951). Values and value orientations in the

theory of action: An ex ploration in definition and
classification. In T. Parsons & E. Shils (Eds.), Toward a
general theory of action (pp. 390–406). Cambridge, MA:
Harvard University Press.

Kluckhohn, F. R. (1950). Dominant and substitute profiles

of cultural orientations: Their significance for the
analysis of social stratification. Social Forces, 28,
376–394.

Kluckhohn, F. R., & Strodtbeck, F. L. (1961). Variations in

value orientations. Westport, CT: Greenwood Press.

Kostelnick, C. (1995). Cultural adaptation and information

design: Two contrasting views. IEEE Transactions on
Professional Communication, 38, 182–196.

Kress, G. R., & van Leeuwen, T. (1996). Reading images: The

grammar of visual design. London: Routledge.

Lea, M., & Spears, R. (1991). Computer-mediated

communication, de-individuation and group decision making.
International Journal of Man-Machine Studies, 39, 283–310.

Lederer, W. J., & Burdick, E. (1958). The ugly American.

Greenwich, CT: Fawcett.

Lee, O. (2000). The role of cultural protocol in media

choice in a Confucian virtual workplace. IEEE Transactions
on Professional Communication, 13, 196–200.

Marcus, A. (1993). Human communication issues in advanced

UIs. Communications of the ACM, 36, 101–109.

Marcus, A. (2001). International and intercultural user

interfaces. In C. Stephanidis (Ed.), User interfaces for
all: Concepts, methods, and tools (pp. 47–63). Mahwah, NJ:
Lawrence Erlbaum Associates.

Marcus, A., & Gould, E. W. (2000). Crosscurrents: Cultural

dimensions and global web user-interface design.
Interactions, 7, 32–46.

Markus, H. R., & Kitayama, S. (1991). Culture and the self:

Implications for cogni tion, emotion, and motivation.
Psychological Review, 98, 224–253.

Nisbett, R. E. (2003). The geography of thought: How Asians

and Westerners think differ ently … and why. New York: Free
Press.

Nunamaker, J. F., Dennis, A. R., Valacich, J. S., Vogel,

D., & George, J. F. (1991). Elec tronic meeting systems.
Communications of the ACM, 34, 40–61.

Ong, W. J. (1975). The writer’s audience is always a

fiction. Publications of the Modern Language Association of
America, 90, 9–21.

Parsons, T. (1951). The Social System. Glencoe, IL: Free

Press.

Pike, K. L. (1954). Language in relation to a unified

theory of the structure of human behav ior. Glendale, CA:
Summer Institute of Linguistics.

Rafaeli, S. (1988). Interactivity: From new media to

communication. In R. P. Hawkins, J. M. Wiemann, & S.
Pingree (Eds.), Advancing communication science: Merging
mass and interpersonal processes (pp. 110–134). Newbury
Park, CA: Sage.

Rafaeli, S., & Sudweeks, F. (1998). Interactivity on the

nets. In F. Sudweeks, M. McLaughlin, & S. Rafaeli (Eds.),
Network and netplay: Virtual groups on the internet (pp.
173–189). Menlo Park, CA: AAAI Press.

Reeves, B., & Nass, C. (1996). The media equation: How

people treat computers, television, and new media like real
people and places. New York: Cambridge University Press.

Said, E. W. (1994). Orientalism. New York: Viking Books.

(Original work published 1978)

Shneiderman, B. (2000). Universal usability. Communications

of the ACM, 43, 84–91.

Schwartz, S. H. (1994). Beyond individualism/collectivism:

New cultural dimen sions of values. In U. Kim, H. C.
Triandis, C. Kagitcibasi, S-C. Choi, & G. Yoon (Eds.),
Individualism and collectivism: Theory, method, and
applications (pp. 85–119). Thousand Oaks, CA: Sage.

Sheppard, C., & Scholtz, J. (1999). The effects of cultural

markers on web site use. In Proceedings of the 5th
Conference on Human Factors and the Web. Retrieved January
23, 2004, from http://zing.ncsl.nist.gov/
hfweb/proceedings/sheppard/index.html

Singelis, T. M. (1994). The measurement of independent and

interdependent self-construals. Personality and Social
Psychology Bulletin, 20, 580–591.

Stewart, E. C., & Bennett, M. J. (1991). American cultural

patterns: A cross-cultural perspective (Rev. ed.).
Yarmouth, ME: Intercultural Press.

Stephenson, G. M. (1981). Intergroup bargaining and

negotiation. In J. C. Turner & H. Giles (Eds.), Intergroup
behavior (pp. 168–198). Chicago: University of Chicago
Press.

Ting-Toomey, S. (1996). Intercultural conflict styles: A

face-negotiation theory. In Y. Y. Kim & W. B. Gudykunst
(Eds.), Theories in intercultural communication (pp.
213–235). Newbury Park, CA: Sage.

Ting-Toomey, S., Oetzel, J. G., & Yee-Jung, K. (2001).

Self-construal types and con flict management styles.
Communication Reports, 14, 87–104.

Triandis, H. C. (1975). Social psychology and cultural

analysis. Journal for the Theory of Social Behavior, 5,
81–106.

Triandis, H. C. (1978). Some universals of social behavior.

Personality and Social Psy chology Bulletin, 4, 1–16.

Triandis, H. C. (1980). (Ed.). Handbook of cross-cultural

psychology. Boston: Allyn & Ba con.

Triandis, H. C. (1983). Dimensions of cultural variation as

parameters of organiza tional theories. International
Studies of Management and Organization, 12, 139–169.

Triandis, H. C. (1988). Collectivism vs. individualism: A

reconceptualization of a ba sic concept in cross-cultural
psychology. In G. K. Verma & C. Bagley (Eds.),
Cross-cultural studies of personality, attitudes, and
cognition (pp. 60–95). London: Macmillan.
Triandis, H. C. (1989). The self and social behavior in
differing cultural contexts. Psychological Review 96,
506–520.

Triandis, H. C. (1995). Individualism and collectivism.

Boulder, CO: Westview Press.

Trompenaars, F., & Hampden-Turner, C. (1998). Riding the

waves of culture: Under standing diversity in global
business (2nd ed.). New York: McGraw-Hill.

Victor, D. A. (1992). International business communication.

New York: HarperCollins.

Viswanath, K., & Zeng, L. B. (2002). Transnational

Advertising. In W. B. Gudykunst & B. Mody (Eds.), Handbook
of International and Intercultural Communication (2nd ed.).
Thousand Oaks, CA: Sage.

Walther, J. B. (1996). Computer-mediated communication:

Impersonal, interper sonal, and hyperpersonal interaction.
Communication Research, 23, 3–43.

Walther, J. B., & Burgoon, J. K. (1992). Relational

communication in computer-me diated interaction. Human
Communication Research, 19, 50–88.

Watson, R. T., Ho, T. H., & Raman, K. S. (1994). Culture: A

fourth dimension of group support systems. Communications
of the ACM, 37, 44–55.

Williams, R. (1994). The non-designer’s design book: Design

and typographic principles for the visual novice. Berkeley,
CA: Peachpit Press.

Yeo, A. W. (2001). Global-software development lifecycle:

An exploratory study. In Proceedings of the SIGCHI
Conference on Human Factors in Computing (pp.104–111). New
York: ACM Press. Chapter

Managing Multicultural Content

in the Global Enterprise

Jorden Woods

Co-Founder, Global Sight Corporation, San Jose, CA

INTRODUCTION AND THE IDEAL ENTERPRISE

This chapter is primarily concerned with the application of
multicultural

content best practices to a global enterprise.

Specifically, the chapter deals

with how to apply multicultural best practices not only to

a single product or

deliverable, but to all consumable content generated and

maintained in an

international organization on an ongoing basis. To better

understand what the challenges are and how the application
of

best practices affects processes within a global

corporation, I first look at a

vision of a fictitious ideal enterprise in the context of

multicultural content.

To clarify, this ideal enterprise is that organization

which is capable of pro

viding every significant customer with the same high

quality user experi

ence regardless of country of origin or ethnic background.

The easiest way

to visualize this ideal enterprise is to view it through

the eyes of its custom

ers, partners, and employees on a global basis.

Customer Experience

Every customer experience is clearly a very personal

situation that cannot

be uniquely determined or parameterized. However, it is

possible to gener

alize a customer’s experience given enough interaction in

combination with 123

customer profiling and user feedback. In this generalized

sense, the ideal
experience can be conceived as one in which the customer is
provided with

or has easy access to useful information and/or

functionality that meets the

timely and relevant manner. In this ideal setting, it

should be clear that be I a Zulu farmer, a princess

in the Thai royal family, or a French nuclear technician I

should value my

interactions with the company’s representatives and its

content. In short

each interaction will meet my particular needs, be they: •

Linguistic, expected or preferred language or dialect. •
Cultural, expected or preferred cultural nuance in terms of
values, historical allusion, use of colors and icons,
relationships, hierarchy, religion, worldview, and so on. •
Technical, expected or appropriate weights, measures,
protocols and conventions. • Economic, expected or
preferred currency, taxation, shipping methods, and so on.
• Legal, expected or appropriate statutory or regulatory
conventions. • Geographical, expected or appropriate
climactic or seasonal norms. Unfortunately, by not fully
understanding their customer’s specific

needs, global enterprises often unwittingly provide many of

their global

customers with inappropriate information or functionality

in a culturally

threatening or linguistically intimidating style. To

further define this ideal experience, consider the
following scenario:

a U.S. company has launched an existing North American

wireless product

for the Asian marketplace. Primary markets for launch

include China, Ja

pan, Korea, Taiwan, Malaysia, Thailand, Indonesia, and

Australia. The list

below details the key issues that would need to be taken

into account to min
imize customer alienation and generate keen user interest:
• Localized product interface. • Culturally appropriate
product size, shape, and color. • Culturally and
linguistically appropriate name and branding. • Competitive
distribution channels and partnerships for product
placement, accessibility, and promotion. • Timely localized
information about the product in multiple for mats: Web,
print, wireless, and so on. • Country or region specific
services which appeal to the primary customer demographic.
• Easily accessible, in-language customer service. •
Adoption of local protocols and standards to maximize
interoperability with locally relevant systems,
accessories, and in frastructure. By ensuring that the
product meets these requirements in each of its tar

get markets, the customer will feel that they are

purchasing a product that

appears to have been developed specifically to address his

or her needs. Based on the foregoing analysis, it should
now be clear that simply re

packaging the device with translations of the original

content will be insuffi

cient to generate an ideal customer experience.

Employee or Internal Process Experience

There are many internal process scenarios that can lead to

the ideal cus

tomer experience outlined in the previous section. For

example, employ

ees may be working under tremendous stress in a situation

where no

processes exist. Such a situation, by the very nature of

its inefficiency, is un

likely to generate a return on investment. Instead, I now

focus on an ideal

environment within a hypothetical global enterprise. To

facilitate the corporate employees in meeting the diverse
needs of a

global customer base, the ideal enterprise makes

information, processes,

and resources available for the intelligent generation and

update of multi

cultural information. So, when communicating and

interacting with a Zulu

farmer, a Thai princess, or a French engineer, appropriate

information is

available for each customer’s particular needs, and

efficient processes and

resources are available to deliver it quickly and

accurately. More specifically, I examine the particular
case of the U.S. company that

is intent on launching an existing North American mobile

device into the

Asian marketplace (China, Japan, Korea, Taiwan, Malaysia,

Thailand, In

donesia, and Australia). In this case, the ideal company’s

marketing and en

gineering staff would have easy access to the appropriate

linguistic,

cultural, economic, and technical information to

effectively market and

modify the product for each target country prior to product

launch. The ideal company’s business and information
technology (IT) re

sources ensure the quality and timeliness of the

information and functional

ity delivered. They would do this by developing and

providing a process

infrastructure for automating and streamlining the tasks

dictated by the

company’s business needs. Furthermore, resources, such as

experts in the

areas of cultural, legal, and tax accountancy, would be

available and would

have easy access to the company’s process flow. Therefore,

when simultaneously launching a product into markets as

different as Indonesia, China, and Australia, the company’s

employees are

able to rely on a breadth of information, process

infrastructure, and exper

tise to rapidly and routinely carry out their tasks.

Deadlines are met, ex

penses meet estimates, the uptake of the product in the

target markets is in

line with expectations, and the company’s brand is

enhanced. This scenario should be contrasted with one in
which, information, re

sources, and processes are not in place. In such a

scenario, deadlines are

only rarely met, expenses balloon out of control, uptake of

product is slow,

and the company’s brand is eroded or tarnished. Examples of

the latter

abound in the literature with cases such as the following:

• Nike’s initial humbling losses to Adidas in the European
market place based on a misreading of their “rebel”
branding as “hooli ganism.” • The failure of the Big Three
U.S. automakers to significantly infil trate Japan’s
Japanese language-centric, metric-standard, rightside drive
road system with English-language, English-system,
left-side drive cars. • The United States’ disastrous
campaign of feeding the hungry in Afghanistan with food
packs the same relative size and color as cluster bombs and
with information distributed to a nearly illiterate
population without pictures and only in English. •
Walmart’s slow and expensive expansion into Latin America
in which the U.S.-centric culture did not translate and
stores were stocked with electronics based on 110 voltage
which would not operate locally. • The U.S. Milk Council’s
unforgettable “Got Milk?” campaign, which, for the U.S.
Hispanic market, was unsuccessfully conveyed as “Are you
Lactating?” Success, however, can be spectacular as seen
in: • The ability of the Japanese Keiritsu to make rapid
progress into the consumer electronics, automotive, and
steel building sectors on a global basis. • SAP’s
domination of the Enterprise Resource Planning (ERP) space
on a global basis from its home base in Germany. • Nokia’s
rise from the small country of Finland to dominate the
global cell phone marketplace. With this ideal company as a
backdrop, it is now useful to explore the

challenges to success in greater detail.

CHALLENGES TO SUCCESSFUL GLOBALIZATION

As the examples in the previous section have clarified,

there are multiple

challenges to the successful development and management of

multicultural

information and functionality for global markets. These

challenges, how

ever, once analyzed can be seen to fall into three primary

categories: pro

cess challenges, integration challenges, and technical

challenges. The following sections explain each of these in
more detail.

Process Challenges

Historically, companies have globalized their operations

organically, with

new countries being added on an opportunistic basis through

distributor

relationships, through acquisitions, and, in some cases,

through the devel

opment of formal subsidiaries. This approach, by focusing

on each country

as a separate entity, has typically created organizations

that are highly de

centralized and only loosely integrated. As a result, it is

not uncommon for each office within an extended enter

prise to have different business goals, employ different

business processes,
and use different technology infrastructure to meet the
needs of the local

marketplace. Furthermore, it is generally the case that

divisions within a

single country operate autonomously or with only small

consideration for

the processes, goals, and technologies employed by other

divisions (con

sider marketing, sales, IT, and engineering, e.g.). From

this analysis one can quickly see that the average global
company is

not only loosely integrated internationally but is also

loosely integrated lo

cally. Due to the Web, there has been more focus recently
on cross-func

tional teams; however, it is more often the case that

divisions operate

independently or in serial, rather than parallel, fashion.

Today’s products and services change continuously, operate
under very

short deadlines, and demand tight communication and tight

integration

among all the members involved in the development,

promotion, sales, and

delivery processes. Under the operating model of a highly

decentralized

and only loosely integrated global organization,

differences between highly

incompatible business processes and technology

infrastructure are exacer

bated by differences in local language, culture, tax law,

legal approach, time

zones, and governmental regulations. In addition, business

process within individual countries will vary based
on the type or purpose of the content that is being
localized. For example,

the globalization business processes associated with a

press release in a sin

gle country may be radically different from the

globalization business pro

cesses associated with a product catalogue or interactive

customer support

within that same country. These innate incompatibilities

and differences create serious inefficien

cies in process and work flow. Leveraging content from one

region or coun

try to another can then only be overcome by organizations

expending large

amounts of both human and capital resources. The sum of

three variables—rate of content change, volume of content,

and number of languages or countries—can quickly reach

critical values

which overwhelm the process and resource infrastructure of

the corpora

tion. In short, it can become prohibitively expensive and

operationally im

possible for a company to maintain its global content so as

to meet the high

standards of quality and timeliness expected by customers,

partners, and

employees in the global marketplace. The greatest casualty

for a company succumbing to process bottlenecks

is that the company becomes unable to meet its time to

market commit

ments. Once this critical capability is compromised, the

company can suffer

in the following ways: • Brand and message fragmentation. •

Distribution of inaccurate, out-of-date, or contradictory
information. • Decreases in customer loyalty due to breach
of trust. • Noticeable functionality lag between
applications deployed in different markets. • A negative
return on investment due to globalization initiatives. Such
conditions can therefore hopelessly undermine a
globalization ini

tiative and send teams back to the drawing board in search

of better solu

tions. Overcoming process challenges are therefore highly

critical to

achieving success; however, there are other factors that

can halt a globaliza

tion initiative in its tracks, and these challenges fall

under the category of in

tegration challenges.

Integration Challenges

In a globalization initiative, there are two main points of

technology inte

gration which must be tackled to be successful. These areas

are (a)

interoperability with source content repositories and

applications to en

able content acquisition, and (b) interoperability with

service provider

tools and applications (i.e., translators, reviewers,

editors). Both revolve

around technology interoperability. The first involves

interoperability

with the technology infrastructure within the enterprise,

whereas the sec

ond involves interoperability both internal and external to

the enterprise

with desktop applications. As Internet technology has

matured, IT organizations have deployed
greater and more diverse varieties of applications within
the enterprise to

both manage and benefit from the increasing functionality

of this medium.

As a result, there has been a proliferation of

databaseand/or file-driven ap

plications for creating and personalizing content, such as

content manage

ment systems (CMSs), personalization systems, and

application servers. To facilitate online commerce and to
generate interactivity with business

processes, companies have also deployed procurement

systems, logistics

applications, and order processing systems. In addition,

other applications

have been deployed to interact with legacy systems, such as

databases,

mainframes, and ERP and MRP systems, as companies have

increasingly

moved their business operations to an Internet-based

environment. Each of these environments contains content
that may need to pass

through a globalization process such that it can be made

available to an im

portant target customer segment. Overcoming this challenge

is therefore

the first step in a globalization process. However,

integration does not end

with the source and target repositories; it is also

critical that content pro

vided in the globalization process be compatible with

desktop tools used by

contributors in the work flow process. For localization,

the tools of choice typically include localized or
specialty

word processors and translation tools such as Microsoft

Word, Trados Work

bench, Star Transit, or Déjà Vu. An inability to

interoperate with these prod

ucts may make it impossible to carry out translation and

localization of

content and therefore will quickly make globalization

impossible. As this analysis illustrates, it is generally
the case that the format for

content processing is often different from the storage

format. As a result,

after localization, it is critical that the format be once

again modified to

enable the storage repository to store and manage the

content. This pro

cess cycle, therefore, places very significant restrictions

on the level of

technology interoperability. If interoperability is low,

the difference can

only be made up through human labor, which is expensive

and, more im

portantly, time consuming. Technical challenges, however,

only begin with interoperability. There are

many others that must be considered to maintain the

integrity of content dur

ing a localization process. These other technical

challenges are detailed next.

Technical Challenges

On close inspection, each of the information storage and

processing environ

ments within an enterprise contains content in different

formats and has dif
ferent rules or Application Programming Interfaces (APIs)
for accessing it. As

global content leveraging can only begin by first sourcing

content from the

organization’s content repositories, it is critical that

any globalization solu

tion be capable of handling three different sets of

parameters simulta

neously. These parameters include content language

encodings, content

storage formats, and application and data repository APIs.

An inability to work with any one of the multitude of
permutations of these

parameters can lead very rapidly to the demise of any

globalization initiative. As an example, consider first a
case of content stored in XML, being

sourced from an Oracle database, in English (ASCII-7

language encod

ing), in the United States and being targeted to an I-mode

server, in Japa

nese, for distribution in Japan. To enable this transaction

to occur, there

need to be three successful transformations. First, the

content needs to be

localized into Japanese and then stored in either the

Shift-JIS or ISO

2022-JP language encoding. Next, the markup format needs to

be trans

formed from XML to cHTML (condensed HTML), and a template

needs

to be applied with a different screen format configuration

due to the dif

ferent device requirements. Finally, the content will need

to be mapped
from the Oracle database in the United States to a Japanese
version of

SQL-Server in Japan. Now consider an alternate scenario

occurring simultaneously with the

process just described: a Chinese HTML page stored in an

enterprise CMS

being targeted to a WAP server in France, or JSP-based

information in a

personalization server under a UNIX/Oracle configuration in

the United

States being targeted to a Brazilian, NT-based ASP server.

Obviously, the

combinations are enormous and interoperability across all

major lan

guages, formats, and applications/repositories is the only

road to both cur

rent and future success. Three additional issues now become

clear: 1. The systems that are interacting in this example
need to support double byte character sets, that is,
character encodings such as those for Chinese, Japanese,
and Korean. 2. Transformations need to be made on items
such as date and time format, naming conventions, and
weights and measures. 3. Storage of information within
these systems must allow for linguis tically correct
translation (i.e., no concatenation allowed). In short, it
is critical that these systems be internationalized so as
to in

nately handle these three issues. Without an

internationalized infrastruc

ture, there is the high probability for data corruption,

loss of content

integrity, and the generation of meaningless information.

By conquering all three of these challenges simultaneously,
an organiza

tion now has the ability to create the foundation for

content globalization

success. This foundation is explained in greater detail in

the next section on

best practices.

BEST PRACTICES

Key Terms

I often find that any discussion of multicultural content

is best understood

once five key terms are explained and agreed on. These
terms and their

So, localization is dependent on translation; however, it

is greater than it.

Globalization and personalization are dependent on

internationalization,

localization, and translation; however, they are greater

than all of them

combined. Understanding the relationships inherent in this

hierarchy, in

terestingly enough, lays the foundation for discussing the

primary best

practices for successful content globalization.

The Three Pillars

Based on my experiences over the last 5 years, three

critical best practices can

be summed up as the Three Pillars for successful

globalization. Each of these

needs to be done in succession, and when all three are in

place, a stable founda

tion is formed for a globalization initiative. The Three

Pillars are as follows: I. Strategy: Development of a high
level documented strategy which has the global buy-in of
all significant stakeholders; II. Infrastructure:
Development and management of a fully interna tionalized
infrastructure;

III. Global processes: Deployment and enforcement of

standardized and coordinated content localization processes
on a worldwide ba sis. Pillar I: Strategy. A globalization
initiative, by its very nature as a world

wide project, is strategic in nature. No matter how small,

there are ramifica

tions in multiple countries, in multiple languages and

cultures, which affect a

company’s perception and positioning. As a result, a

globalization initiative of

ten becomes a highly politicized event, with multiple

groups jockeying for con

trol, visibility, and direction. As a best practice,

setting the strategy is similar to the design phase for an

art or engineering project. This stage is critical because

without it, the direc

tion and goals of the initiative would otherwise be poorly

defined. Without

well-defined goals or direction, communication across

countries, lan

guages, and functional groups becomes nearly impossible.

Where possible,

the message will be, at best, confusing and will not enable
groups to engage

in meaningful discussion. Instead, the strategy for the

initiative should be developed by a working

group of key, high-level individuals and then openly

discussed as a living

document among all significant stakeholders. It is

mandatory that the indi

viduals driving the initiative be at as high a level as

possible in the organiza

tion to ensure that the results of the document be binding

and be

enforceable throughout the organization. Conversation among

the different groups will then have the chance to

stimulate truly global thinking among the participants.

Such thinking will,

in turn, uncover challenges and lead to the creation of

broad-based solu

tions. The strategy document can then be iterated until a

compromise is

reached which clearly articulates benefits to the global

stakeholder commu

nity. Once approved, this vision document will form the

basis for the road

map for the initiative. Pillar II: Infrastructure. Once the

business vision for the globaliza

tion initiative is in place and blessed by the

stakeholders, it is then possible

to consider the infrastructure implementation issues for

the project. It is at

this point that the team queries its infrastructure and

decides whether or

not it has sufficient and/or appropriate resources,

systems, and processes to

support the initiative. Unfortunately, it is all too common

that teams develop a business vision

and then find out during the implementation process that

their goals are

unsupportable. Due to the technical nature of electronic

media, it is critical

that the business vision be coupled with technological due

diligence. Any

gaps between what is called for in the vision and what is

available can then

be remedied prior to project initiation. As an example,

consider a team that has decided to roll out multicultural
content into the Eastern European and Asian marketplaces.
This team de

cides on its direction, goals, and road map but then

neglects to investigate

the resourcing and technical aspects of the project. Once

underway, the

team realizes that it does not know how to support double

byte character

sets (DBCSs) for the Asian marketplaces or the varied

scripts of Eastern Eu

rope. Also, its applications for developing content do not

support storage,

input, or rendering of the character sets. Furthermore, the

team’s lack of

access to multicultural resources leads unknowingly to the

development of

template colors and icons that are highly offensive in the

target regions. Such a scenario can be avoided by carefully
evaluating the human and

technological resource infrastructure to appropriately

develop, manage,

and update content for the target regions. Thinking through

these issues

and making changes to the infrastructure prior to project

kick-off will then

reduce the chances for surprises during the implementation

phase, or will

make the surprises introduced more manageable. Pillar III:

Global Processes. Although often overlooked during a

planning phase, the life of a globalization project is

typically much longer

in duration than the initial project development cycle. As

a result, prepar

ing for and deploying the infrastructure to support updates

and ongoing

work is just as critical for success as the initial

strategy and infrastructure

considerations. As was noted in the earlier section on

challenges, one of the greatest hur

dles to any globalization initiative is the process

challenge. Not only can this

challenge create a massive financial and resource drain, it

can nullify all ef

forts such that no noticeable gains are produced. The first

step to overcoming process-based challenges is to design
con

tent localization processes (i.e., work flows, content

preand postprocess

ing, etc.) that meet the business needs of the enterprise

on global, regional,

and local levels. Once this design is completed these

processes can then be

automated via a standardized infrastructure. The final

step, and another

key to sustainable globalization success, resides in having

visibility to, and

the power to enforce, these processes on a worldwide basis.

With global processes in place, the update of content to a
worldwide au

dience becomes routine and, in most cases, predetermined.

Crises are

greatly reduced, as most situations have already been

addressed, and there

are mechanisms in place for dealing with problems and

surprises. Further

more, worldwide teams are tightly integrated by the global

processes, which

means that work is completed quickly, efficiently, and to

the satisfaction of

all parties involved.

Central Oversight With Local Empowerment (COLE)

In addition to the Three Pillars, the fourth key best

practice is known as cen

tral oversight with local empowerment (COLE). To understand

this con

cept better in context, it is first instructive to discuss

alternative develop

ment strategies: centralized and decentralized development.

Centralized. Under a centralized development scenario, a
company, di

vision, or group manages all of its content creation,

technology, and human

resources in one location. Advantages to this approach

include the following: • Resources are coordinated with one
voice. • All business processes receive good visibility. •
Consistency can be enforced. • Costs and scheduling receive
good oversight. • Single technology infrastructure
minimizes technical challenges. Though appealing because of
these strengths, this approach has a signif

icant downside, based primarily on the fact that the

centrally developed

content is actually to be targeted to other regions. As has

been noted, these

other countries may have significantly different language,

cultural, tax, le

gal, regulatory, and economic environments that will need

to be taken into

consideration. If these factors are not taken into account,

then the entire

centralized initiative can be called into question. The

list below summarizes the weaknesses of the centralized
approach: • Costs for multicultural consultancy can rise
exponentially. • Content may not conform to the needs of
countries targeted. • Scheduling may not meet the needs of
country offices. • Technologies and media offered may not
meet standards in target countries. • Country offices may
attempt to redo work created centrally, leading to rising
costs. Centralized development may work in limited cases;
however, in the

long term this approach usually leads to significant

problems. Another ap

proach, which is opposite in philosophy to the centralized

approach, is typi

cally called the decentralized approach. It is discussed in

more detail next. Decentralized. Under the decentralized
approach, a company, such as

a multinational, empowers each of its country offices or

subsidiaries to act

autonomously to produce content for its local needs. This

approach clearly

addresses the weaknesses of the centralized approach by

putting each re

gion in the driver’s seat for its content. As a result, the

strengths of this ap

proach include the following: • Country content is focused

on the specific needs of each region. • Locale-specific
issues, such as language, culture, and currency, are
addressed automatically. • Scheduling of content releases
meets the needs of each office. • Promotions for partners,
holidays, and local events are locally con trolled. •
Content is actively tied into the country offices’ sales
and marketing plans. As with the centralized approach,
however, there are issues that create

problems for the decentralized approach. Most of these

problems revolve

around the development of severe process and integration

challenges that

hinder the company’s ability to coordinate distributed

teams and maintain

costs. It should be noted as well, that the greatest

challenges arise in an on
line environment due to the ease with which information
from different

countries can be accessed and compared. The list below

summarizes the

problems associated with the decentralized approach: • Lack

of central oversight makes coordination on a regional or
global basis nearly impossible. • Significant
incompatibilities may arise in the technological and
process infrastructure so that different offices cannot
work together effectively. • Costs skyrocket due to
redundancies in resources and technology across offices. •
Global branding, messaging, and look and feel becoming
highly fragmented. • Inconsistencies and contradictory
information across countries can create increased customer
service and customer loyalty issues. Interestingly, one can
see that the strengths of each approach, central

ized and decentralized, cancel out the weaknesses of the

other. A hybrid

approach, which combines the best of both of these worlds,

is known as

central oversight with local empowerment. This approach is

detailed in

the following subsection.

Central Oversight With Local Empowerment

As discussed in the previous sections, the strength of the

centralized ap

proach lies in its ability to coordinate and oversee

processes with a single

voice and technology infrastructure, whereas for the

decentralized ap

proach, its strengths lie in enabling local offices to

fine-tune content to meet

the needs of their locale. The optimal approach would then

provide a cen

tralized infrastructure that oversees and coordinates all

processes, while all
processes would route content through the local offices to
enable them to

develop and/or review content for their markets. This

approach, known as central oversight with local
empowerment, has

proven to be one of the most powerful assets to companies

such as HP, GE,

Xerox, and Dell, all of which have globalized effectively.

It should also be

clear from this discussion that this approach is fully

complementary to the

principles of the Three Pillars. Armed with these best

practices, companies now have the knowledge

and direction to enable them to develop and manage

multicultural con

tent sustainably. The following sections go into greater

detail into how

these best practices can be implemented within the

enterprise to generate

financial success.

THE DETAILS OF IMPLEMENTING BEST PRACTICES

So how does an enterprise, division, or group use the

concepts of best

practices to move itself closer to the vision of an ideal

enterprise? The fol

lowing sections detail the steps a company should take with

each of the pil

lars in order to implement a content globalization

initiative with a high

probability for success.

Pillar I: Strategy

At the highest levels, the strategy for a globalization

project is quite a com
plicated undertaking, as it incorporates large amounts of
financial analysis,

risk factors, competitive issues, and so forth. These

issues are not in the

scope for this chapter or this book, so I assume that the

business globaliza

tion strategy is in place prior to the development of the

content globaliza

tion strategy. To simplify the discussion of the content

globalization

strategy, consider it as moving through three distinct

phases: (a) definition

of goals and scope, (b) communication and cooperation

between stake

holders, and (c) road map and metrics. Definition of Goals

and Scope. The first step in creating a strategy

for a content globalization initiative is to synthesize

from the business

strategy three primary constituents: the goals of the

initiative; the

breadth of the initiative (i.e., the number of countries

and teams in

cluded); and the depth of the initiative (i.e., the types

of products and

content to be included). As an example, consider the

following fictitious plan for a U.S. mobile

device company that is planning to increase sales in Asia:

Problem: XYZ Company has noted that the Asian market is the
largest mar ket in the world for mobile electronic devices
such as phones and pagers. However, XYZ Company has only a
small presence in Japan and Australia and some penetration
through distributorships in the other major countries.
Goal: XYZ Company will increase its market share from
current levels (< 1%) to at least 5% in the primary Asian
countries within 12 months of prod uct launch. It will
carry out an extensive launch and promotion campaign within
3 quarters by selling its key North American products. The
initiative should reach breakeven within the first 6 months
and then become profit able with a growth rate of 10–15%
afterward. Breadth of the initiative: Research indicates
that the primary Asian markets include the following eight
countries: China, Japan, Korea, Taiwan, Malay sia,
Thailand, Indonesia, and Australia. Teams from the two
offices, Japan and Australia, as well as those in the
distributorships will be leveraged. Other offices will be
set up over the course of the project. Depth of the
initiative: Three key products and all their supporting
collateral will be launched into the markets. The company
home page, product, and support sections as well as two
commerce applications will also be globalized to support
the launch and ongoing marketing and sales activities. With
the goals, depth, and breadth of the project defined,
enough infor

mation exists about the initiative to begin to communicate

about it with the

other country teams. It is at this point that the

stakeholders for content glob

alization can be defined, contacted, and brought into the

strategic process. Communication and Cooperation Between
Stakeholders. Choosing

stakeholders is a critical step that can have political

ramifications if not car

ried out wisely. To begin, it is recommended that a senior

individual from

the group leading the initiative create a committee to

represent the global

ization project. After the group has been created, it

should be officially rec

ognized and given a charter by members of senior management

in order to

imbue it with legitimacy. At this point, invitations should

be made to international deputies to en

able each country office to provide input. Individual

country stakeholders

should be chosen on the basis of their position and ability

to create action in

their organization. Alienating key individuals who may be

threatening or

including individuals with little power is not recommended.

To kick off the strategic session, it is recommended that
all important

stakeholders be brought together so that they can meet and

discuss with

each other in person. Such a level of interaction is

critical at the kick-off

of the project but may not be necessary on an ongoing basis

unless the

initiative is very large, politically charged, and/or

complex. Clearly, how

ever, ongoing interaction and communication is critical to

enable the

team to work through its issues and develop a plan

outlining the recom

mended strategy. Time should be set aside to deal with the

concerns of each office and de

velop solutions that are both practical and applicable to

as many countries

as possible. Should problems persist or should issues

escalate beyond the

capabilities of the team, it may be necessary to go to

outside experts for help

and a path forward. Consider once again, the example of the

North American mobile device

company launching its products in Asia: Consider that the

CEO gives the mandate to the Senior Vice President of
Worldwide Marketing, Nancy Liu, to carry out the marketing
side of the product launch. Nancy would then invite her
marketing counterparts in corporate and product marketing,
as well as channel partners, distributors, and information
services, to be stakeholders in the initiative. With a
significant company presence in Japan and Australia, these
two offices would clearly have input on the marketing and
technology aspects, and so several members would be invited
from each team. Other stakeholders would be drawn from the
one key distributor for the remaining countries plus the
operational lead from the team in the new Chinese office.
Issues to deal with would include a strong belief from the
Japanese and Australian offices that they are
self-sufficient, not needing any help or input from the
United States. The distributor would be interested in
charging for additional overhead associated with the
launch. Meanwhile, the Chinese team would claim not to have
time or resources, but would most likely hire consultants
to complete the task. Part of Nancy’s role would be to push
for a strategy and process that incorpo rate COLE. She
would need to work with the offices to have them help each
other and buy into a single overarching framework for the
project. Addi tionally, all major cultural, linguistic,
legal, technical, and regulatory issues would need to be
discussed to ensure that these would be taken into account
in the process. Once a solution is reached, the team would
document the strategy and put forth a road map for its
execution. Creating metrics and a road map enables the team
to measure their per

formance and schedule out the use of resources. It also

enables the team to

synchronize their efforts more effectively and work toward

common goals. Road Map and Metrics. A team-developed road
map is critical to the

success of the initiative because it ensures that the

global team is focused on

the same goals during a well-defined time frame. Milestones

and deadlines

keep the project on track and facilitate monitoring of team

performance

and the overall viability of the initiative. Due to this

structure, it is also less

likely for team miscommunication to occur as the objectives

are well docu

mented and mutually agreed on. Metrics should be put in

place both for the deployment of the initiative,

often known as the development phase, and the ongoing

operations, or

maintenance phase, to provide visibility to the project’s

success. Different

metrics may be of value to different initiatives and

organizations; however,

there are three metrics that should be applicable in most

circumstances: 1. On time/on budget metric—this metric is a
measure of how well the team conformed to its projected
timeline and budget for initiative deployment. 2. Return on
investment (ROI) metric—this metric measures the financial
return as a function of the money invested in the
initiative. 3. Customer loyalty index or similar—this
metric measures customer perception with regards to how
well the company’s initiatives and behavior coincide with
customer desires. Results from these three metrics allow
the team to understand which ap

proaches were successful and which were not and to develop

its own specific

best practices regarding meeting customer needs,

fine-tuning intracompany

collaboration, and optimizing automation of its

globalization processes. As an example, consider once again
the North American mobile device

company launching its products in Asia: After a number of

good international meetings, Nancy Liu is able to under
stand the dynamics between the different offices and decide
on the identi ties of her key stakeholders. Maintaining a
good dialogue between these individuals or teams, she then
leads the crafting of an outline of tasks, re sources, and
responsibilities that will evolve into a road map and
milestone chart for the initiative. Under this dynamic, the
team would very quickly realize that they are all im
portant to the success of the project but that they must
all compromise if the project is to be implemented. For
example, the Australian and Japanese of fices may feel that
they are self-sufficient but will quickly realize that
resourcing needs, project costs, collaboration issues, and
interoffice syn chronization is best shared with Nancy’s
team at headquarters and other of fices. Meanwhile, Nancy
will see that leveraging the other offices as much as
possible will speed time to market and reduce
responsibilities and require ments of her team. Once the
road map is laid out, milestones defined, and the metrics
put in place, the teams are now on the same page and can
move to the detailed cost ing, resourcing, and timing
issues. Once these issues are resolved, the team can now
move forward with understanding the limitations and
strengths of their infrastructure. With a strategy and road
map in place, the team is now clear on how it

would like to proceed. Before actual implementing, however,

it is critical

that the team evaluate its infrastructure and process to

ensure that it can ac

tually handle the demands of its plans.

Pillar II: Internationalized Infrastructure

Internationalization as Best Practice. Most companies,
unfortunately,

are unaware of the true capabilities of their

organizations, both good and

bad, with regard to the readiness of their people, systems,

and processes to

function internationally. This is a striking assessment

given the vast numbers

of global companies; however, it is critical to realize

that most global compa

nies are actually a loose federation of separate, local

entities. With this vision in mind, it should be clear that
most global entities de

sign products and services for a particular market, such as

the United

States, and then ask each additional targeted market to

create or customize

similar offerings locally. This process is highly

inefficient as little is gener

ally leveraged from the original products or services. As a

result, costs for lo

calization are high, time frames for launch or update are

long, and quality
can be questionable. The topic of product localization and
internationalization has been cov

ered in depth earlier in this book; however, it is

important to note that inter

nationalization and localization do not only cover product

design and

coding. Actually, internationalization is an issue that

pervades every part of

an organization, from the copy written for a manual to the

topics of meet

ings in boardrooms. Internationalization is a best practice

that, once implemented across

people, systems, and processes, allows a company to

collaborate globally

and harness the full potential of its worldwide presence as

a competitive ad

vantage. Without it in place, the costs and resourcing

necessary to work with

incompatible systems that will corrupt and destroy data can

eliminate any

project ROI and make success nearly impossible. Consider,

for example, a very large and respected U.S. company that
de

signs and builds commerce systems based on the EDI

standard. Prior to inter

nationalizing their product infrastructure, they developed

a U.S. product

and then proceeded to send it to overseas offices to

localize in country. After a

few releases, the company realized that these had

multiplied to over 30 re

leases in the field, none of which shared the same code

base, contained the

same feature set, or were well documented internally or

externally. To add to the confusion, each of the company’s
offices indicated that

even after extreme customization few if any of the products

were well ac

cepted in their local marketplace and had very low uptake.

Suffice it to say

that internationalization helped to consolidate all

products worldwide onto

the same code base and rapidly increased product acceptance

and sales in

all countries offered. Implementation Plan. Once a project

has been fully scoped, it is possi

ble to understand if the infrastructure is in place to

support its implementa

tion. For example, a checklist should be developed which

contains sample

questions such as the following: • Do all systems,

processes, and people support the creation, transport, and
storage of content for multiple target markets in different
languages? • Is the concept of country or language
supported by each component of the system? • Can the system
accommodate delivery of different content to different
markets based on user requirements? • Is all text separated
from code for ease of localization? • Has all concatenation
(splicing together of parts of a sentence to create a full
sentence) been removed from the infrastructure? • Can
different weights and measures as well as date format be
stored and displayed? • Do authors create content for an
international marketplace or a particular country? • Has
the content work flow process been developed to include
trans lation as well as in country content development? The
main purpose of these questions is to uncover areas of the
infrastruc

ture that may prove problematic to the implementation of

the overall vi

sion. As an example, consider that after the strategy has

been developed,

Nancy Liu and her support teams carry out an audit on their
assets. After a comprehensive audit, the team discovers the
following issues that

need to be resolved: • The operating system for the mobile

devices does not support dou ble-byte characters (such as
Chinese, Japanese, and Korean). • Datasheet content has
been written in Quark, and it is unknown how to make this
information available in Asian languages. • The Web sites
in each country (United States, Japan, Australia, and
China) are operating using very different content creation
and publishing environments. • 50% of existing U.S.-based
web applications do not support dou ble-byte characters,
contain no fields for important Asia informa tion, do not
support multiple currencies, dates, and address formats,
and use a significant amount of concatenation. • U.S.
product colors and cultural icons are inappropriate for all
tar geted markets and will have to be reworked. • The
primary channel for content distribution in Japan is
I-Mode, a mobile device format for which the company has
very limited content. • There is significant project risk
as the team has limited experience with simultaneously
launching and maintaining products in differ ent countries
with different languages and cultures. • There is no
infrastructure in place for coordinating and updating team
members on a worldwide basis with regard to important
activities and milestones. With this information, Nancy and
her team are able to create a business

plan, requesting budget for the internationalization and

localization pro

jects that they need to undertake. They also can create

specific Requests for

Proposal (RFPs) and submit them to third-party vendors that

have exper

tise in internationalization and localization. Given the

team’s lack of experi

ence in these areas, the third-party providers should

enable the team to hit

their aggressive timelines. Subsequently, the team is also

able to schedule in

realistic times and phasing for the completion of these

initiatives prior to

product launch. With a clear understanding of the range of

implementation challenges
that they need to overcome, the team can now plan and
prepare for the on

going life of the globalization initiative. It cannot be

overstressed that the

ongoing management of a globalization initiative is the

greatest challenge

to a team or organization. Planning and implementation have

specific

deadlines and a limited scope; however, the ongoing project

is one of con

stant change and growth. Managing, adapting, and

controlling the ongo

ing global processes is the ultimate key to success.

Pillar III: Global Processes

Efficient global business content processes typically rely

on localization.

There are two primary reasons for this. First, people are
three to four

times more likely to purchase products when they are

presented with in

formation that is in their native language (DePalma, 1999).

Second, con

tent localization is on average one third the cost of

content creation

(Drakor, 2001). Here it is important to note that

localization is appropriate where source

content can be leveraged into a new market and still be

appropriate. So, for

example, a global press release regarding a company’s

record quarterly

earnings would be appropriate as it is of global interest.

However, informa
tion regarding a product that is only available in the
United States and has

limitations that make it unusable in other countries (i.e.,

mobile devices that

do not conform to local protocols) would not. Additionally,

it should be kept

in mind that information that does not conform to a

consumer’s preferred

environment (appropriate currency, voltage, cultural

attributes, tax and le

gal system, government regulation, etc.) can make purchase

of products

and services very difficult or impossible. It is also

important to understand

that content publishing for global markets may rely on a

hybrid approach in

which specific information is localized and then combined

with locally cre

ated content. Reasons for such a hybrid approach could

include (a) im

proved management of a company’s global brand and messaging

while

simultaneously ensuring that information is locally

relevant; or (b) better

utilization and leverage of enterprise-wide resources.

Consequently, before tackling globalization processes, I
first explain

conventional content localization processes as they nearly

always form the

foundation for a globalization initiative. Conventional

Localization Process. Conventional localization process

flow includes tasks such as the following (see Fig. 5.1):

1. Change detection on source content repositories. 2. Code
and content separation. 3. Content and format
normalization. 4. Translation. 5. Translation memory and
terminology leveraging. 6. Contextual review. 7. Work flow
management and monitoring. 8. Synchronization of content
with other locales. 9. Quality assurance.

10. Mapping of localized content to target repositories. As

should be clear from this list, and as has been borne out
by extensive

analysis, localization costs are generally about three to

five times the cost of

translation alone. Each of the steps in the previous list

can be a very time-intensive and gru

eling task if carried out manually under tight deadlines on

an ongoing ba

sis. As volumes increase, timelines tighten, and the number

of target

markets increases, costs escalate exponentially due to

rising resource

needs, management overhead, and increased human handling,

interaction

and environment complexity. If complexity increases to the

point where

the localization process is too slow to meet deadlines or

too inefficient to

create relevant content, then a return on investment (ROI)

becomes impos

sible and the investment will never be recouped. The key to

localization process savings is taking situations that are
com

plex and highly leveraged and simplifying them so that they

become simul

taneously routine and rapid. A combination of process

streamlining and

automation of the tasks in the list above (see Fig. 5.2)

can lead to a 15–25%

reduction in costs for localization of new content, and an

80–90% reduction FIG. 5.1. Localization process.
in localization costs associated with content updates. A
good rule of thumb

is that even for automation of small projects, say 100

pages of content lever

aged into three locales with 10% changes per month, cost
savings should av

erage hundreds of thousands of dollars on an annual basis.

Once again, consider Nancy’s team at the point where they
are develop

ing their plans for the ongoing support for the

globalization initiative. Spe

cifically, the team must have a plan in place that

addresses content,

information, and support needs post launch. Nancy already

has understood that decentralized management and op

eration of the Asian launch will be very expensive and

prone to failure. As a

result, she has developed a plan based on Central Oversight

with Local Em

powerment (COLE). Under such a plan, she has centralized

her primary

technology infrastructure, project management, and project

drivers in the

United States. Each country office in Asia will be

responsible for content re

view, local content creation, and, in some cases, content

translation. FIG. 5.2. Streamlined process. Given the many
different types and the large volumes of content that will

need to be localized rapidly on an ongoing basis, Nancy

explores different

technology options for boosting efficiency and quality. To

this end, she be

gins a competitive assessment of localization firms,

translation tool provid
ers, and globalization management system (GMS) providers.
By choosing

an appropriate partner, she hopes to ensure that she can

coordinate and le

verage her critical corporate information assets on a

global basis and consis

tently hit aggressive localization deadlines. Nancy also

creates a competitive process to review vendors for the man

agement of her source content assets. To this end, her team

reviews enter

prise content management systems (ECM), portal solutions,

and source code

control solutions. By combining an appropriate solution for

content creation

with an infrastructure for global content leverage, Nancy’s

team can be more

confident that they will be building on best practices and

generating effi

ciency and scalability through process automation and

streamlining. The next sections explore in detail the most
critical implementation as

pects for generating and maintaining success during the

ongoing growth

and change of a globalization initiative.

WORK FLOW AND PROCESS AUTOMATION

Content change is typically the initiating action in any

globalization pro

cess, however, there are many other steps that must take
place before a

globalization process is complete. A typical globalization

process can be

subdivided into four primary stages: (a) process initiation

and content ac
cess, (b) content preprocessing, (c) human-based work flow
activities, and

(d) content postprocessing.

Content Accessibility

Digital content can be stored in one or both of two ways.

It can be stored as a

file in a file system (e.g., a Word document), as data in a

database (e.g.,

product descriptions as cells in a database table), as a

combination of these

two approaches (e.g., an ASP, JSP, or Cold Fusion

document). As a result,

before a document, web page, or specific content string can

enter a global

ization process, it is imperative that first the content be

accessible. Access can be complicated in multiple ways. The
most common ways

that content access can be rendered difficult include

proprietary storage

formats, distributed storage, undocumented or poorly

documented stor

age, proprietary presentation formats, access controls

(ACLs), encryp

tion, encoding, and access through other applications. In

some cases,

companies reduce content accessibility and visibility to

improve security;

in other cases, it is done to improve content or process

management,

whereas in still other cases, it may be simply accidental

as a result of an im

plementation approach. Uncovering your company’s or

client’s particular challenges with regard
to content accessibility is critical for the development of
any sustainable

globalization solution.

Content Preprocessing

Digital content can be acquired in a variety of formats.

These can include

popular word processing formats, such as Word or

WordPerfect; powerful

business presentation formats, such as Quark or Framemaker;

Web-centric

formats, such as HTML, XML, and Javascript; coding formats,

such as

C++, Perl, and Java; or potentially database formats

accessible via SQL. A key point to note is that any format
can be represented as a combina

tion of code and content. The code typically gives

instructions regarding

presentation and potential functionality whereas the

content is a combina

tion of textual information plus graphics. In a

localization process involving nontechnical translators and
reviewers,

it is generally unlikely that any changes will need to be

made to the code. As a

result, to streamline a localization process, it is

critical that content be sepa

rated from the code to simplify the tasks of translation

and review. There is a catch, however, for without the
presentation layer or the func

tionality, the context for the content may be lost, thereby

complicating lo

calization tasks. As a result, content to be localized

needs to be isolated while
at the same time retaining contextual information. This
approach maxi

mizes productivity while minimizing the need for rework. In

addition to format, content can be acquired in different
language

encodings. So, for example, content acquired in the

Japanese language will

often be encoded as shift-JIS, whereas content in French

would be encoded

as ISO-8859-1, and content captured in simplified Chinese

would be en

coded as GB2312. As a result, language translation for

digital information

leads to a transformation of the encoding as well. To

simplify the processing of language, it is recommended as a
best prac

tice that all content to be processed be first converted to

Unicode. Unicode

is a single encoding for all the world’s languages which

simplifies language

translation by limiting all language processes to a single

encoding with

well-defined rules. To summarize, there are six steps which

must take place

to make digital content ready for processing: 1. Separation

of code from content. 2. Storage of code for contextual
presentation/functionality. 3. Conversion of content to
Unicode. 4. Segmentation of content into sentences and/or
paragraphs. 5. Separation of content into translatable and
nontranslatable seg ments. 6. Word count for
budgeting/timing purposes. Additionally, it is recommended
that all content not only be stored as

Unicode but be embedded into an XML-based “container.” This

container

should contain metadata (information about the data) for

preserving the in
tegrity and fidelity of the document being processed under
a round-trip oper

ation. That is, if I begin with a Word document, I want to

ensure that I get a

Word document back with the same formatting as the original

after localiza

tion; the same should be true for a Quark document, an HTML

page, or a JSP. XML is recommended for the container
format, as it is a neutral format that

is rich enough to preserve the formatting for the most

complex documents. To

maximize productivity, it is also recommended that all the

activities involved

with the preprocessing of content be automated as much as

possible.

Human-Based Work Flow Activities

Once information has been preprocessed, it is now possible

to route it into a

human-based work flow for translation, review, and

potentially local con

tent additions. Manual work flow processes contain many

bottleneck points

that lead to high process inefficiency. Automation of work

flow processes,

on the other hand, can greatly streamline operations

leading to much

higher throughput, better productivity, and higher quality.

Business processes typically exist within companies;
however, they are

usually poorly defined or are often reinvented each time

they are repeated.

For example, a company may publish press releases every

week and then

distribute versions in 10 different languages across 18

countries, or a com

pany may make daily updates to its online product catalog

and localize

these updates into the European marketplace. In each case,

there are par

ticular initiating activities, which should then be

followed by set processes

(i.e., distribution into certain languages and

marketplaces). Unfortunately,

the necessary processes often are not set, and lack of

clarity leads to mis

takes, missed deadlines, and poor quality. Automation of

work flow processes is not always possible. For automation

to be effective, there are a minimum of two prerequisites

that should be in

place. First, content under a globalization initiative must

be categorized or

easily categorizable. Second, business rule templates and

an infrastructure

for initiating business rules must be put in place. Both

are necessary, for

business rules cannot be initiated without content

categorization and con

tent categorization alone cannot lead to automation of

processes. Categorization is simply a fancy term for the
type of content, for exam

ple, a press release or a product catalog item. With this

descriptor in place,

it is then possible to associate the content with a

particular business pro

cess. A codified version of a business process is typically

referred to as a

business rule. So, for example, consider the press release

scenario out
lined earlier in more detail.

Press Release Business Process

Every Thursday, an English-language financial press release

is created

for distribution for the following Tuesday in 10 languages

for 18 mar

kets. The original document is created in Quark format and

is two pages

long. Once the press release is created, it is then sent to

five teams in suc

cession, these are: • An editor who checks the style and

formatting and ensures that the content is globally
relevant and easily translatable. • Translators with
particular language expertise and expertise in marketing
and financial translation then translate the content. • In
country marketing managers then review the press release to
ensure that messaging is consistent with their country. •
The in-country legal team then checks the content to ensure
that the content will not create any legal exposure. • Then
information is given a final review by an in-country public
relations team representative. Once this process is
understood, it can be turned into a standard business

rule template that can then be initiated automatically. The

business rule can

be summarized as follows: • Every Thursday initiate a work

flow process for new content in Quark format in the
financial press release folder on a particular server. •
Route this content through five steps: de-Americanize,
translate, review, legal, and final review. • Associate
relevant individuals with each step of the work flow
process. • Content must be fully processed by Monday of the
following week. Now, with a combination of content
categorization and business rules,

the process of globalizing this press release can be

carried out automati

cally on a weekly basis. Obviously, some level of human

supervision is nec

essary; however, all of the manual hand-offs, phone calls,

format

problems, and second-guessing associated with the fully

manual process

are no longer necessary.

Content Postprocessing

The final stage of the process is often termed

postprocessing. It is in this

stage that the content is put into its final form prior to
distribution. Postpro

cessing usually occurs across three axes: format, encoding,

and location. So,

if as in the previous example, the press release begins as

a Quark document,

it should complete its localization process also as a Quark

document with

the same formatting as the original. For print and web

documents, there

can be significant reformatting work necessary as different

languages gen

erally have different linguistic rules and take up

different amounts of space

on a page. So, for example, German, French, and Japanese

translations of

English content often show 20–50% horizontal text expansion

or text swell,

whereas Chinese and Korean typically see horizontal text

contraction and

vertical text expansion. Such changes in text size often

require different formatting for such

items as pagination, table sizes, text boxes, and so on.

These changes can be

made manually postprocess, or they can be facilitated

during document lo
calization during the work flow. In any case, formatting
will change between

the original language version and the local language

version; thus, modifi

cations will need to be made. Given that the language has

gone through a translation process, it is criti

cal that the encoding for the content be appropriate for

the language. It

should be noted, that though Unicode is a good format for

processing mul

tilingual content, it is very rarely a good format for

exporting content for

distribution. The reason for this seeming contradiction is

that today, very few stan

dard desktop applications use Unicode as a native content

format. As a re

sult, Web-based content and that for Quark, Frame, and Word
must, in

most cases, be converted to national language encodings

(NLEs). NLEs are

the native encodings that most desktop applications use to

render different

languages on an electronic device. Common NLEs include Big5

for com

plex Chinese, KOI8 for Russian (Cyrillic), and Latin-1 for

French. In addition to format and encoding, postprocessing
must also specify a

storage location for the content. Storage location

determination is critical

for process optimization because without it, the

information will either be

lost or could overwrite other important information.

Storage location infor
mation should specify not only which repository the content
is to be stored

into, but also a unique address point within this

repository that follows some

sensible content categorization scheme. By streamlining and

automating the preprocessing, work flow, and post

processing stages of content globalization, a team can

rapidly deliver a re

turn on investment to a globalization project. Another

critical factor, which

can greatly reduce costs and time and improve quality, is

the use of linguis

tic technologies. The types and benefits of linguistic

technologies are cov

ered in detail in the following section.

LINGUISTIC TECHNOLOGIES

Today, there exist three broad classes of linguistic

technologies that can be

of benefit to content globalization initiatives: (a)

translation memory (TM)

engines and repositories, (b) terminology management

systems, and (c)

machine translation (MT) engines. Each of these

technologies has the potential to be employed singly or in

tandem to reduce costs, speed time to market, and improve

quality. The fol

lowing subsections explain each technology in more detail.

Translation Memory (TM)

In its simplest form, translation memory, or TM as it is

commonly known, is a

matching between a sentence and its human-translated

equivalent in another
language. A TM is often built as a consequence of the
translation process, with

more “matches” or “source-target pairs” created as

translation proceeds. Over

time, a repository of such pairs can be mined or

translations “remembered” to

reduce the cost and time of translation while increasing

consistency. The power of TM relies on the “leveraging” of
past translations to reduce

the amount of current and future translation. Experience

has shown that

approximately 20-30% of a single Web site or a large

document can contain

redundant information. As a result, a TM produced from a

single site or

document can reduce costs and workloads by at least 20%. A

realistic limit

for a good TM produced over a few years is that 50% of

content can be lever

aged directly from the repository. With the addition of a

fuzzy TM match

ing engine, or an engine that can deduce matches based on a

percentage

similarity, workloads can be reduced by an additional 50%.

Additional factors that can increase the power of a TM
include con

trolled language authoring and the application of the TM to

content on the

same subject or produced consistently for the same team.

Controlled lan

guage authoring increases the efficacy of TM by ensuring

that content is

constructed linguistically the same way each time the same

concept is writ
ten. In contrast, application of the TM to similar content
increases efficacy

by ensuring that the concepts to be leveraged are similar

from the start. Interestingly, due to its reliance on human
translation, TM is considered by

many to be the most reliable and useful linguistic

technology that exists today.

Terminology Management

Terminology management technology, or glossary management

as it is

sometimes known, is similar to TM; however, rather than

dealing with sen

tences or paragraphs, it is concerned solely with words or

short phrases. Ad

ditionally, terminology is often a linguistic asset that is

dictated prior to a

translation process rather than being a by-product of it.

Terminology is ei

ther created by a particular company’s marketing department

or it is dic

tated by usage within a particular industry. To be useful,

a terminology management system must be able to do more

than aggregate information. As terminology is a by-product

of marketing

or of an industry, it is important that the teams that use

it can edit it and

change it over time as terms are added, deleted, or

modified. Furthermore,

the terminology management system should enable the teams

to add sen

tences to specify context, where appropriate. Given its

nature, terminology management is a critical technology for

companies that rely on the precise usage and/or translation

of key words.
Certain examples where term usage is very important include
product

names, terminology or jargon for any industry (medical,

financial, insur

ance, technology, etc.), and menu items for applications.

In regulated in

dustries, incorrect use of terminology can cost companies

millions of

dollars in lost time and additional expenses, while in

other industries it can

lead to confusion and potentially loss of trust and/or

revenue. Integrating terminology management with TM is seen
within globaliza

tion as a best practice for maximizing consistency and

translation quality

while simultaneously holding down costs.

Machine Translation

Machine translation (MT) is one of those technologies that

appear on the

surface to have tremendous potential, but is rarely of

significant value in

practice. The concept behind MT is that a computing device

should be able

to “read” a sentence and then translate it correctly into

another language

based on a complex integration of grammar rules and a

cross-language lex

icon. Were this theory accurate, the result would be that

translation could be

done at a rate of hundreds of words per minute for pennies.

Unfortunately, from a technology point of view, language is
most often

too complex, nuanced, and illogical for today’s technology

to accurately in

terpret between languages. As a result, MT is typically

relegated to situa

tions in which information is highly technical and very

specific in nature,

such as in a technical manual, or is of questionable value,

such as say in

“chat” between consumers on a Web site. Though it is

tempting to implement MT systems, experience has shown

that low-cost systems create more problems than they solve,

while the costs

associated with implementing “high-value” MT systems can

exceed the costs

for human translation. Though this is the situation today,

I would not count

MT out entirely, but instead believe that as computing

power increases, there

will be significant technologies that will be produced in

the next few years.

These technologies will combine TM, terminology management,

and MT to

great benefit of the localization and globalization

community.

TRANSLATION SERVICE PROVIDERS

Paradoxically, though the skill of speaking multiple

languages is nearly

universally applauded as a significant accomplishment, good

translation is

often seriously undervalued by most companies and

organizations. Poten

tially, this seemingly contradictory situation is fostered

by the false belief

that being able to speak a second language is sufficient

for being able to

translate between two languages. Actually, good translation

is an art that is a capability of only a small per

centage of bilingual speakers. In addition, there are very

few translators

that can be considered experts across two languages;

instead, translators

are categorized by their strengths and experience. For

example, a particu

lar translator may be highly specialized in translation

between German and

English for the telecommunications industry, whereas

another may be an

expert in legal interpretation between French and Spanish.

On the global scene, it is arguable that more wars,
misunderstandings,

and conflicts have arisen from poor or incorrect

translation exacerbated by

different cultures than from any other source. As a result,

translation should

not be taken lightly and should, in most cases, be provided

by experts. Keep

in mind that your brand, your image, your messaging, and

your revenue in

another market may rely on the quality of your translators.

Also, if you have

adopted a model of local content creation in international

markets, note

that translation is three times less expensive, on average,

than native-lan

guage authoring. Translation firms abound in many

varieties, from single language pair

specialists to huge full service operations employing

thousands of people.
In determining the appropriate translation firm for your
needs, keep in

mind four key parameters: • Translation expertise,

including language pairs and specialties (au tomotive,
medical, telecom, etc.). • Quality and approaches to
maintaining quality. • Cost per word or per project. • Turn
around time. To keep your choice honest, keep these
benchmarks in mind: • A typical translator can translate
2,000–2,500 words per day. • A typical reviewer can review
approximately 10,000 words per day. • European languages
typically cost US$0.20–0.30/per word. • Asian languages
typically cost US$0.30–0.40/per word. • A web page costs
approximately $100–200 to fully localize. Also of
importance, if quality or multicultural issues are of
critical impor

tance, be willing to pay significantly more than the rates

quoted here. Note

that there are hundreds of stories of companies spending

pennies to trans

late their brand into another market and losing millions of

dollars due to

poor perception, misunderstanding, and confusion.

OVERVIEW AND CONCLUSIONS

Though the complexity of content globalization is often

either underrated

(incorrectly believed to be synonymous with poor

translation) or overrated

(considered impossible), the implementation of best

practices to globaliza

tion initiatives can enable teams to attain significant

return on investment

in short time frames. By intelligently combining planning,

an international mindset, global

teamwork, and technology, the world’s best companies have

leveraged their

global operations to yield significant competitive

advantage. Though the

road to success is lined with many challenges, persistence

and knowledge

can turn content globalization into a cornerstone of your

company’s long

term sustainability and growth. Armed with the concepts in

this chapter, you are now empowered to walk

this road with confidence, avoiding its pitfalls, and

attuned to the concepts

that will generate success and advancement.

DePalma, D. (1999). Strategies for Global Sites. Forrester

Research Report, For rester Research.

Drakos, N. (2001). Blueprint for Globalization ROI. Gartner

Research Report, Gart ner Research.
DESIGN ISSUES AND USABILITY ENGINEERING

Axtel, R. E. (1991). Gestures: The do’s and taboos of body

language around the world. New

York: Wiley.

Berlitz, C. (1982). Native tongues. New York: Grosset &

Dunlap.

Boiarsky, C. (1992, September). Strategies for successful

international communication:

Using cultural conventions to inform your documents. Paper

presented at the Record

of the International Professional Communications

Conference, New York.

Center for Universal Design. (1997). What is universal

design. Retrieved May 20,

2002, from http://www.design.ncsu.edu/cud/univ_design/ud.htm

Cooper, J. C. (1978). An illustrated encyclopedia of

traditional symbols. London: Thames

& Hudson.

Deregowski, J. B. (1974). Pictorial perception and culture.

In Image, object, and illu

sion (pp. 79–85). San Francisco: Freeman.

Dreyfuss, H. (1984). Symbol sourcebook: An authoritative

guide to international graphic

symbols. New York: Van Nostrand Reinhold.

Fernandes, T. (1996). Culture4 and design in international

user interfaces. New York:

Wiley.

Forslund, C. J. (1996). Analyzing pictorial messages across

cultures. In D. C. An

drews (Ed.), International dimensions of technical

communication (pp. 45–58).
Arlington, VA: Society for Technical Communication.

Frye, R. H. (1981). Artistic technical training. IEEE

Transactions on Professional Com

munication, PC-42(2), 86–89.

Fuglesang, A. (1982). About understanding: Ideas and

observations on cross-cultural com

munication. Uppsala, Sweden: Dag Hammarskjold Foundation.

Gombrich, E. H. (1969). Art and illusion: A study in the

psychology of pictorial representa

tion. Princeton, NJ: Princeton University Press.

Grove, L. K. (1989). Signs of the times: Graphics for

international audiences. Proceedings

of the International Professional Communications

Conference, New York.

Hagen, M., & Jones, R. (1978). Cultural effects on

pictorial perception: How many

words is one picture really worth? In R. D. Walk & H. L.

Pick (Eds.), Perception and

experience (pp. 171–212). New York: Plenum.

Halio, M. P. (1992, October). Helping users navigate in

multimedia documents: The affec

tive domain. Proceedings of SIGDOC’92, New York.

Hall, E. T. (1959). The silent language. Garden City, NJ:

Anchor Press.

Hao, R. (1999). Really—How do Asians learn? Performance

Improvement, 38(3),

13–15.

Hartshorn, R. W. (1989). Writing for International Markets.

Paper presented at the

Proceedings of the 37th Technical Writers Institute, Troy,

NY.
Jones, S., Kennelly, C., Mueller, C., Sweezy, M., Thomas,
B., & Velez, L. (1992). De

veloping International User Information. Bedford, MA:

Digital Press.

Kano, N. (1995). Developing international software for

Windows 95 and Windows NT.

Redmond, WA: Microsoft Press.

Kittendorf, D. K. (1981, May). The cartoon as technical

communication. Proceedings of

the 28th International Technical Communication Conference,

Arlington, VA.

Kostelnick, C. (1995). Cultural adaptation and information

design: Two contrasting

views. IEEE Transactions on Professional Communication,

38(4), 182–196.

Kumhyr, D., Merrill, C., Spalink, K., & Taylor, C. (1994).

International report: Concepts

and procedures. Durham, NC: Internationalization and

Translation Services.

Merrill, C. K., & Shanoski, M. (1992, October).

Internationalizing online information.

Paper presented at the Proceedings of SIGDOC’92, New York.

Modley, R. (1937). How to use pictorial statistics. New

York: Harper.

Morris, D. (1977). Manwatching: A field guide to human

behavior. New York: Harry N.

Abrams.

Mrazek, D., & Baldacchini, C. (1997). Avoiding cultural

false positives. Interactions,

July + August, 1997.

Neurath, O. (1936). International picture language: The

first rules of isotype. London:
Kegan Paul.

Ogilvy, D. (1985). Ogilvy on advertising. New York: Vintage

Books.

Rahmat, O. (1999, July). I ain’t no gringo. New Media,

27–36.

Rochester, J. (1992, September). Visual aids for a foreign

audience. Proceedings of the

International Professional Communications Conference, New

York.

Spragins, E. S. (1992, October). Developing hypertext

documents for an international au

dience. Proceedings of SIGDOC’92, New York.

Thorell, L. G., & Smith, W. J. (1990). Using computer color

effectively: An illustrated refer

ence. Englewood Cliffs, NJ: Prentice Hall.

White, J. V. (1990). Color for the electronic age. New

York: Watson-Guptill. Chapter

“Sunday in Shanghai,

Monday in Madrid?!”

Key Issues and Decisions in Planning

International User Studies

Susan M. Dray

David A. Siegel

Dray & Associates, Inc., Minneapolis, MN

INTRODUCTION

That you are reading this book suggests that you already
understand why

you might want or need to do international user studies.

Suffice it to say that

success in the international marketplace requires detailed

knowledge about

your international users and the context of usage, and

evaluation of the de

sign (preferably iterative) with representative

international users. You must

go beyond the generic information that is available about

cultural differ

ences. You must go beyond basic localization of your

product, whether it is

hardware, a Web site, or software. You must actually get

empirical data spe

cific to your users and your product. This is simply

extending the principles

of user-centered design to an international context.

Although international user and usability research is

expensive, with

out this data to drive product definition and design,

products have a

much higher risk of failing in the international

marketplace. Not only is 189

the risk of a “miss” higher, but “misses” are more likely

to be total. Over

and over, we have found that international research has

provided the

type of “aha!” insights that made the difference between a

product or sys

tem being viable or completely ill-adapted to the

international con

text—insights that could not have been obtained any other

way, except

maybe by a product post mortem. Of course, it is much

easier and

cheaper to do it right the first time than to try to

recover from mistakes by

retrofitting changes to address international concerns or

issues. So al

though this research is indeed more costly than domestic

user research

for many reasons, it is also indispensable.

If you are already an experienced international user

researcher, we

hope that it will be interesting and valuable for you to

compare your expe

riences and approaches with ours, but we have a special

sympathy for

those new to international research. If you are looking

forward to your

first international usability evaluation or field research,

you are likely to

experience a mix of excitement and anxiety as you begin to

think about

the unfamiliar territory you are about to venture into. You

begin to realize

the number of new things you will have to manage and the
likelihood that

you will have to make various adjustments in your usual

procedures to

make the study “work” in another country, in another

language, in an

other culture. Even more anxiety-provoking, as you realize

how high the

stakes are, is the fear that you will overlook a major

issue in logistics or in

methodology. You definitely do not want to find this out

when you arrive

in County A, with 3 days to complete the research before

your scheduled

departure for Country B.

We can empathize with both the excitement and the concern,

and that is

why we are writing this chapter—to give you the benefit of

what we have

learned through doing a variety of user studies, including

usability evalua

tions and ethnographic research, in many countries in North

America, Eu

rope, Asia, and South America. We have tried to write the

chapter we wish

we had had available to us when we started with

international work many

years ago, and although this chapter is not a cookbook, it

will introduce you

to the issues you need to address as you plan and prepare

for studies in a cul

ture other than your own. We have organized the chapter

around the key is

sues in preparation in particular, because preparation is

the key to

international user studies, and to managing the trade-offs

that arise. We

also briefly cover some of the things you need to know

about actually con

ducting the study in-country. Of course, we focus on those

things that are

different from doing studies in your own country. In

addition, because cost

is usually a factor in planning these studies, we pay

particular attention to

planning issues that have a significant impact on cost,

both explaining what

drives these costs but also offering suggestions about

strategies to make the

research as cost effective and valuable as possible.

A Few Caveats

This chapter is based on our own experiences in doing user

studies interna

tionally and is intended to provide practical assistance.

Although there is a

vast literature on internationalization and on key cultural

differences be

tween countries, we do not review that here. Instead, we

focus on the specific

issues, especially trade-offs, that people doing

international user studies

need to address in their practical planning. We recommend

and trust that,

as a person interested in learning more about the countries

where you will

be working, you will consult the literature for more

information and

backgroud specific to those countries. This should be a

part of your prepa

ration for international user research.

Additionally, our experiences are shaped, for better or

worse, by our own

cultural conditioning. Although we try to be sensitive to

any biases this may

lead to, we are also aware that we may not always see these
clearly. There

fore, if you are not from the United States, and you find
something here
that is discrepant from your experiences, we encourage you
to contact us so

we may learn from you.

“It’s All About Preparation”

It is a truism that you can’t do too much preparation for

an international

user study. Preparation is always important, but it is

especially crucial for

conducting a study in a country outside your own. It is

challenging to make

arrangements remotely, time for problem solving when you

arrive is lim

ited, and it is unlikely you will have the opportunity to

postpone or re-do the

research with a modified plan if something significant does

not work. It

takes longer than you think it will to prepare and carry

out an international

user study, certainly longer than when you are doing

domestic studies. This

is true whether you are planning a visit study, such as an

ethnography, natu

ralistic observation, or contextual inquiry, or whether you

are planning us

ability evaluations. And even with all of your preparation,

there will STILL

be unexpected things to deal with once you get in-country.

MAJOR PLANNING AND PREPARATION ISSUES

Selecting Where to Go

When you are planning a study in another country, one of

the first questions

is often “Where are we going to test?” or “Where are we

going to visit?”

Sometimes circumstances dictate where you will have the

opportunity to do

the research, but often our clients are looking for some
rational basis for

making the decision. The question translates into some

sub-issues like how

to balance depth of coverage in one area with

representation of different

major geographical regions, and which countries are the

right ones to visit

in a given region. It is good to remember that one city in

a country is not

necessarily “representative” of that country as a whole, or

that one country

does not represent a continent. However, because you can

never afford to

go everywhere, you need some guidelines about how to sample.

Of course, in making this decision for a particular study,

you have to be

clear about what is driving the need for the research. This
will often deter

mine both the geographical scope, as well as whether the

entire scope has to

be addressed in one major study, or if there can be a

series of geographically

focused projects. Are there preexisting concerns specific

to one geograph

ical area, or is the product aimed at one particular

overseas market? Obvi

ously, this can limit the number of destinations you need

to consider. Often,

however, the product is intended for a global market, and

there is a general

need to better understand the global diversity of users and

their responses

to the design. Often, development life cycles mean that you

will have one

major opportunity for international research for a

particular development

cycle. Sometimes Web sites, such as corporate web sites,

have a long life, un

dergo frequent updating, and have different national

versions, creating op

portunities for multiple independent studies with a more

focused

geographical scope, which can be conducted over time in an

ongoing pro

gram of research. However, even for Web sites that are

undergoing contin

ual modification, there are usually limited opportunities

for input into a

major redesign. So, in these cases, it is important to make

the most of the

opportunity for international research, by making sure that

your geograph

ical sample gives you the maximum benefit.

Sampling for variation. Perhaps the key principle to

remember when de

ciding where to do your research is that when you can not

sample

exhaustively, it is always good practice to sample for

variation. This means

that you allocate your resources to sample the most diverse

locations you

can manage. Obviously, this means diversity in relation to

dimensions of

difference that you have reason to believe are likely to

influence the usage

and usability of your product.

An important nuance in relation to this principle of

sampling for varia

tion has to do with the issue of what you can and cannot
generalize from a

study. On one level, you will learn things specific to the

success of the design

in a given location, and there will of course be a degree

of uncertainty about

how much you can generalize from these findings to areas

you have not spe

cifically visited. But on another level, because it greatly

expands the uni

verse of variation compared to doing research in your home

country alone,

international research almost inevitably exposes new

dimensions of differ

ence among users that you would not have anticipated. You
may not know

where people you have not sampled fall on that dimension,

but once you be

come aware of a new dimension of difference, it becomes

part of your gen

eral conceptual schema. You can at least take steps to

ensure that it is

addressed in your product planning and design research for

other geo

graphical areas. This is a form of “generalization” that

increases the sophis

tication of your design research and thoroughness of your

product

planning for all areas. To put it in a nutshell, you don’t

have to see every

manifestation of difference—you just have to see enough

differences so you

begin to understand how broad the universe of differences

is and what some

of the dimensions might be.

As an example, many studies have taught us that there are

differing na

tional patterns in such things as use of the postal system

for sending pay

ments or receiving shipments of products of any type,

whether it is common

for packages to be left at the door, and so on, and that

these differences

seem to relate to differences in e-commerce-related

behavior and attitudes.

Once sensitized to issues like this, you can add them to

the checklist of

things you consider when designing a product for any

geographical area.

Though the principle of sampling for variation may be the

primary guid

ing principle, there are many other factors that should and
will affect your

decisions about where to go. Here are some examples, along

with some of

the considerations they may introduce into your planning.

Cost. The cost of research varies greatly from region to

region and coun

try to country. In general, this is due not simply to

differences in prices for
services among countries, but also to factors that will
influence how effi

ciently you can spend your time in the country. For

example, in some coun

tries, it is difficult or impossible to get working people

into a research

facility during working hours, so your research team will

need to spend

more time in-country for a given sample size.

Whereas actual costs depend on the actual research plan,

and generaliza

tions are possibly misleading, we can still offer some

general impressions

about relative costs. We have generally found costs in Asia

to be higher than

costs in Europe or North America, with Japan the highest

cost area. The

high cost of translation in Japan is a significant

contributing factor. Costs in

Singapore have tended to be an exception to the rule of

higher costs in Asia,

but this may be partly due to our circumstances as U.S.

consultants, as the

fact that it is often appropriate and feasible to work in

English there greatly

reduces our costs. We have found that costs in Latin

America tend to be

lower, although rental of computer or video equipment, if

necessary, can be

more expensive there than in the United States.

Obviously, you will have to take into account the differing

costs of travel
to different locations. Also, cost must be weighed against
the relative value

of going to particular places. Singapore may be less

expensive than Seoul,

but the same factors that make Singapore easier and less
expensive may

make it less interesting than a less “Westernized” city.

Finally, another

trade-off will be between the cost per location and the

number of locations

you can visit. You might be able to go to Mumbai (formerly

Bombay) and

Singapore for the cost of a similar study in Japan. In

short, you will have to

carefully weigh the trade-offs of cost versus the strategic

importance of dif

ferent destinations.

While considering costs, it is worthwhile to point out the

cost efficiencies

that can be derived from visiting more than one country per
project. While

the in-country research costs will be duplicated from one

country to an

other, there are some “economies of scale.” It is often

possible to save on

airfares by purchasing an around-the-world ticket, as

opposed to separate

round trips to several countries. Also, there will be some

economies of scale

in the planning effort for a project conducted as a single

grand tour.

Market share. If you sell a lot of products in a particular

country, this may
be one appropriate way to focus your efforts. Going where
you have signifi

cant market share will allow you to recruit people who use
your product.

Your product will also be more likely to be recognized or

known even by

those who don’t actually own or use it, thanks to

advertising and observed

usage by others. Depending on this experience, as well as

branding, and so

forth, this may make it a good bet or a bad idea to go to a

place where you

have significant market share.

New markets. If you are intending to design a product for a

new market, or

you are introducing it into a new area, you may want to

gather user informa

tion from that country or region. At the very least, it is

important to gather

ethnographic field data to help you in the earliest stages

of product devel

opment to know whether your product even “fits” into that

region, or

whether you need to make fundamental changes in your

approach. You will

also want to check out your design decisions and do

usability evaluations,

preferably in an iterative manner throughout development,

or, if you want

to live dangerously, when you are approaching a final form.

Technology penetration. You may be able to decide which

country or coun

tries to visit based on the penetration of the technology

in question, such as

cell phones or the Internet. Depending on your

circumstances, strategy

might lead you to focus either on places where the

technology is used a good

deal, or places where the technology has limited

penetration. Your market

research department probably has a lot of information on

this, especially if

you are interested in more specialized technology

penetration. For some

things, you can find out for yourself on the Web, however.
For instance, if

you are looking for Internet use, you might want to check
out

http://cyberatlas.internet.com/. This site is a guide to

online facts, gathered

from a variety of sources.

Sales or field representative support. You may choose a

country because you

have a particularly helpful (or politically important)

sales or field represen

tative there. Sales/field representatives are often

critical to recruiting, espe

cially of current or potential customers and of

enterprise-level customers. If

you do not include them, you might find that the political
support for your

study can evaporate. However, when actively enlisted in

your project,

sales/field representatives can be incredibly helpful. They

can help line up

customers for field visits, identify cultural issues that

will require localiza

tion of your test plan, help with finding equipment, and

assist in translation

by doing at least informal back-translations for you to

check the accuracy of

translated materials. They also know your company and its

positioning

within their country. It is important to educate them as to

what you need

and to help them understand what you are planning so they

buy into your

study. In some studies we’ve done, the client has chosen

the locations based

on where they had good sales/field support, and this was

critical to the suc

cess of the studies. You may want to consider identifying

ways to make it

worthwhile for them to help you by including whatever perks

and recogni

tion are organizationally appropriate.

Sometimes the interest in the study comes from one area,

such as when

local field representatives have reported user problems in

their territory. In

cases like these, a particular field office may be the

biggest advocate for user

research. Using this as the entrée to international work

may mean that you

do not sample your entire international universe of users,

but by working

with these interested field representatives, you may build

broader support

for international research and set a precedent in your

company.

Special challenges. If you know anything about conditions

in a country that

suggests your design will be challenged in a particular

way, you should

weigh that information heavily in prioritizing where to go,

even if there isn’t

high market share or if it is more costly. For instance, if

the product you are

testing might work differently with different writing

systems, you may

choose to test it in a country with a pictographic system

as well as one with a

phonetic system. Or, as we mentioned earlier, you may want

to test your

e-commerce site in countries where there are known issues

in people’s

readiness to order online.

Window of opportunity. You may find that you have a chance

to piggyback

an international study onto a trip planned for another

reason. For instance,

you may have to attend a sales meeting or go to a

conference in another

country. If you have the opportunity, you may be able to

extend your trip to

include a user study as well, thereby minimizing travel

time and cost. Alter

natively, you may be able to combine several studies in the

same trip. For in

stance, we have combined usability testing of a product

during the day with

family ethnographies at night in order to collect as much

data as possible in

a short time (Dray & Mrazek, 1996). This is extremely

exhausting, but it can

help save money. Another possible synergy that can increase

opportunities

for international research while possibly reducing the

effective cost, or even

generating additional financial support, is to partner with

marketing and

coordinate international user-centered design research with

marketing re

search. For example, as part of the same overall logistical

effort, you might

combine a usability evaluation study with a focus group

study conducted on

the same trip in the same facility.

Arranging for Research Resources In-Country

Once you have identified where you are heading, you must
find and con

tract with the resources necessary in that locale to assist

you. The solution

you choose will probably be heavily dependent on whether

your research

will be done in a facility or in the field (homes or

workplaces). Some of the

services you may need to obtain in-country include facility

rental, recruit

ing, translation (both of materials beforehand and

simultaneous translation

during the research), audio–video and computer equipment,

and facilita

tion of the actual interaction with users. Unfortunately,

working out this
part of the plan and making your arrangements may be
significantly more

difficult than when you are doing something similar in your

own country.

Even though the Web has made it much easier to locate

facilities and re

cruiting companies, the arduous task of vetting them and

figuring out ex

actly who to use can still be quite time-consuming.

There are several different strategies or combinations of

strategies for

meeting these needs, and the appropriate solution for you

will depend on

many factors, such as the nature of the research, what

resources your com

pany may already have in-country to assist you, what your

own experience has

been with international user research, and what your

capabilities are for con

ducting the research in a language other than your own.

Here we focus on se

lecting and contracting with a facility. We focus on a

scenario involving use of

a market research facility, because this involves obtaining

the most compre

hensive set of services overseas and because it is the most

common solution.

Although you may be able to find a dedicated usability lab,

commercial labs

are still rare in many countries, and we know of no

usability firms that can

provide installed usability labs in a wide range of

locations around the world.
Assuming you contract with a market research facility, you
can at least ex

pect them to be your single point of communication for

arranging most or

all other necessary resources. There are market research

firms in most

countries you might want to visit which offer one-way

mirror observation

rooms with fixed wall-mounted cameras and can provide or

arrange for re

cruiting services, facilitation, translation, and

interpretation. However,

they will typically need significant guidance and oversight

to prepare for

user-centered design research such as usability evaluation,

as compared to

interviews or focus groups, with which they are probably

more familiar.

In order to get meaningful bids from facilities, you have

to ensure that

your requirements are well understood. Pay special

attention to commu

nicating aspects of usability research these firms may be

unfamiliar with.

While recognizing that written specifications are only the

start of the

process, you must begin by creating as detailed a

specification for them

as you can, covering the research plan and requirements. Be

sure to spec

ify such things as a minimum configuration for the computer

(processor

speed, RAM, disk space, etc.) and/or Internet access

(dial-up, dedicated

or high-speed line, and minimum speeds) and audio–video

require

ments. In many market research facilities, the standard

video arrange

ments include only a fixed focal length camera. If you need

a more

elaborate setup, such as direct screen capture through a

scan converter

and a “picture-in-picture” recording with an inset of the

user, and as

suming you do not have a portable lab, the facility will

probably have to

subcontract this. Because this setup may well be unfamiliar

to them, you

must explain it very clearly, so that they can obtain an

accurate bid. Dif

ferent countries have different video standards (PAL,

SECAM, NTSC),

so it is important to specify in advance which output you

require. If it is

not possible to record in your preferred format, you will

have to plan for

conversion of the videotapes. Depending on where you are

from, it may

be less costly to have these services provided locally and

then have the

tapes shipped to you, but if it is not feasible to have

this done before you

leave, and you are uncomfortable leaving the tapes behind,

you will have

to arrange for this back home.

It is essential to do everything possible to avoid

misunderstandings about

what a facility is bidding on, and given the challenges to

communication

that are inherent to international work, you must recognize

the real risk of

this occurring. As consultants, we have rarely found the

level of detail we get

from clients, even in formal RFPs (requests for proposal)

to be sufficient for

the purpose of obtaining bids from overseas facilities.

This can be much less

of a concern when we are working with a facility we already

have a relation

ship with from previous work. However, we commonly find

that we need to

root out vagueness and ambiguity in these documents,

usually by repeated

conversations with the client. We then rewrite

specifications in as much de

tail as we can, using simple, clear, straightforward

English, and follow up

with telephone conversations to ensure that the

specifications and the na

ture of the research we want to do is understood.

Be prepared to find out that some of these requirements are

not feasible

as initially stated, or need to be adapted to the

circumstances of the region.

For instance, we have done testing of the same Web site

using high-speed

lines in some countries, and dial-up lines in others,

because the local users

in the latter would not have high-speed access to the

Internet. Often, soft

ware will not run on older computers. Be sure to address

this clearly in ad

vance so you are sure that you will be able to conduct your
tests once you

arrive. As a final example, you may find that screener

requirements are

based on assumptions about the user demographics that need

to be modi

fied for the country (more on recruiting later).

If it is not already obvious, this entire process is

extremely communica

tion intensive. This is especially true if you are getting

multiple bids from

multiple countries. Allow the time needed for ensuring that

facilities under

stand the requirements and for clarifying their responses.

E-mail is helpful,

but personal contact over the telephone is indispensable at

key points in the

process of negotiation, especially for getting a sense of

how easy the interac

tion with the facility is. Depending on where you are

trying to do the re

search, you will have to be prepared for some telephone

calls at very

unusual hours. You will also probably spend a lot of time

discussing

trade-offs among facilities with your team or client to

make a joint decision

about final location/facilities.

When you are looking for a facility to help with a user

visit study or a us
ability evaluation, here are some of the things you need to
consider:

Previous experience with this site. If you or anyone in

your organization

(such as people from a local field office) has worked with

this group previ

ously, don’t overlook their input. What was their

experience? How similar

was previous work to that being considered now? Even if no

one in your own

organization has used them, do you know anyone who has? Can
they give

Claimed experience. Have they done similar types of

studies? The concept

of usability, and even the term, is entering the language

of market research

around the world. However, often claims of experience with

“usability” do

not mean what a professional trained in usability

evaluation would expect

(Siegel & Dray, 2001), and are more likely to refer to

survey or interview re

search, where they asked about perceptions of usability,

rather to behav

ioral research. Therefore, be sure to ask them to describe

their usability

research in as much detail as they can (not so much who

they did work for,

but what methods they used) so you can assess whether their
perspective

and definition of usability matches yours. Alternatively,

it may simply be

safer to assume that you will have to supervise the

research, or arrange for a

usability consultant to do so. 1

One factor that may be a more relevant dimension of

experience to as

sess in qualifying a facility is their level of experience

in working with over

seas clients. This is important for them to be able to

understand your

general level of expectations and the degree of

coordination that will be

necessary, as well as to give evidence of a track record

meeting the needs of

international clients, such as by providing bilingual

facilitators and simulta

1 The same is true, by the way, if you are arranging for

user visit studies. True user-centered

design field research tends to be somewhat different from

typical “day-in-the-life” or

“ethnographic” studies conducted for marketing groups in

companies. Therefore, it is impor

tant to check out what methodology the group has used in as

much detail as possible.

neous translation. If this is not already an important part

of their experi

ence, look elsewhere.

A second experience factor to assess is their background in

doing re

search on technical products and software, as opposed to

household con

sumer products like foods and cosmetics. This provides at

least some

indirect evidence of their ability to arrange for resources

appropriate to a
usability study focusing on technical matters, such as
facilitators comfort

able exploring technical issues, translators who are

comfortable with com

puter terminology, recruiters who can understand the point

of your

screener requirements, and so forth.

Location. Access to public transport is critical in many

locales. How easy will

it be for participants to get to the facility? Is it safe

at all hours of the day? Are

women and men equally safe? This can influence how

successful the recruit

ing will be and may influence the schedule itself. Is it

located near a reason

able hotel, and will it be easy for you to get to and from
the facility? This is

particularly important in usability evaluations because for

visit studies, you

are typically traveling to the users’ locations, be they

homes or businesses.

Your perceptions. When you interview a candidate facility

on the phone,

you will get a sense of their fluency and understanding. It

is important to ask

who, exactly, you will be working with because it may well

not be the same

person whom you are negotiating with. Interview the people

with whom you

would actually be working and see how easily you

communicate. Look for

the facility to ask penetrating questions about your

requirements and to
identify things that may have to be adapted for the study
to work in their

context. This is both part of the value that they add, and
a way to check that

they really do understand what you are trying to

accomplish. A facility that

confidently says, “Sure, we can do this” without raising

questions or even

objecting to some parts of our initial plan does not

inspire our confidence.

Price. When you look over the bids you get back after this
careful process,

check to see if they are really comparable. Often, bids

will include slightly

different constellations of services and charges, so it can

be difficult to make

comparisons. If you have one “outlier”—either much higher

or much lower

than the others—check with that facility to make sure you

have clearly com

municated the requirements. Price alone is not a good basis

for your ulti

mate decision, but it clearly does play a role. Don’t

automatically dismiss the

high bid—they may be taking account of something that you

will indeed

have to deal with but that the others did not recognize or
identify for you.

As we have already explained, testing in some countries,

for instance, in

Japan, can be much more costly than in other locales. If

you need to test in

high-cost areas, prepare yourself or your client for

“sticker shock” and dis

cuss with the facilities what things can realistically be

cut back without com

promising the actual testing. Some facilities, often

smaller ones, may be

willing to negotiate with you to find a mutually acceptable

price.

Remember that the facility will be obtaining bids for some

services, such

as video equipment for screen-in-screen video. We have

already empha

sized the importance of making sure they understand your

requirements.

The other side of that is to make sure they are not bidding
on more than

you require. For example, we have questioned what seemed

like high bids

for video equipment, to find out that the bid included

charges for three

video staff (technician, producer, and director) to be on

duty throughout

the study. The price came down greatly when we made it

clear those ser

vices were not needed.

Modifying a Recruiting Strategy and Screener

to Reflect Local Cultural Issues

It is typical that you will find out while choosing the

facility that some as

pects of your initial plan don’t make sense in one or more

of the locales you

plan to visit. This is the first step in localizing the

test or visit plan. As men
tioned previously, you want to be specific about your
initial requirements,

but be flexible to accommodate what you find to be the norm

in that locale.

You will want to ask the facility for their input on the
feasibility of your initial

goals for recruiting.

Some of the things you want to address include the

following:

Adaptation of your screener. This is one of the most

crucial tasks in prepara

tion, and one in which there are interesting subtleties.

The facility’s recruit

ing staff can advise you on how to create a screener, or

modify the one you

have used in your own country, to find the people you need
for the study.

You may need to revise the quotas or requirements to

reflect the actual de

mographics or usage patterns in a particular locale.

Alternatively, your op

erational definitions, for instance, of “expert computer

users” or “frequent”

Internet users may need to be changed. Or the number of

years that a per

son has been online may vary depending on the country,

because the types

of activities or applications used by “experts” may vary,

or because the

Internet is far less commonly used in some places.

In one example, we were working with a screener that

required represen

tation of different defined levels of Internet experience

across socioeco

nomic levels. The facility provided statistics to show us

that in their country,

the distribution of home telephone service, and thus of

dial-up Internet ac

cess, was such that the sample we described would have to

be refocused to

ward the higher income levels. Without this information,

the client would

have been concerned about a sample that appeared to be

skewed toward

higher income levels than they had expected. Similarly, the

difficulty of

achieving balanced gender representation in different

occupational cate

gories can vary greatly from one country to another.

When you do research in more than one country, it is almost

inevitable

that you will want to compare findings across countries.

However, there is

no simple way to ensure that you are really sampling the

same people in

different countries. Rigidly insisting on a literal

translation of the screener

that you use for studies in your own country will not
ensure “equivalence”

because the meaning of that sample will differ across

countries. When the

local context seems to require changes in how you structure

your sample,

you will have to carefully consider how to be sure you are

sampling the

group you are really interested in. You will need to think
about whether

your screener concepts are absolute objective definitions,

or if they are

culturally relative definitions. Be prepared to take some

time for concep

tual negotiation with your client or team, discussing

whether adjustments

in the screener are giving you a sample that is the analog

in this context of

the sample you might put together somewhere else. Also,

remember that

this process may give you insights about real differences

in who your users

are in different countries, insights that you might want to

follow up on with

quantitative data.

It is obviously most efficient for the recruiter to assist

you in modifying

the specific requirements before the screener is translated

so you have an

opportunity to reword the screener. When the screener is

translated, the re

cruiter may reorganize it into their standard format.

Review it very carefully

to make sure that the logic of the decision rules remains

intact and that the

key concepts have not been distorted in translation or

reorganization. This

is not a place to skimp on back-translation, which we

discuss later.

Recruiting strategy. The recruiter can also tell you the

best way to find your

target users. In some countries, cold calling by recruiters

is illegal or not ef

fective. Recruiters in these countries know how to get

people to come in for

testing, however, as well as how many they need to recruit

for a given quota

of completed tests or visits. They will also know how much

you need to pay

or how else to provide effective incentives to increase

your show-rate or visit

acceptances.

Just as some recruits are more difficult in your own

country, the same is

true in other countries. The facility should be able to

advise you on the ways

to reduce the cost of challenging recruits. Sometimes, if

you can compro

mise your recruiting strategy somewhat, either reducing the

total number,

or relaxing restrictive requirements, you can make the

recruiting easier,

and therefore, less costly.

In some countries, partly due to the difficulties of

getting working people

into facilities during weekdays, there is a tendency for

samples to be overly

loaded with students and housewives. There can also be a

very blurry dis

tinction between people who are self-employed and people

who are unem

ployed. This is something that we address explicitly, by

setting quotas and

by negotiating the definitions of employment status

criteria.
If you are doing a visit study, discuss with the recruiter
whether your

plans are realistic. Will you be able to get people to

allow you to come to

their homes or offices? If not, are there alternatives,

such as doing an inter

view with them on the phone or at the facility initially to

build rapport, and

then visiting them? Can you do without recordings, video or

audio, if that is

a stumbling block? Do you need to make changes in the

length of the inter

view, or the size or composition of the visit team to

accommodate cultural is

sues? We typically find that a cross-cultural visit team is

most effective, but

you need to make sure that this will be acceptable in a

particular locale. Also

make sure whether you can have a mixed gender team. In some
places, this

may be less acceptable. Because the purpose of visit

studies is usually to un

derstand the context of the users, it is critical that you

identify ways to get

this information, even if it means modifying your initial

plans.

Scheduling. This goes hand-in-hand with recruiting. It is

understandable

that your team may want to spend as little time in-country

as possible to

minimize time and travel costs. However, in many locales,

if you want work

ing people as evaluators, you must run evaluations in the

evenings. In some

places, you can also run on weekends, but in others this is

never done. For

working people in Japan, for instance, you must plan to

schedule on the

weekend because it is difficult to schedule during the

week. In contrast, in

some Latin American countries, we have found, we were able

to do usability

evaluations weekdays and evenings, but not on the weekends.

Your facility

can help you identity a workable scheduling plan. Your

specific user study

protocol will also influence scheduling. Often, it is

possible to combine us

ability testing and ethnographic visits by using different

time slots. For ex

ample, in a project in South America, where it was

difficult to do facility

based usability evaluations on the weekend, it was fairly

easy to do home vis

its on the weekend.

If you are visiting families, you need to think about when

all needed

members of the family will be present (assuming this is

important to your

study), and what you can do to “break the ice” and build
rapport as quickly

as possible so you can get the information you need. When

we do family

studies, we often take food in with us so we can meet

earlier, over the dinner

hour, for instance. This is a common practice in some

places, such as the

United States, but is very unusual in other locales. It can

also be difficult to

make arrangements for this in some places. For instance,

when we did fam

ily visits in Germany, we had difficulty finding a

restaurant that could pro

vide food “to go” for us (Dray & Mrazek, 1996). Although
this was an

unusual practice, we found that it worked well, especially

as it was the “crazy

Americans” who were bringing the food. It might have been

more difficult

for a local person to break the norm in this way.

Planning for no-shows. Whenever you plan a usability

evaluation, you have

to plan for no-shows. Rates of no-shows vary widely from

locale to locale. In

some places, no-show rates are extremely low, so scheduling

one extra slot,

as a backup, is all that is necessary. Other places,

however, have predictably

high no-show rates. These may be influenced by specific

local conditions,

such as whether it rains (making transport slower), or what

day it is. As an

other example, payday may be a bad day to schedule because

in some places

it is common for people to pay their bills in person that

day, snarling traffic

and making travel times unpredictable.

There are a variety of ways to deal with high no-show

rates. Some facili
ties simply advise scheduling multiple extra time slots.
Others use “floaters”

as backups. In some places, facilities typically recruit

multiple people per

slot (as many as three) to deal with high no-show rates.

This can make it

more difficult to control the sample in terms of

segmentation unless they

are able to recruit multiple people from the same segment

for the same

time slot, which is unlikely and impractical to require. It

adds to the difficul

ties in exporting your segmentation model, which we have

already dis

cussed. When multiple people show up for the same time

slot, and you have

an opportunity to choose, it is therefore very important

that you carefully

review their characteristics against your quotas and

against the range of

people on the list for the time slots yet to come, to make
best use of the de

grees of freedom that remain to you.

No-shows for visit studies are less common, but not unheard
of. It is

much harder to overschedule, given the amount of time each

visit takes.

Therefore, when we are doing visit studies, inside or

outside our own coun

try, we routinely call the day before to answer questions,

check the instruc

tions for reaching the user(s), and establish preliminary

rapport. If visiting
a company or office, we also send ahead any nondisclosure
or informed

consent form so that these are not a shock to participants.

This helps to es

tablish that we are, indeed, legitimate, and makes

cancellations less likely.

Planning for extra time. When scheduling your first test or

visit, it is impor

tant to allow the team some time to adjust to a different

time zone. Usually,

we try to travel so we have at least a day, and preferably

a weekend, prior to

the beginning of the study to adjust, although sometimes

teams need more

than a single day for adjusting to both time and culture if

they have not been

in this locale before. A good compromise is to schedule an

equipment check

and dry-run for the first day. We discuss these activities

in more detail later.

In addition to initial time to adjust to new time zones, it

is important to

allow more time between sessions than you might expect to

spend in your

own country. The process of doing usability or ethnographic

studies in a

country other than your own, especially one with a language

you don’t

speak, requires much more concentration than one in your

own country,

even if you are very experienced. You will find that it

takes more time to

process what you are observing, and this needs to be

planned for. In addi

tion to the things you normally do in a debriefing, it is

important to have the

time to compare your impressions with those of the

facilitator and inter

preter(s). For example, it may be necessary to review how

the interpreter

translated some key ideas to make sure you really have

captured the sense of

the words. This is also a valuable opportunity for coaching

the facilitator or

planning any modification for the next session.

Preparing or Adapting the Protocol and/or Test Plan

A critical step is to adjust the specific methodology to

the specific locale, bal

ancing carefully the need to keep things as similar as

possible, for comparison

purposes, with the need to modify to fit local culture.

This needs to be done at

multiple levels, from the overall goals to the specific

tasks you plan to use.

Goals and underlying questions. Obviously, the first step

in localizing a test

plan is really understanding what the underlying questions

are that must be

answered and then figuring out how to get at those answers

in the context of

each culture. If you are the one who developed the test
plan, this may seem

obvious. However, if you did not develop it yourself, or

are working with a

team where there may be different perceptions of the

critical information, a
crucial first step is to become intimately familiar with
the issues the team

needs to understand, so that the focus will remain intact

when you localize

the test plan. This is always important in user studies,

but especially so in

preparing for international work. Because of all the other

factors that may

divert you, you will need an especially strong conceptual

compass.

Tasks. You may need to modify task scenarios to reflect

more realistic or

typical tasks, especially if you are not basing your tasks

on previous

ethnographic visits. At the very least, certain elements,

such as data to be en

tered, names, and/or address formats, will need to be

localized. However,

entire tasks may need to be adapted to the local context.

It is beyond the

scope of this chapter to discuss the many issues involved

in how you select

task scenarios for usability evaluations. For our purposes,

it is sufficient to

simply underline one key issue: the distinction between

choosing tasks to

represent typical user activities versus choosing tasks

specifically designed

to probe certain aspects of the design. Sometimes, these

two different ways

of choosing or prioritizing tasks do not overlap. A problem

can arise when a

task that you think of as a typical activity is actually

unfamiliar to users, un

common in their context, or presented in ways that do not

match how they

typically think about the task. We generally want user

difficulties to reflect

design issues. We do not want them to be mere artifacts of

problems with

understanding the task instructions.

For all of these reasons, it pays to discuss the specific

tasks you plan to use

in usability evaluations with the facilitator (again,

before they are translated

if at all possible). However, it is important to have

realistic expectations. Al

though the facilitator can and should help with

localization of terminology,

data formats, and so on, and although they may be able to

raise concerns

about some of the tasks at a conceptual level, do not

expect them to be ex

pert spokespeople for describing the technology usage

patterns and ways of

thinking of all their fellow citizens, especially if you

are targeting a special

ized population. For some applications it can help to have

your tasks re

viewed by a local person who knows the domain, if possible.

Of course, we

should never rely too heavily on a single informant. To

some extent, you

will have to be prepared to modify task scenarios based on

your observa

tions of how they are actually working during the study.

Learning that you

have to change your assumptions about task scenarios can in

fact be one of

the big benefits of your research.

Methodology. You may need to consider changes to your

methodology to

adapt it to cultural constraints. These include such things

as differences in

how people express their subjective experience, thoughts,

and reactions in

a social context. There are also cultural differences in

how ready people are

to express criticism or confusion, or to attribute problems

to the design

rather than to their own failings. In addition, how

inhibited people may be

in the presence of a foreigner seems to differ from country

to country. All of

these factors influence how well such basic techniques as

“think out loud”

will work. Concerns about these cultural factors are

compounded by the

need to adapt your methodology to working through the

intermediary of a

facilitator who is often not a usability professional. All

of these consider

ations have a tendency to lead people to think that only a

more structured,

less interactive, more quantitative or “metric” evaluation

approach is feasi

ble. This is unfortunate because the team may need the kind
of design guid

ance that is more likely to come from a more interactive

exploratory or

“formative” approach to evaluation than from metric

testing, which focuses

on assessing success in task completion and may be more

appropriate as a

“validation” test. Furthermore, if you are planning

ethnographic work in

other countries, you have no choice but to find a way to

facilitate rich inter

active exploration with users, despite cultural and

language barriers.

A thorough discussion of these issues and how to deal with

them is be

yond the scope of this chapter. However, if you are

interested in highly

qualitative forms of user and usability research, we urge

you not to give up

on these too readily. Fortunately, we have found that, with

careful planning,

choice of method, and preparation of the facilitator, it is

possible to use

flexible, interactive approaches, such as various

exploratory usability evalu

ation methods and, in the realm of ethnography, contextual

inquiry. For

example, in a country like Japan, commonly considered a

context where

“think-out-loud” does not work very well, we have found

that co-discovery

can be effective at eliciting verbalizations that give the

needed clues to the

user’s cognitive processes. Your facilitator should be able

to help you plan
alternative strategies (once the nature of usability
evaluation is under

stood). For example, when we first used co-discovery in

Japan, our facilita

tor helped us realize that it was important that the two

participants be of

equal social rank, or the one who was subordinate would not
speak freely.

We had the Japanese facilitator assess social rank based on

cues that to us, as

non-Japanese, were difficult to assess, but to her, were

obvious.

Working through a local facilitator is often part of the

solution to cultural

barriers, but at the same time, it introduces

methodological issues of its

own. If the test is a highly structured metric test, then

real-time interaction

with the user will be less, and the intermediary role of

the facilitator will be

more structured or more narrowly “procedural,” and thus in

a sense less in

fluential and easier to manage. This contributes to the

tendency for some

people to believe that international evaluations almost

necessitate this type

of approach, in contrast to exploratory evaluation methods.

For example, it

may be assumed that the facilitator’s role should be

limited to introducing

the test procedure, handing the user cards with written

tasks, and then in

tervening minimally and only if the user becomes “stuck.”

In contrast, we
have found that it can indeed be possible to engage in a
much richer qualita

tive evaluation, even when working through a facilitator.

This depends on

thorough preparation of the facilitator and on choosing a

workable method

for communication with the facilitator during the data

collection. We dis

cuss these issues later in the chapter.

Arranging for Translation

We have already briefly referred several times to the

importance of transla

tion. In this section, the issue of bridging between

languages is our primary

focus. Bridging between languages involves three

components: facilitation,

translation of written materials, and simultaneous

translation (also called

interpretation) of the sessions themselves.

Bilingual facilitation. We say more about the role of the

facilitator shortly,

but in relation to the language issue, here we simply want

to point out that it

is crucial for the facilitator to be able communicate both

with the team and

with the users. The facilitator has to be comfortable

enough in both lan

guages to communicate and understand nuances, or you will

have to com

municate through a translator.

We believe that all testing should be done in the local

language. Only
rarely, in our experience, has it been defensible to use
English when work

ing with people for whom English is not their first

language, such as when

working in a technical domain where English is in fact the

language of ev

eryday activities for users. However, even with people like

this, not all of the

verbalization you want to capture consists of technical

content. You want

people to express their thoughts and reactions as

spontaneously and freely

as possible, which argues for providing facilitation in the

local language.

A compromise solution sometimes attempted is to facilitate

the evalua

tion in your own language, assisted by an interpreter when

necessary. If you

intend to use the interpreter for backup only, this may

require you to skew

your sample by requiring that participants speak your

language. The idea

of facilitating in your own language, with all

communication mediated by

the interpreter is also attempted sometimes, but can be

even more prob

lematic. Not only does it greatly impair the immediacy and

spontaneity of

communication, but it is laborious, makes it difficult for

any probes to come

at the right moment, and creates distractions for the user.

Furthermore, as

we have already mentioned, in some countries, users are

much more likely

to interact freely and openly with a member of their own

culture.

These problems bring us back to the desirability of

providing facilita

tion in the local language. This creates the need for

simultaneous transla

tion, which we discuss next. Of course, it also presents

challenges in regard

to selection and preparation of the facilitator, and we

discuss this area in a

subsequent section.

Written translation. All of your materials, starting with

the screener and in

cluding all documents for the testing or visit itself, will

need to be translated

into the local language. This is typically a job for a

professional, preferably

one who is familiar with technical terms that may be

included in your mate

rials. The translator may be associated with the facility

you use, or may be an

independent or freelancer. In either case, your logistics

plan has to allow

time for this step. It is difficult to get a reliable bid

on this type of translation

because it is charged on a per page basis, and at the time

you are getting

bids, you may have only a vague idea of how many pages will
be involved. In

our experience, however, this is not an especially costly

item.

One document that is part of most protocols is an Informed

Consent/

Non-Disclosure Agreement (NDA) for the research

participants to sign. Fa

cilities typically supply their own “standard” versions of

these forms, which

might avoid the need for translation, but you should check
to make sure

their version meets your requirements. You may decide that

you need to use

your own form for your purposes, even if they automatically

have partici

pants sign their form when they arrive.

It is common to back-translate materials to make sure that

they are accu

rately translated. This can be seen as an added cost, but

is a good idea, espe

cially if you do not have access to someone who is fluent

in the target

language(s) who can give you an informal check to make sure

the transla

tion has captured the spirit and nuances of the original.

Often, the project timeline requires that work begin on

translation of ma

terials while the study methodology is still evolving. This

can result in multi

ple cycles to update the translations. This makes version

control extremely

important. It is amazing how rapidly different versions can

multiply and po

tentially cause confusion, especially when multiple people

in different coun

tries all need to be looking at the same version at the

same time. The
challenge is compounded when you are preparing a
multicountry study.

Simultaneous translation of sessions. The simultaneous

translator, or inter

preter, provides a voice-over that observers hear through

earphones and

which is also recorded as a track on the audioand videotape

of the sessions.

The interpreter may be in a soundproof booth in the

observation room, may

be sitting behind the team in the observation room, or may

be monitoring the

session remotely via a video hookup or even audio only. In

any case, this per

son must be highly fluent and able to maintain

concentration for long peri

ods of time, passing along every comment, and conveying

nuances in real

time to the team. Skill in translating documents is not the

same as skill in si

multaneous translation. Clearly, this is a job for a highly

skilled professional.

As a result, this critical role is also often quite costly.

The costs vary from mar

ket to market, with Japan being the highest in our

experience. Translation

costs can often be equal to or greater than the daily

facility rental costs. How

ever, do not be tempted to cut corners on the interpreter,

as problems with

interpretation can undermine the value of the entire

project.

Cost is also increased by the fact that, in some countries,

interpretation is

commonly done by a team of translators, who can take turns

during a

lengthy session, or who can alternate sessions during a

long day. Even when

this is not the case, it is critical to allow enough time

between sessions to give

the interpreter a break. Also, plan food and refreshments

for this person or

team during the breaks, so they can maintain their focus on

the translation.

Training Local Facilitators and Preparing Facilities

A key to success is to make sure that the local personnel

are fully trained in

the methods you plan to use. This is especially true if you

are using a local fa

cilitator who is experienced in market research methods,

but not in usabil

ity. Whereas part of this takes place before you arrive,

part of it can only be

done after you are on the scene.

Training of local facilitators in usability methods.

Assuming that you are not

working with a trained usability professional who is fluent

in the local lan

guage, you will have additional work to prepare the

facilitator. As we have

already mentioned, the role of and demands on the

facilitator depend on

the style of testing you plan to do. If the test is a

highly structured metric

test, then you may be able to focus primarily on training

the facilitator in the
mechanics of the evaluation process. However, if you plan
to conduct a

more exploratory usability evaluation, you will need an

appropriate process

to train the facilitator, as well as a strategy for

communicating with the facili

tator during the session.

When we are working with a facilitator for the first time,

we typically pro

vide training in the concepts of usability evaluation, both

before arrival, and

via hands-on training, role playing, and dry runs on site.

We usually spend

time on the phone with the facilitator prior to our

arrival, giving him or her

an overview of the concepts, with a focus on how usability

evaluation differs

from methodology they may be more familiar with. We also

send written

materials and pointers to Web sites, including materials we

have available

on our own Web site, to give them a background in what will

be expected.

Depending on the circumstances, we have used some other

strategies.

Sometimes it is possible to send a video of a related test

so they can see how

we have done it. This works best if it is a tape of an

earlier evaluation of the

same product. Occasionally, we conduct a mock session with

the facilitator

over the phone, using a public Web site and a set of

standard tasks. For in
stance, we may have them think aloud while trying to find
and order a book

at a popular bookselling Web site in their country, or have

them find some

information, such as the weather forecast, for their

hometown.

Once we arrive, we go over this information again and then

conduct a

dry run in person. During the dry run, we often will first
demonstrate the

process with the facilitator playing the role of the user,

then have them

conduct a session, with one of us taking the role of the

evaluator. If possi

ble, and appropriate to the technology being evaluated, we

end by having

the facilitator conduct a dry run using a facility staff

member as the user.

This process is valuable on several levels. Of course, it

helps the facilitator

learn the test plan thoroughly, while at the same time

revealing last-min

ute modifications that may be necessary. Second, it is an

opportunity for

the facilitator to learn the paradigm of usability testing

with its focus on

user behavior and cognitive process rather than opinion,

and some of the

basic methods of exploratory usability evaluation, such as

ways to elicit

thinking out loud and ways to probe nondirectively.

Finally, it is also a

chance for them to understand the underlying goals and

questions that

are implicit in the test plan and get an idea of the types
of issues that inter

est us in relation to this particular design. This is

critical, as it helps them

get a sense of where to take the initiative to follow up

and draw people out

to help the team answer key questions.

This level of preparation has proven its worth many times.

It provides fa

cilitators with a good sense of what to expect before

getting into the actual

usability evaluation sessions, and helps them function much

more effec

tively as collaborators. It also helps make the entire

process much more re

warding and professionally stimulating for them.

Furthermore, there may

be a longer range benefit from the effort you put in with

your facilitators. In

our case, people we have trained for this role have become
part of our net

work of collaborators, allowing us to do subsequent

international studies

much more easily. The same may apply for you.

The preparation prior to the research, however, is not

enough to turn

the facilitator into a usability expert. In addition you

need to plan for how

you will communicate with the facilitator during the

session when neces

sary, or even if you will attempt to supervise and direct

the process of testing
from behind the mirror. Some of the methods that people use
are passing

notes, breaking for a quick consultation during the

session, or even knock

ing on the door and calling the facilitator out. These

methods work best in

situations where there are clear predetermined break

points, or there is a

fairly clear threshold for intervention, such as when the

process has clearly

broken down and the facilitator needs some guidance about

how to get back

on track. However, they lack the flexibility and immediacy

necessary for

probing those ephemeral opportunities for judicious probing

that tend to

occur in usability evaluation and often lead to deep

insights—moments

which the facilitator may miss.

To address this, we have often used a technique called the

“bug in the

ear.” This is a wireless intercom that uses an earplug

speaker/microphone.

It allows the usability expert behind the mirror to

communicate immedi

ately to the facilitator. However, this takes practice. If

this is not used skill

fully and judiciously, in terms of timing, quantity, and

wording of

messages, it can be potentially confusing and disruptive to

the facilitator.

However, we have found that, given some time for

familiarization and
practice, it can be very effective. We routinely use this
technique when the

two of us work together on a usability evaluation, with one

serving as facili

tator and one behind the mirror and, therefore, have had
the opportunity

to practice ways to time our communications and to word

specific probes

to be most useful to the facilitator.

If a facilitator is not comfortable with the “bug in the

ear,” if the per

son’s level of comfort with your language is not high

enough, or if the tech

nique is not working effectively, we have resorted to other

compromise

solutions, such as the common ones of passing notes under

the door or

having the facilitator step out of the evaluation room for

clarification. De

pending on the risks of our being present with the user, we

have also sat

with the facilitator, to pass notes when necessary.

Sometimes we have done

this while wearing the bug ourselves, so we can relay

suggestions from the

observers (subject to our own “editorial” authority).

Finally, even if the

bug in the ear can not be used in a rich and flexible way
to cue the facilita

tor for specific probes, it can still be quite helpful when

used in a very lim

ited way to help manage the flow of the session, such as by

using it simply
to signal the facilitator to move on, or reminding the
facilitator to encour

age thinking out loud. It is helpful to agree in advance to

a small number

of simple code words you will use for these purposes.

Checking and correcting problems in equipment set up prior

to test. It is helpful if

you can be present when the equipment is being set up. For
instance, if you

have arranged for a picture-in-picture recording, it may

require some

tweaking to set up properly. Because you know what aspect

of the screen is

the least likely to have critical interactions, you may

need to specify where to

position the inset. It also helps to be present at setup so

you can understand

what to do should troubleshooting become necessary in the

middle of a ses

sion, since, typically, it is not cost effective to have

subcontracted video tech

nicians stay onsite for the duration of testing.

Train interpreters. Another key to success is briefing

interpreter(s) on the pro

cess of usability testing. This allows them to understand

more fully why you

need to hear everything, and why even offhand comments can

be the keys to

user confusion that you need to understand in order to make

changes in your

product. Because of the cognitive load on them,

interpreters may normally let
these chance comments go by without translating them, so it
is especially im

portant that they understand the value of these comments to

you.

We usually conduct our training of interpreters shortly

before the actual ses

sions. We give an overview of usability evaluation (or

ethnographic visits, if that

is what we are doing) and explain carefully the types of

things that are most

useful to hear. We try to help them understand that the

tone of voice, as well as

the words, is important to convey. This helps the

interpreters to balance their

load, by focusing on the relevant nuances of tone or

expression that are most

useful. We try to have the interpreters present for the dry

runs that we do with

the facilitators, and may even use the interpreter as an

evaluator in a dry run to

help them get a better sense for the setup. If the

interpreter is observing the

evaluations only through a video connection (or if they

cannot see it at all) this

helps them to visualize the set up, which can help with
their interpretation.

Multicountry Studies: Serial or Parallel?

We have found that it is very beneficial to have continuity

on the observation

team for multicountry studies. This means that at least one

key person

needs to observe all of the studies—visits or usability

evaluations—in all of
the countries. This may seem more costly, but in reality,
it is very helpful to

ensure that you can make real sense of the findings. Even
though it is valu

able to have “new eyes” along as well, it is critical that

someone has seen all

of the users so that any differences in procedure or result

can be put in con

text. In studies where this is not the case, it can be

extremely difficult to

know whether to attribute a difference in findings to a

difference between

locales or to an inconsistency between teams.

In addition, what you see depends on what is salient to you

and this in

turn is influenced by the patterns you have already seen.

Things that would

not have been noticed by themselves can become noteworthy

when they

stand out against other patterns. This is the familiar

figure/ground phe

nomenon. For example, in a recent study in two countries in

the same

larger geographical region, we noted that in one country

people tended to

make an initial attempt to guess at a URL, whereas in the

nearby country,

when faced with the same web navigation task, almost all
the users followed

a strategy of surfing links starting from their home page.

In both cases, this

observation was incidental to the main point of the study.

Therefore, if we
had not seen these two contrasting patterns side by side,
the chance that we

really would have taken note of them would have been less.

It goes without saying that if you are to have continuity,

it is not recom

mended that you conduct your study in different countries

in parallel. Of

course, it takes more time to do them serially. If time is

of the essence, you may

want to examine other ways to expedite the data gathering,

before resorting to

parallel teams, such as reducing the number of visits or

evaluations, or select

ing the locations to allow more rapid travel between them.

If you can not pro

vide the continuity yourself, consider having a consultant

play this role.

However, if it is simply not possible or feasible to

conduct the study in se

ries and to maintain some continuity across teams, then you

will need to de

vote extra effort to planning for some standardization of

methodology, and

ongoing coordination among teams. We recommend that you

plan to main

tain communication with and among the teams during the data
gathering,

so that modifications in the procedure can be coordinated

as appropriate,

and insights and possible trends can be shared. This can

help reduce the

risk that if two teams come back describing different

themes and trends, it is
not merely because their thinking happened to get into
different channels

and they were not aware of what the other teams were doing.

CONCLUSIONS

In this chapter, we tried to give you a flavor for some of

the major decisions

and trade-offs that you are likely to encounter when doing

international us

ability evaluations or ethnographic visits. Certainly there

are many other is

sues one could discuss, but it is not possible to cover

them all in one chapter,

and these are some of the biggest ones. It would have been
easier for us had

we known these things in advance, rather than having to

learn them the

hard way. The community of professionals doing

international user re

search is growing, and we therefore have increasing

opportunities to learn

from each other. With this in mind, we welcome a dialog

with you as you

pursue your quest for international usability.

Dray, S. M., & Mrazek, D. (1996). A day in the life:

Studying context across cultures.

In E. M. del Galdo & J. Nielsen (Eds.), International user

interfaces (pp. 242–256).

New York: Wiley.

Siegel, D. A., & Dray, S. M. (2001). New kid on the block:

Marketing organizations

and interaction design. Interactions, 8(2), 19–24. Chapter

Cost-Justifying Usability
Engineering for Cross-Cultural

User Interface Design

Deborah J. Mayhew

Deborah J. Mayhew & Associates, West Tisbury, MA

Randolph G. Bias

University of Texas at Austin, Austin, TX

INTRODUCTION

In this chapter, we discuss the cost justification of

adding usability engineer

ing to projects aimed at developing software to be used in

multiple different

countries and cultures. Although both of us have done

significant work in

the area of cost-justifying usability engineering in

general, neither of us has

significant concrete experience with designing for

cross-cultural audiences.

We have drawn on the writings of and conversations with our

colleagues

who do have significant experience in cross-cultural

design, and extrapo

lated from our collective experience about cost justifying

the application of

usability engineering in cross-cultural contexts.

Cost justification of usability engineering in general is

addressed in de

tail elsewhere (Bias & Mayhew, 1994; Bias, Mayhew &

Upmanyu, 2003;

Mayhew & Bias, 2003). Here, we focus on what is unique

about adding us
ability engineering to a global project, and how to adapt
the general cost

justification technique in this case. To do this, we first

provide an overview

of the general usability engineering process. Then, we

provide some discus 213

sion and examples of unique aspects of the usability

engineering process for

cross-cultural projects. Based on this background, we then

provide an over

view of the general cost justification technique and some

examples of

adapting the technique to cross-cultural projects.

Overview of a Generic Usability Engineering Lifecycle

The Usability Engineering Lifecycle (Mayhew, 1999) is a

structured and sys

tematic approach to addressing usability within the product

development

process. It consists of a set of usability engineering

tasks applied in a partic

ular order at specified points in an overall product

development lifecycle.

Several types of tasks are included in The Usability

Engineering Life

cycle, as follows:

• Structured usability requirements analysis tasks.

• An explicit usability goal setting task, driven directly

from requirements analysis data.

• Tasks supporting a structured, top-down approach to user

interface design that is driven directly from usability
goals and other requirements data.

• Objective usability evaluation tasks for iterating design

toward usability goals.
Figure 8.1 represents in summary, visual form, The
Usability Engi

neering Lifecycle. The overall lifecycle is cast in three

phases: Require

ments Analysis, Design/Testing/Development, and

Installation. Specific

usability engineering tasks within each phase are presented

in boxes,

and arrows show the basic order in which tasks should be

carried out.

Much of the sequencing of tasks is iterative, and the

specific places where

iterations would most typically occur are illustrated by

arrows returning

to earlier steps in the lifecycle. Following the chart are

brief descriptions

of each lifecycle task.

Phase One: Requirements Analysis

In this phase, user research is conducted and collated. A

description

(User Profile) of the specific user characteristics

relevant to user interface

design (e.g., computer literacy, expected frequency of use,

level of job expe

rience) is obtained for the intended user population. This

will drive tailored

user interface design decisions and also identify major

user categories for

study in the Task Analysis task. A study of users’ current

tasks, workflow pat

terns, and conceptual frameworks is made (Task Analysis),

resulting in a de
scription of current tasks and workflow and an
understanding and

specification of underlying user goals. These will be used

to set usability

goals and drive Work Reengineering (also known as

Information Architec

ture) and user interface design.

The user interface capabilities and constraints (e.g.,

windowing, di

rect manipulation, color, etc.) inherent in the technology

platform cho

sen for the product (e.g., Apple Macintosh, MS Windows,

product

unique platforms) are determined and documented (Platform

Capabil

ities/Constraints.) These will define the scope of

possibilities for user in

terface design. Relevant general user interface design

guidelines

available in the usability engineering literature are

gathered and re

viewed (General Design Principles.) They will be applied

during the de

sign process to come, along with all other project-specific

information

gathered in other tasks in this phase.

Based on all usability requirements as described above,

both qualitative

goals reflecting general usability requirements and

quantitative goals defin

ing minimal acceptable user performance and satisfaction

are developed

(Usability Goals.) These usability goals focus later design

efforts and form

the basis for later iterative usability evaluation.

FIG. 8.1. The usability engineering lifecycle (Mayhew,

1999, used with per

mission).

Phase Two: Design/Testing/Development

Level 1 Design. Based on all requirements analysis data and

the usabil

ity goals extracted from them, user tasks are redesigned at

the level of orga

nization and workflow to streamline work and exploit the

capabilities of

automation (Work Reengineering). No user interface design

per se is in

volved in this task, just abstract organization of

functionality and workflow

design. This task is sometimes referred to as Information

Architecture.

Next, initial high-level design alternatives are generated.

At this level, navi

gational pathways and major displays are identified, and

rules for the con

sistent presentation of navigational controls, work

products, processes and

actions are established (Conceptual Model Design). Screen

design detail is

not addressed at this design level.

Mockups (paper-and-pencil or live prototype) of Conceptual

Model De

sign ideas are prepared (Conceptual Model Mockups.)

Detailed screen de

sign and complete functional design are not addressed at

this stage. The

mockups are evaluated and modified through iterative

evaluation tech

niques such as formal usability testing, in which real,

representative end us

ers attempt to perform real, representative tasks with

minimal training and

intervention, imagining that the mockups are a real product

user interface

(Iterative Conceptual Model Evaluation.) The design,

mockup, and evalua

tion tasks are conducted in iterative cycles until all

major usability “bugs”

are identified and engineered out of the Level 1 (i.e.,

Conceptual Model)

design. Once a Conceptual Model is relatively stable,

system architecture

design can commence.

Level 2 Design. A set of product-specific standards and

conventions

for all aspects of detailed screen/page design is developed

(Screen Design

Standards), based on any industry and/or corporate

standards that have

been mandated (e.g., Microsoft Windows, Apple Macintosh,

etc.), the data

generated in the Requirements Analysis phase, and the

product-unique

Conceptual Model Design arrived at during Level 1 Design.

Screen Design

Standards will insure coherence and consistency—the

foundations of us

ability—across the user interface.

The Screen Design Standards (as well as the Conceptual
Model Design)

are applied to design the detailed user interface to

selected subsets of prod

uct functionality. This design is implemented as a live

prototype (Screen

Design Standards Prototyping). An evaluation technique such

as formal us

ability testing is carried out on the Screen Design

Standards prototype (Iter

ative Screen Design Standards Evaluation), and then

redesign/reevaluate

iterations are performed to refine and validate a robust

set of Screen Design

Standards. Iterations are continued until all major

usability “bugs” are

eliminated and usability goals seem within reach.

At the end of the design/evaluate iterations in Design

Levels 1 and 2,

you have a validated and stabilized Conceptual Model

Design, and a vali

dated and stabilized set of standards and conventions for

all aspects of de

tailed Screen Design. These are captured in the document

called the

product Style Guide, which already documents the results of

requirements

analysis tasks (Style Guide). During Detailed User

Interface Design, fol

lowing the Conceptual Model Design and Screen Design

Standards in the

product Style Guide will insure quality, coherence, and

consistency, the
foundations of usability.

Level 3 Design. Detailed design of the user interface to

the complete

set of product functionality is carried out based on the

refined and vali

dated Conceptual Model Design and Screen Design Standards

docu

mented in the product Style Guide (Detailed User Interface

Design).

This design then drives product design and development. A

technique

such as formal usability testing is continued during

product develop

ment to expand evaluation to previously un-assessed subsets

of function

ality and categories of users, and also to continue to

refine the user

interface and validate it against usability goals

(Iterative Detailed User

Interface Design Evaluation).

Phase Three: Installation

After the product has been installed and in production for

some time,

feedback is gathered to feed into enhancement design,

design of new re

leases, and/or design of new but related products (User

Feedback).

Overview of Unique Aspects of Usability Engineering

for Cross-Cultural User Interfaces

Various aspects of The Usability Engineering Lifecycle, as

described ear
lier, will need to be modified or added to, and may be more
difficult and ex

pensive, when designing for a cross-cultural audience.

There will always be

a variety of approaches possible to addressing the unique

aspects of includ

ing usability engineering in a cross-cultural project. Dray

and Siegel (chap.

7, this volume) address many of these alternative

approaches. Here, we just

refer to them briefly to provide context for this chapter

on cost justification.

In the Requirements Analysis phase, fundamental and

application-in

dependent cross-cultural differences in design, such as use

of color, sym

bols and icons, date and currency formats (referred to

extensively in

other chapters in this volume) will need to be addressed.

The good news

here is that most likely all of this information about a

culture (i.e., lan

guage issues, formatting conventions, use of color, etc.)

can probably be

accurately learned from a single native individual who also

speaks very

fluent American English (or whatever is the language of the

designing

organization). It may be possible, and would certainly be

desirable, to

hire such a person to be on site with the project team full

time. Of course,

if there are multiple target cultures, multiple natives

must be hired. An
other alternative here is to find a firm that specializes
in researching and

providing this type of cross-cultural information and

partnering with

them. Still another approach is to simply assume that when

in a target

country doing the application-unique types of requirements

analyses

that are a normal part of The Usability Engineering

Lifecycle (i.e., User

Profile and Contextual Task Analyses), your usability

engineer will be at

tuned to these more general cultural issues, and as a part

of other activi

ties, pick up these differences. Depending on which

approach is chosen,

different costs will be incurred. Even though gathering

this type of re

quirements data is necessary and must be planned for,

strictly speaking,

it need not be included in the cost justification of a

usability engineering

program, because as explained earlier, it is really a basic

prerequisite for

designing for another culture.

While a User Profile questionnaire can be distributed via a

web site or

e-mail, it may need to be translated into the target

language, and recruiting

appropriate participants to respond may be more complex

than in one’s

native country. It might be most effective to hire a firm

in each target coun
try to recruit questionnaire participants, and communicate
the sampling

requirements to them. The services of a translator may be

necessary to com

municate the recruiting task to the local firm.

An alternative to an end user questionnaire for user

profiling is to con

duct interviews with representatives of target users (e.g.,

managers) in each

country and ask them to characterize the target user

population, although

just as when using this technique in one’s native country,

this sort of second

hand data is always less reliable. If you do this, you may

still need to plan on

translation costs to conduct these interviews.

Task Analyses will be considerably more costly and

difficult in other

countries. One really needs access to actual end users to

conduct an effective

Task Analysis. One approach is for someone in the designing

organization

to travel to each of the target countries and, if

necessary, employ a transla

tor to both translate all task analysis materials

beforehand and translate in

real time during the Task Analysis sessions with potential

users. Time

should also be planned to coordinate with and train the

translators regard

ing what is important to pick up and what is not. You would

need to factor in
travel time as well, as it may not only take much longer
for your usability en

gineer to get to the target country than to travel within

the United States (or

whatever is the home country), but there may be a day or

two required to re

cover from jet lag and the time zone change as well.

Alternatively, locals skilled in Task Analysis techniques

might be hired in

the target countries and communicated with effectively,

which may entail

translation costs and long distance phone expenses.

In addition, it may take more time to conduct Task Analyses

in coun

tries other than the United States. For example, in some

countries you

might not be able to get people to participate during work

hours. In oth

ers, you might not get them on weekends. If someone is

traveling to the

target country to conduct task analysis interviews, you

should thus prob

ably plan more elapsed travel time to get the interviews

completed than

it would take in the United States. Also, in some cultures

no-shows may

be much more likely than in the United States, especially

if they must

come to some other facility away from work, so you may have
to schedule

as many as two to three users for every data point you hope
to get, and

correspondingly plan the additional time to run sessions

with whoever

does show up.

In the Design/Testing/Development phase, during design

activities it

would be very helpful to have a native member of each

target country on the

team. Planning needs to be done in this phase to build a

global architecture

that can be most efficiently translated into each local

target country. In later

stages of design, design will need to be localized for each

target country.

Usability tests should be conducted with native end users.

One approach

is to hire locals in each target country skilled in

usability testing and com

municate effectively with them, which may entail

translation costs and long

distance phone expenses. Alternatively, you might send

someone in the de

sign organization to the target countries and hire

translators to work with

them. Just as with performing Task Analyses, it may take

more travel time,

as well as more time in-country to collect enough data

points in a test, as the

rate of no-shows may be higher than it would be in one’s

own country, and it

may be harder to schedule test users.

In either case, you need to recruit appropriate users for

testing, and costs

similar to those described earlier in this chapter for

recruiting for User Pro
files and Task Analyses need to be factored in.

An alternative is hiring skilled locals to perform

heuristic evaluations of

prototype designs. In this case, translation costs may be

incurred to trans

late the report into the native language of the development

organization.

Many of the same issues arise in the Installation phase

where usability

feedback is desired after implementation. Users are remote,

and there is

additional time and expense involved in soliciting their

direct feedback.

Again, alternatives are hiring skilled locals, or having

someone from the de

sign organization travel and work with a translator. Some

feedback tech

niques can be conducted remotely, using the Web or e-mail,

and again, for

these techniques it might be best to hire a local firm to

recruit appropriate

participants based on sampling specifications. Again, there

may be addi

tional costs for various translation services, and

increased travel time and

in-country time relative to within one’s own country.

FRAMEWORK FOR COST JUSTIFICATION

This rest of this chapter briefly presents a general

approach for cost-justify

ing usability engineering efforts for software development

projects. The

subject of cost-justifying usability engineering in general

is covered exten

sively in a book called Cost-Justifying Usability (Bias &

Mayhew, 1994.) In par

ticular, chapter 2 in that book elaborates, in much more

detail, on the

simple framework offered in this chapter. Other chapters in

that book offer

further discussion on the topic and case studies. Readers

interested in

learning more about how, when, and why to cost-justify

usability engineer

ing efforts, are referred to that book for more detail than
can be provided

here. (Also see Bias et al., 2003; Mayhew & Bias, 2003) The
purpose of this

chapter is to provide an overview of how that more general

cost justification

technique can be applied to cross-cultural development

projects in particu

lar. To that end, a simplified overview is provided,

followed by some sample

hypothetical analyses for cross-cultural development

projects.

Purposes for Cost Justification

There are two possible purposes for conducting a

cost–benefit analysis of

usability engineering efforts for software development

projects: (a) to “sell”

usability engineering in general, and (b) to plan for a

usability engineering

program for a particular development project

Very general cost–benefit analyses of hypothetical

usability engineering
efforts can be prepared as a strategy to win general
support for trying out

usability engineering techniques in a development

organization. When an

organization has no experience with usability engineering,

cost–benefit

analyses can help to win resources to experiment with it

for the first time.

In organizations more mature with respect to usability

engineering,

cost–benefit analyses can be used to plan an optimal

usability engineering

effort for a particular development project.

To help settle on a final usability engineering plan for a

specific develop

ment project, you could start out by calculating the costs

of the most aggres

sive usability engineering program that you might like to

implement,

including rigorous techniques for all lifecycle tasks. Then

you could estimate

benefits, using very conservative estimates of benefits. If

benefits outweigh

costs dramatically, as they usually do when critical

parameters are favor

able, then you could easily make a good argument for even
the most aggres

sive usability engineering program, because only the most

conservative

claims concerning potential benefits have been made and, as

such, can be

defended easily. In fact, you can then go back and redo the
benefits esti
mates using more aggressive, yet still realistic, benefit
assumptions, and

show that in all likelihood an even more dramatic overall

benefit will be re

alized, even from a significant investment in usability

engineering.

If, however, benefits and costs in the initial analysis

seem to match up

fairly closely, then you will want to scale back your

initial, aggressive usabil

ity engineering program, maybe even to just a bare-bones

plan. Perhaps

you should plan to do only a quick and dirty user profile

by interviewing

user management, a very quick and dirty task analysis

consisting of just a

few rounds of contextual observations/interviews with

users, and then do

just one iterative cycle of usability testing on a complete

detailed design, to

catch major flaws and be sure you have achieved the

predicted benefits. You

are still likely to realize the very conservative

assumptions made concerning

benefits with just these minimal usability techniques, and

thus you could

predict with confidence a healthy return on investment

(ROI) from a more

minimal approach to usability engineering. This is probably

wiser in the

long run than making overly optimistic claims concerning

potential bene

fits, spending a lot to achieve them, and then perhaps not

achieving them

and losing credibility.

Thus, you can use a cost–benefit analysis in this fashion,

to develop a sen

sible usability engineering effort for a cross-cultural

development project,

likely to pay off as predicted.

When an organization is first experimenting with usability

engineering

techniques, and is still skeptical about their value, it is

wise to make very

conservative cost–benefit arguments, based on a relatively

low cost usability

engineering effort and very modest estimates of benefits,

and then to show

after the fact that much larger benefits were in fact

realized. Once an orga

nization has several positive experiences investing in

usability engineering,

it will be more receptive to more aggressive proposals for

usability engi

neering programs, and to more optimistic benefits estimates

in the cost–

benefit analyses that argue for them. (See Bias & Mayhew,
1994, chap. 2, for

additional concrete examples of how to use cost–benefit

analyses to tailor

usability engineering programs.)

General Approach

To cost justify any usability engineering plan,

cross-cultural or not, you sim

ply adapt a very generic and widely used cost–benefit

analysis technique.

Having laid out a detailed usability project plan based on

lifecycle tasks (see

earlier section titled “Overview of a Generic Usability

Engineering Life

cycle,” and also Mayhew, 1999), it is a fairly

straightforward matter to calcu

late the costs of that plan. Then you need to calculate the
benefits. This is a

little trickier and where the adaptation of the generic

analysis comes into

play. Then, you simply compare costs to benefits to find

out if, and to what

extent, the benefits outweigh the costs. If they do to a

satisfactory extent,

then you have cost justified the planned effort.

More specifically, first a usability engineering plan is

laid out. The plan

specifies particular techniques to employ for each

lifecycle task, breaks the

techniques down into steps, and specifies the personnel

hours and equip

ment and travel costs for each step. The cost of each task
is then calculated

by multiplying the total number of hours for each type of

personnel by their

effective hourly wage (fully loaded, as we explain later)

and adding in any

equipment, travel, and other costs. Then, the costs from

all tasks are

summed to arrive at a total cost for the plan.

Next, overall benefits are predicted by selecting relevant

benefit categories,
calculating expected benefits by plugging project specific
parameters and as

sumptions in to benefit formulas, and summing benefits

across categories.

The list of possible benefits to consider is long, as

usability engineering

leads to tangible benefits to all concerned. The

development team realizes

savings, as problems are identified early when they are

cheap to fix. The

customer support team realizes a reduced call support

burden. More usable

e-commerce Web sites have higher buy-to-look ratios, a

lower rate of aban

doned shopping carts, and increased return visits. More

usable applica

tions have fewer errors and higher user productivity.

Training costs are

lower on more usable applications.

At this point, to simplify the discussion of cost

justification for cross-cul

tural software, we limit the discussion to web sites. It

should be noted, how

ever, that similar analyses can be applied to

shrink-wrapped software, and

software developed for internal use by specific companies.

The benefit cate

gories might be slightly different in these cases, but the

overall analyses are

highly analogous. Other sources provide the details of

analyses in these other

cases (Bias & Mayhew, 1994; Bias et al., 2003), and could
be adapted with the
details of cross-cultural usability engineering provided in
this chapter.

The potential benefit categories selected in a particular

cost–benefit

analysis depends on the basic business model for the web

site. Sample bene

fit categories potentially relevant to e-commerce sites

include: (a) increased

buy-to-look ratios, (b) decreased abandoned shopping carts,

(c) increased

return visits, (d) decreased costs of other sales channels,

(e) decreased usage

of “Call Back” button, and (f) decreased development costs.

Sample benefit categories potentially relevant to sites

based on an ad

vertising business model include: (a) increased number of

visitors, (b) in

creased number of return visitors, (c) increased length of

visits, (d)

increased click-through on ads, (e) decreased failed

searches, and (f) de

creased development costs.

Sample benefit categories potentially relevant to product

information sites

include: (a) increased sales leads, and (b) decreased

development costs.

Sample benefit categories potentially relevant to customer

service sites in

clude: (a) decreased costs of traditional customer service

channels, and (b)

decreased development costs.

Sample benefit categories potentially relevant to intranet

sites support

ing internal business users include: (a) decreased user

training costs, (b) in

creased user productivity, (c) decreased user errors, (d)

decreased user

support, and (e) decreased development costs.

Note that the relevant benefit categories for different

types of sites

vary somewhat. Thus, in a cost–benefit analysis, one wants

to focus atten

tion on the potential benefits that are of most relevance

to the bottom line

business goals for the site.

Also note that these benefits represent just a sample of

those that might

be relevant in these types of sites, and does not address

other possible ben

efits of usability in other types of sites. Others might be

included as appro

priate, given the business goals of the site sponsor and

the primary

concerns of the audience, and could be calculated in a

similar fashion as

these are (see next section).

Finally, overall benefits are compared to overall costs to

see if, and to

what extent, the overall usability engineering plan is

justified.

SAMPLE COST–BENEFIT ANALYSES FOR CROSS-CULTURAL

USABILITY ENGINEERING PROJECTS

As far back as 1991, five of the six major U.S. computer

vendors were bring
ing in over 50% of their income from international sales
(Russo & Boor,

1993). Smaller countries had an even larger percentage of

their sales come

from outside their own country (Nielsen, 1990a). In the

decade plus since

then, the Internet and the World Wide Web have made it even
easier to

market and sell in other countries. However, understanding

cross-cultural

differences, and incorporating them into both products and

marketing

channels like the Web, has a profound impact on the success

of interna

tional marketing efforts, just as incorporating an

understanding of domes

tic users and their tasks into products and marketing

channels impacts

overall success and ROI in one’s own country.

Consider a hypothetical usability engineering plan in the

context of two

different cross-cultural web development projects, and then

see how you

could conduct cost–benefit analyses of that plan for each

project.

A Cross-Cultural E-Commerce Web Site

Imagine a web development organization is planning to

redesign an exist

ing e-commerce site, which is not producing the ROI hoped

for. Traffic

and sales statistics are available from the current site.

In particular, as
sume the original site was designed in English for a U.S.
audience, but is

being used by multiple other cultures, and is not

performing as well as

hoped in any country, but especially in the international

countries. Part of

the goal of the redesign is to tailor the site for use in a

small number of tar

get countries. The usability engineer prepares a proposed

usability engi

neering project plan to be conducted for each target

country (including

the country of the development organization), and then

performs a

cost–benefit analysis of that plan.

First, the final results of cost–benefit analyses for a

“base” country (in this

example, the United States) and one international country

are presented in

the rest of this section. Then, in the sections that

follow, the derivations of

those final results are described. The same type of

analysis would apply to

each international country, although the numbers would vary

from country

to country depending on costs in individual countries. It

should be noted

that the cost for usability engineering programs would

probably be less for a

set of countries than the sum of the costs computed

separately for each

country, as certain parts of the program could be

generalized or “amor
tized” across countries. For example, creating usability
testing materials

would likely take the most time in the first country, but
be a simple matter of

translation for other countries. However, for simplicity’s

sake, here we as

sume and present the analysis for just one international

country.

Table 8.1 shows the overall calculation of the cost of a

usability engineer

ing plan for a project including a base country (in this

example, the United

States) and one international country (imagine a country

in, say, South

America, where the cost of living and pay scales are

significantly lower than

in the United States). In Table 8.1, the first column

identifies the overall

project phases. The second column identifies which

Usability Engineering

Lifecycle tasks (see “Overview” section at chapter

beginning) are planned in

each phase. This column also breaks down each task

according to general

categories of costs, first for the base country (in this

example, the United

States), and then for the international country.

The third, fourth, fifth, sixth, and seventh columns

identify the number

of hours of time of different types of professionals that

are required to com

plete each task. The last column then summarizes the total
cost of each cost
category within a task. Note that values in the cells
representing hours and

costs for the international country are expressed as

differences relative to the

base country, referred to in Table 8.1 as “International

Delta.” For exam

ple, under User Profile, you will see that relative to the
base country, 12

more hours of the usability engineer’s time (beyond the 80

hours required

for doing a User Profile at home) are expected to be

required.

Grand cost totals for the whole plan are given at the
bottom of the table.

All dollar amounts are given in the currency of the base

country, in this ex

ample, U.S. Dollars. 225 T A B L E 8 . 1 C o s t o f U s a

b i l i t y E n g i n e e r i n g P l a n P H A S E T A S K
U s a b i l i t y E n g i n e e r H o u r s @ $ 1 5 0 D e v
e l o p e r H o u r s @ $ 1 5 0 U s e r H o u r s @ $ 4 0 U
s e r H o u r s @ $ 2 0 L a n g u a g e / C u l t u r e C o
n s u l t a n t H o u r s @ $ 7 5 T O T A L C O S T R e q u
i r e m e n t s A n a l y s i s U s e r P r o f i l e B A S
E H o u r s 8 0 1 2 . 5 $ 1 2 , 5 0 0 R e c r u i t m e n t
F e e s $ 2 , 0 0 0 I N T E R N A T I O N A L D E L T A H o
u r s 1 2 – 1 2 . 5 1 2 . 5 1 2 + $ 2 , 4 5 0 R e c r u i t
m e n t F e e s – $ 1 , 0 0 0 M a t e r i a l s T r a n s l
a t i o n + $ 1 , 0 5 0 T a s k A n a l y s i s B A S E H o
u r s 1 0 4 3 0 $ 1 6 , 8 0 0 R e c r u i t m e n t F e e s
$ 8 0 0 T r a v e l / P h o n e E x p e n s e s $ 3 , 2 7 5
I N T E R N A T I O N A L D E L T A H o u r s 8 0 – 3 0 3 0
3 2 + $ 1 3 , 8 0 0 ( C o n t i n u e d o n n e x t p a g e
)

226 T A B L E 8 . 1 ( c o n t i n u e d ) P H A S E T A S K
U s a b i l i t y E n g i n e e r H o u r s @ $ 1 5 0 D e v
e l o p e r H o u r s @ $ 1 5 0 U s e r H o u r s @ $ 4 0 U
s e r H o u r s @ $ 2 0 L a n g u a g e / C u l t u r e C o
n s u l t a n t H o u r s @ $ 7 5 T O T A L C O S T R e c r
u i t m e n t F e e s – $ 4 0 0 M a t e r i a l s T r a n s
l a t i o n + $ 3 0 0 L o c a l F a c i l i t a t o r + $ 2
, 5 0 0 S i m u l t a n e o u s T r a n s l a t i o n + $ 2
, 0 0 0 V i s a + $ 1 8 0 T r a v e l / P h o n e E x p e n
s e s + $ 1 , 5 7 5 P l a t f o r m C o n s t r a i n t s B
A S E H o u r s 8 8 $ 2 , 4 0 0 U s a b i l i t y G o a l S
e t t i n g B A S E H o u r s 2 0 $ 3 , 0 0 0 I N T E R N A
T I O N A L D E L T A H o u r s + 8 + $ 6 0 0 227 D e s i g
n / T e s t / D e v e l o p I n f o r m a t i o n A r c h i
t e c t u r e B A S E H o u r s 6 0 $ 9 , 0 0 0 I N T E R N
A T I O N A L D E L T A H o u r s 1 6 1 6 + $ 3 , 6 0 0 C o
n c e p t u a l M o d e l D e s i g n B A S E H o u r s 6 0
$ 9 , 0 0 0 I N T E R N A T I O N A L D E L T A H o u r s 1
6 4 0 + $ 5 , 4 0 0 S c r e e n D e s i g n S t a n d a r d
s B A S E H o u r s 6 0 $ 9 , 0 0 0 I N T E R N A T I O N A
L D E L T A H o u r s 1 6 4 0 + $ 5 , 4 0 0 L i v e P r o t
o t y p e D e v e l o p m e n t B A S E H o u r s 2 0 1 2 0
$ 2 1 , 0 0 0 I N T E R N A T I O N A L D E L T A H o u r s
1 6 4 0 + $ 5 , 4 0 0 ( C o n t i n u e d o n n e x t p a g
e )

228 T A B L E 8 . 1 ( c o n t i n u e d ) P H A S E T A S K
U s a b i l i t y E n g i n e e r H o u r s @ $ 1 5 0 D e v
e l o p e r H o u r s @ $ 1 5 0 U s e r H o u r s @ $ 4 0 U
s e r H o u r s @ $ 2 0 L a n g u a g e / C u l t u r e C o
n s u l t a n t H o u r s @ $ 7 5 T O T A L C O S T U s a b
i l i t y T e s t B A S E H o u r s 1 0 4 1 6 $ 1 6 , 2 4 0
R e c r u i t m e n t F e e s $ 4 0 0 T r a v e l / P h o n
e E x p e n s e s $ 2 , 7 7 5 I N T E R N A T I O N A L D E
L T A H o u r s 5 6 – 1 6 + 1 6 + 1 6 $ 9 , 2 8 0 R e c r u
i t m e n t F e e s – $ 2 0 0 M a t e r i a l s T r a n s l
a t i o n + $ 3 , 0 0 0 L o c a l F a c i l i t a t o r + $
2 , 5 0 0 S i m u l t a n e o u s T r a n s l a t i o n $ 2
, 0 0 0 V i s a + $ 1 8 0 T r a v e l / P h o n e E x p e n
s e s + $ 1 , 3 5 0 R e d e s i g n B A S E H o u r s 4 0 $
6 , 0 0 0 I N T E R N A T I O N A L D E L T A 1 6 $ 3 , 6 0
0 H o u r s 1 6 229 D e t a i l e d U s e r I n t e r f a c
e D e s i g n B A S E H o u r s 6 0 $ 9 , 0 0 0 I N T E R N
A T I O N A L D E L T A H o u r s 1 6 1 6 + $ 3 , 6 0 0 L i
v e P r o t o t y p e D e v e l o p m e n t B A S E H o u r
s 2 0 8 0 $ 1 5 , 0 0 0 I N T E R N A T I O N A L D E L T A
H o u r s 1 6 1 6 + $ 3 , 6 0 0 U s a b i l i t y T e s t B
A S E H o u r s 8 4 1 6 $ 1 3 , 2 4 0 R e c r u i t m e n t
F e e s $ 4 0 0 T r a v e l / P h o n e E x p e n s e s $ 2
, 7 7 5 I N T E R N A T I O N A L D E L T A H o u r s 4 0 –
1 6 1 6 1 6 $ 8 , 0 8 0 R e c r u i t m e n t F e e s – $ 2
0 0 M a t e r i a l s T r a n s l a t i o n + $ 3 , 0 0 0 L
o c a l F a c i l i t a t o r + $ 2 , 5 0 0 S i m u l t a n
e o u s T r a n s l a t i o n + $ 2 , 0 0 0 ( C o n t i n u
e d o n n e x t p a g e )

230 T A B L E 8 . 1 ( c o n t i n u e d ) P H A S E T A S K
U s a b i l i t y E n g i n e e r H o u r s @ $ 1 5 0 D e v
e l o p e r H o u r s @ $ 1 5 0 U s e r H o u r s @ $ 4 0 U
s e r H o u r s @ $ 2 0 L a n g u a g e / C u l t u r e C o
n s u l t a n t H o u r s @ $ 7 5 T O T A L C O S T V i s a
+ $ 1 8 0 T r a v e l / P h o n e E x p e n s e s + $ 1 , 3
5 0 R e d e s i g n B A S E H o u r s 4 0 $ 6 , 0 0 0 I N T
E R N A T I O N A L D E L T A H o u r s 1 6 1 6 + $ 3 , 6 0
0 T O T A L B A S E 7 6 0 2 0 8 7 4 . 5 0 0 $ 1 6 0 , 6 0 5
I N T E R N A T I O N A L D E L T A 3 1 6 0 – 7 4 . 5 + 7 4
. 5 + 2 8 4 + $ 9 2 , 2 7 5 T O T A L I N T E R N A T I O N
A L 1 , 0 7 6 2 0 8 0 7 4 . 5 2 8 4 $ 2 5 2 , 8 8 0

Next, the project usability engineer estimates that the

usability engineer

ing plans described in Fig. 8.1 will produce new site

designs with particular

expected benefits every month in the two countries, as

shown in Table 8.2.

Comparing these benefits and costs, the project usability

engineer ar

gues that the proposed usability engineering plan at home

will likely pay for

itself in the first five months after launch: Base Country

Benefits per month = $33,854 One-time cost = $160,605
PAY-OFF PERIOD = 4.74 months

Given the higher costs for the international usability

engineering pro

gram, plus the different parameters and predictions of

benefits, the pro

ject usability engineer then argues that the proposed

international

usability engineering plan will likely pay for itself

within the first seven

months after launch: International Country Benefits per

month = $40,625 One-time cost = $252,880 PAY-OFF PERIOD =
6.23 months

As the new sites are expected to have a lifetime of

something much lon

ger than 7 months, the project usability engineer expects

the plans to be ap

proved for these two countries in particular, based on this

cost justification. TABLE 8.2 Expected Monthly Benefits for
an e-Business Site

BENEFIT CATEGORY VALUE in Base Country— PER MONTH VALUE in

International Country— PER MONTH

Decreased abandoned shopping carts $15,625 $18,750

Increased buy-to-look ratio $15,625 $18,750

Decreased use of “Call Back” button $2,604 $3,125

TOTAL MONTHLY BENEFITS $33,854 $40,625

Note that the analyses offered here do not consider the

time value of

money. That is, the money for the costs is spent at some
point in time,

whereas the benefits come later in time. Also, if the money

was NOT spent

on the costs of usability engineering, but instead was

invested, this money

would likely increase in value. This investment could be

compared to the

benefits of investing in the usability engineering program.

Furthermore,

both costs and benefits are somewhat simplified and not

exhaustive. Some

expenses, but also some benefits, may have been overlooked.

Usually, the

benefits of usability are so robust that these more

sophisticated and com

plex financial considerations and details are not

necessary. However, if

needed, calculations based on the time value of money are

presented in
Bias et al. (2003), and also in Bias and Mayhew (1994).

Next is a step-by-step description of how the project

usability engineer

arrived at the final results described in this section.

Start With a Usability Engineering Plan

If it has not already been done, this is the first step in

conducting a

cost–benefit analysis. The usability engineering plan

identifies which Us

ability Engineering Lifecycle tasks and techniques will be

employed, and

breaks them down into required staff, hours, and other

expenses. Costs can

then be computed for these tasks in the next two steps. The
sample usability

engineering plan used in these examples and laid out in

Table 8.1 includes

most lifecycle tasks and fairly rigorous techniques for

each task. It assumes

that three tasks (i.e., the task analysis and two rounds of
usability testing) re

quire travel, but that the rest of the work can be done in
the base country lo

cation of the development organization.

Establish Analysis Parameters

Most of the calculations for both planned costs and

estimated benefits are

based on project-specific parameters. These should be

established and docu

mented before proceeding with the analysis. Sample

parameters are given

below for our hypothetical project.

Base (U.S.) Parameters:

• The site is an e-business site (vs. a product information

site, or a site based on an advertising model).

• The current site gets an average of 125,000 visitors per

month in the base country (vs., for example, 500,000 users
or 10,000 users).

• The current buy-to-look ratio in the base country is 2%.

• The current average profit margin on each online purchase

is $25.

• The current rate of usage of the “Call Back” button

(serviced by a local customer service organization) is 2%.

• Average time to service each usage of the “Call Back”

button is 5 minutes.

• Users are paid to participate in usability engineer tasks

at an hourly rate of $40.

• Customer support fully loaded hourly wage is $50.

• Usability Engineer and Developer fully loaded hourly wage

is $150.

International (for one hypothetical country) Parameters:

• The site is an e-business site (vs. a product information

site, or a site based on an advertising model).

• The current site gets an average of 75,000 visitors per

month in the international country.

• The current buy-to-look ratio in the international

country is 1%.

• The current average profit margin on each online purchase

is $25.

• The current rate of usage of the “Call Back” button

(serviced by a local customer service organization) is 4%.

• Average time to service each usage of the “Call Back”

button is 10 minutes.

• Users are paid to participate in usability engineering

tasks at an hourly rate of $20.

• Customer support fully loaded hourly wage is $25.

• Usability Engineer and Developer fully loaded hourly wage

is $150.

• Language/culture consultant is a graduate student living

in the location of the development organization in the base
country, paid a flat rate of $75 per hour.

Fully loaded hourly wages are calculated by adding together

the costs of

salary, benefits, office space, equipment, and other

facilities for a type of

personnel, and dividing this by the number of hours worked

each year by

that personnel type. If outside consultants or contractors

are used, their

simple hourly rate would apply.

The hourly rate for users used in this cost estimate is $40
in the base country

and $20 in the international country. This is not based on

a typical user’s fully

loaded hourly wage at his or her job, which it would be in

the case of a cost justi

fication of traditional software development for internal

users or intranet de

velopment for internal users. Instead, it is premised on

the assumption that

test users in the case of an e-commerce web site have to be

recruited from the

general public to participate in usability engineering

tasks/techniques, and

that they are paid at a rate of $40 (or $20) an hour for
their participation.

The fully loaded hourly rate for usability engineering

staff is based

roughly on a typical current salary and benefits for a

senior-level internal

usability engineer in the United States. The hourly rate of

developers was

similarly estimated.

It should be emphasized that when using the general

cost–benefit analy

sis technique illustrated here, the particular parameter

values used in the

sample analyses should not be assumed. The particular

parameter values of

your project and organization and your base and

international countries

should be substituted for those given. They will almost

certainly be different

from the parameters used in this example. In particular,

professionals

hired in different countries may have very different rates

in the designing

organization’s currency.

Calculate the Costs of the Usability Engineering Plan

Given a usability engineering plan, the number of hours

required of each

necessary type of staff can be estimated for each

task/technique in the plan.

Once you can estimate the number of hours required for each
task/technique,

you then simply multiply the total number of hours of each

type of staff re

quired for each task by the fully loaded hourly wage of

that type of staff, and
then sum across the different staff types. Additional
costs, such as equipment,

supplies, travel expenses and special services, should also

be estimated and

added in for each task. This is how the task costs in Table
8.1 were calculated.

The numbers of hours estimated for each task/technique in

the cost sum

mary table were not pulled out of a hat—they are

hypothetical, but based on

many years of experience. For example, the 104 hours

estimated to conduct

the first usability test in the base country (in this case,
the United States) was

derived as follows: • Design/develop test materials 40 hrs

• 2 days’ travel 16 • Run test/collect data (3 days, 8 test
users) 24 • Summarize/interpret data; draw conclusions 8 •
Document results 16 • TOTAL 104

Similar sample breakdowns of the steps required in each

Usability Engi

neering Lifecycle task/technique can be found in Mayhew

(1999) and in

Bias and Mayhew (1994).

The cost differences for the international usability

engineering program in

dicated in Table 8.1 are also hypothetical, but again, are

extrapolated from the

actual experience of some of our colleagues conducting such

projects. For ex

ample, in the case of the same first usability test, the

additional 56 hours of the

usability engineer’s time is based on a number of

anticipated factors:

• More time spent in preparation for the testing due to the

complex ity of making international travel and support
staff arrangements.

• More travel time (actual in-transit time, plus a weekend

stay-over required to be on site the required number of
weekdays).

• The assumption of more no shows in the target country,

which means more users must be scheduled in order to meet
the goal of eight data points, which in turn means the
usability engineer must plan to spend more time at the
testing site.

• Time spent training the local facilitator before testing

begins.

• Time spent simply picking up basic cross-cultural

differences (re ferred to earlier in this chapter). It is
assumed that basic data were provided, but that additional
data will be pursued and picked up during testing.

Also note that there are cost categories in Table 8.1 that
are given as flat

fees, rather than broken down into staff types, numbers of

hours of each

staff type, and hourly rates. For example, flat fees are
given for services such

as Recruitment of Users, Local Facilitator, and

Simultaneous Translation.

These dollar amounts are extrapolated from the actual

experience of col

leagues in countries where service providers quoted a flat

fee for the service

agreed to, rather than an hourly rate.

The difference in travel expenses for the base and

international coun

tries is primarily due to the expectation of higher

airfares for international

travel. Whereas more travel days are required, both

in-transit and on-site,
in order to run the same number of test users as in the
base country, it is ex

pected that hotel, meal, and rental car costs will be less
in the international

country, making overall travel expenses comparable in the

two countries.

Remember that both the hourly wage figures and the

predicted hours and

expenses per cost category for each task used to generate

the sample analyses

here are just hypothetical examples based on specific

experiences. Again, you

have to use the actual costs and time of personnel and

other expenses in your

country and in your target country(s), in order to carry

out your own analysis.

Select Relevant Benefit Categories

Since this is an e-business site, only certain benefit

categories are of rele

vance to the business goals of this site. For another kind

of site, different

benefit categories would be selected.

In this hypothetical case, the project usability engineer

decides to in

clude the following benefits: (a) increased buy-to-look

ratio, (b) decreased

abandoned shopping carts, and (c) decreased use of “Call

Back” button.

These categories were selected because the usability

engineer knew these

would be of most relevance to the analysis audience, the

business sponsors

of the site. There may be other very real potential

benefits of the usability

engineering plan, but the usability engineer chose just

these for simplicity

and to make a conservative estimate of benefits.

As compared to the existing site design, the usability

engineer anticipated

that in the course of redesign, the usability engineering

effort would decrease

abandoned shopping carts by insuring that the checkout

process is clear, effi

cient, provides all the right information at the right

time, does not violate any

cultural expectations or values, and does not bother users

with tedious data en

try of information they do not want or need to provide. The

usability engineer

expected to improve the buy-to-look ratio by insuring that

the right product

information is contained in the site, that navigation to

find products is efficient

and always successful, and that no cultural blunders are

made in terms of prod

uct names, site use of colors, and so forth. The usability

engineer expected to

decrease the use of the “Call Back” button by making the

information architec

ture match user expectations by designing and validating a

clear conceptual

model, so that navigation of, and interactions with, the

site are intuitively obvi

ous, and by insuring that language translation is not

misleading or confusing.

Accomplishing all these things depends on conducting the

requirements anal

ysis, design and testing activities in the proposed plan

with an eye towards un

covering and addressing unique requirements in the

international country, as

well as on applying general user interface design expertise.

Quantify/Estimate Benefits

Next, the project usability engineer estimates the

magnitude of each

benefit that would be realized (compared to the current

site, which is being

redesigned) if the usability engineering plan (with its

associated costs) were

implemented. Thus, for example, she or he estimates how

much higher the

buy-to-look ratio would be on the site if it were

reengineered for usability as

compared to the existing site.

To estimate each benefit, you must choose a unit of

measurement for the

benefit, such as the average purchase profit margin in the

case of the In

creased Buy-to-Look Ratio benefit, or the average cost of

customer support

time spent servicing each usage of the “Call Back” button,

in the case of the

Reduced Use of the “Call Back” Button benefit. Then—and

this is the tricky

part—you must make an assumption concerning the magnitude

of the ben

efit for each unit of measurement, for example, a 1%

increase in buy-to-look
ratio, or a 1% decrease in the usage rate of the “Call
Back” button. (Tips on

how to make these key assumptions are discussed later.)

Finally, you will cal

culate the total benefit in each category based on the unit

of measurement,

key parameters, and your assumptions about magnitudes of

benefit. When

the unit of measurement is time, benefits can be expressed

first in units of

time and then converted to dollars, given the value of time.

Remember that our hypothetical project involves development

of an

e-commerce site. Based in part on in-house experience, our

project usabil

ity engineer makes the following key assumptions:

• Buy-to-look ratio will increase by .5% of total visitors

in the United States and by 1% in the international
country.

• Abandoned shopping carts will decrease by .5% of total

visitors in the United States and by 1% in the
international country.

• Usage of the “Call Back” button will decrease by .5% of

total visitors in the United States and by 1% in the
international country.

Note that the assumption is that there will be greater

benefits in the in

ternational country than in the base country. Remember that

currently the

international country is using the site designed in English

for a U.S. audi

ence. Any general usability problems are thus compounded by

the fact that

the site is also not in the user’s native language, nor

does it take into account

other cross-cultural differences, such as preferred

measurement units, cul

ture-appropriate imagery, culture-appropriate use of color,

and currency

differences. Thus, due to decreasing the general

“alien-ness” of the site for

the international audience, as well as improving the

general usability of the

site, greater benefits are predicted.

Also note that very conservative assumptions regarding

predicted bene

fits are made, in spite of the aggressiveness and

thoroughness of the usabil

ity engineering project plan. A 0.5% increase in the

buy-to-look ratio in the

base country is not much. However, making conservative

benefits estimates

in a cost justification analysis such as this is always

wise: If you can show an

overall benefit, even when your costs are high and your
claims regarding

expected benefits are very conservative, then you have a

compelling argu

ment for your plan.

Based on these key assumptions, the project usability

engineer then cal

culates benefits in each of the selected benefit categories

as follows:

Increased Buy-to-Look Ratio: Base Country: .5% more

visitors will decide to buy, and will successfully make a
purchase each month 125,000 visitors per month × .5% = 625
more purchases 625 purchases @ profit margin of $25 =
$15,625 per month International Country: 1% more visitors
will decide to buy, and will successfully make a purchase
each month 75,000 visitors per month × 1% = 750 more
purchases 750 purchases @ profit margin of $25 = $18,750
per month

Decreased Abandoned Shopping Carts: Base Country: .5% more

visitors who would have decided to buy anyway will now
complete checkout successfully and make a purchase each
month 125,000 visitors per month × .5% = 625 more purchases
625 purchases @ profit margin of $25 = $15,625 per month
International Country: 1% more visitors will decide to buy,
and will successfully make a purchase each month 75,000
visitors per month × 1% = 750 more purchases 750 purchases
@ profit margin of $25 = $18,750 per month

Decreased Usage of “Call Back” Button: Base Country: .5%

fewer visitors will need to use the “Call Back” button each
month 125,000 visitors per month × 5% = 625 fewer calls 625
calls at 5 minutes each = 52.08 hours 52.08 hours @ $50 =
$2,604 per month International Country: 1% fewer visitors
will need to use the “Call Back” button each month 75,000
visitors per month × 1% = 750 fewer calls 750 calls at 10
minutes each 125 hours 125 hours @ $25 = $3,125 per month

This is how the benefit predictions summarized in Table 8.2

were made.

The usability engineer based the assumptions regarding

benefits in the

base country on statistics available in the literature,

such as those presented

in Mayhew and Bias (1994). In particular, she or he began

with the often

quoted average e-commerce web site buy-to-look ratio of

2–3%. The usabil

ity engineer then based the assumption that this rate could
be improved by

a minimum of 1% (.5% from improving the product search

process and .5%

from improving the checkout process) through usability

engineering tech

niques, on statistics offered as average by a Forrester

report called “Get ROI
From Design” (Souza, 2001). This report suggested that it
would be typical

for as much as 5% of online shoppers to fail to find the

product and offer

they are looking for (other statistics suggest as much as

45% may experience

this problem), and for over 50% of shoppers that do find a

product they

would like to buy to bail out during checkout due to

confusion over forced

registration, being asked to give a credit card number

before being shown

total costs, and because they are not convinced this is the
best available of

fer. The assumption of reduced usage of the “Call Back”

button by .5% was

based on statistics cited earlier suggesting as much as 20%

of site users typi

cally call in to get more information. Most of us have

experienced these

problems and would have little argument with the idea that
they are typical.

Given these statistics, the assumptions made for the base

country seem very

modest indeed. The slightly greater benefits assumptions

made for the in

ternational country were based on the expectation that

decreasing the gen

eral “alien-ness” of the site for the international

audience would provide

benefits beyond improving the general usability of the site.

The basic assumption of a cost–benefit analysis of a

usability engineering
plan is that the improved user interfaces that are achieved
through usability

engineering techniques result in such tangible, measurable

benefits as

those calculated in this hypothetical example.

The audience for the analysis is asked to accept these

assumptions of cer

tain estimated, quantified benefits as reasonable and

likely minimum benefits,

rather than as precise, proven, guaranteed benefits. Proof

simply does not

exist for each specific web site that an optimal user

interface provides some

specific, reliable advantage over some other user interface

that would, or

has, resulted in the absence of a usability engineering

plan.

How can you generate, and convince your audience to accept,

the inher

ent assumptions in the benefits you estimate in a given

cost–benefit analy

sis? First, it should be pointed out that any cost–benefit

analysis for any

purpose must make certain assumptions that are really only

predictions of

the likely outcome of investments of various sorts. The

whole point of a

cost–benefit analysis is to try to evaluate in advance, in

a situation in which

there is some element of uncertainty, the likelihood that

an investment will

pay off. The trick is in basing the predictions of

uncertainties on a firm
foundation of known facts. In the case of a cost–benefit
analysis of a planned

usability engineering program, there are several

foundations on which to

build realistic predictions of benefits.

First, there is ample research published that shows

measurable and sig

nificant performance advantages of specific user interface

design alterna

tives (as compared to other alternatives) under certain

circumstances.

Ben Shneiderman (a well-known author in the field of

usability engineer

ing) states that “Performance decreases 5% to10% when

designers change the

color and position of interface elements. It slows down 25%

when developers

switch terminology on buttons like ‘submit’ and ‘search’”

(Sonderegger, 1998).

In another example, Lee and Bowers (1997) studied the

conditions un

der which the most learning occurred from presented

information and

found that compared to a control group that had no learning

opportunity:

• Hearing spoken text and looking at graphics resulted in

91% more learning.

• Looking at graphics alone—63% more.

• Reading printed text plus looking at graphics—56% more.

• Listening to spoken text, reading text, and looking at

graph ics—46% more.

• Hearing spoken text plus reading printed text—32% more.

• Reading printed text alone—12% more.

• Hearing spoken text alone—7% more.

Clearly, how information is presented makes a significant

difference in

what is retained.

The research does not provide simple generic answers to

design ques

tions. However, what the research does provide are general

ideas of the

magnitude of performance differences that can occur between

optimal

and sub-optimal interface design alternatives. The basic

benefit assump

tions made in any cost–benefit analysis can thus be

generated and de

fended by referring to the wide body of published research

data that

exists. From these studies, you can extrapolate to make

some reasonable

predictions about the order of magnitude of differences you

might expect

to see in web site user interfaces optimized through

usability engineering

techniques. (See Bias & Mayhew, 1994, chap. 2, for a review

and analysis of

some of the literature for the purpose of defending benefit

assumptions.)

Besides citing relevant research literature, there are

other ways to ar

rive at and defend your benefit assumptions. Actual case

histories of the

benefits achieved as a result of applying usability

engineering tech

niques are very useful in helping to defend the benefits

assumptions of a

particular cost–benefit analysis. A few published case

histories exist (e.g.,

Bias & Mayhew, 1994; Karat, 1989). Wixon and Wilson (1997)
and

Whiteside Bennett, and Holtzblatt (1988) report that across

their experi

ence with many projects over many years, they find that
they average an

overall performance improvement of about 30% when at least

70% to

80% of the problems they identify during usability testing

are addressed

by designers.

But even anecdotes are useful. For example, a colleague

working at a

vendor company once told us that they had compared customer

support

calls on a product for which they had recently developed

and introduced a

new, usability-engineered release. Calls to customer

support after the new

release were decreased by 30%. This savings greatly

outweighed the cost of

the usability engineering effort. (For another, even more

dramatic anec

dote, see the example cited by Nielsen, 1993, p. 84,

regarding an upgrade

to a spreadsheet application.)

Finally, experienced usability engineers can draw on their

own general
experience evaluating and testing software user interfaces,
and their spe

cific experience with a particular development

organization. Familiarity

with typical interface designs from a particular

development organization

allows the usability engineer to decide how much

improvement to expect

from applying usability engineering techniques in that

organization. If the

designers are generally untrained and inexperienced in

interface design

and typically design poor interfaces, the usability

engineer would feel com

fortable and justified defending more aggressive benefits

claims. On the

other hand, if the usability engineer knows the development

organization

to be quite experienced and effective in interface design,

then more conser

vative estimates of benefits would be appropriate, on the

assumption that

usability engineering techniques result in fine tuning of

the interface, but

not radical improvements. The usability engineer can also

assess typical in

terfaces from a given development organization against

well-known and ac

cepted design principles, usability test results, and

research literature to

help defend the assumptions made when estimating benefits.

In general, it is usually wise to make very conservative

benefit assump
tions. This is because any cost–benefit analysis has an
intended audience,

who must be convinced that benefits will likely outweigh

costs. Assump

tions that are very conservative are less likely to be

challenged by the rele

vant audience, thus increasing the likelihood of acceptance

of the analysis

conclusions. In addition, conservative benefits assumptions

help to man

age expectations. It is always better to achieve a greater

benefit than was

predicted in the cost–benefit analysis, than to achieve

less benefit, even if

it still outweighs the costs. Having underestimated

benefits will make fu

ture cost–benefit analyses more credible and more readily

accepted.

When each relevant benefit has been calculated for a common

unit of

time (e.g., per month or per year), then it is time to add

up all benefit cate

gory estimates for a benefit total.

Compare Costs to Benefits

Recall that in our hypothetical project, the project

usability engineer com

pared costs to benefits, and this is what was found: Base

Country: Benefits per month = $33,854 One-time cost =
$160,605 PAY-OFF PERIOD = 4.74 months International
Country: Benefits per month = $40,625 One-time cost =
$252,880 PAY-OFF PERIOD = 6.23 months

Our project usability engineer’s initial usability

engineering plan for both
countries appears to be well justified. It is a fairly
aggressive plan (in that it in

cludes all lifecycle tasks and moderate to very rigorous

techniques for each

task), and the benefit assumptions are fairly conservative.

Given the very clear

net benefit, it would be wise to stick with this aggressive

plan and submit it to

project management for approval. In fact, based on the very

modest assump

tions, it might be well advised to redesign the usability

engineering plan to be

even more thorough and aggressive because increased

benefits that might be

realized by a more rigorous approach will likely be more

than compensated

for. In this case, the usability engineer might consider

increasing the level of ef

fort of the requirements analysis tasks, which usually have

a high pay-off.

If the estimated pay-off period is long, or in fact if

there is no reasonable

pay-off period, then it would be wise to go back and

rethink the plan, scaling

back on the rigorousness of techniques for certain tasks,

and even eliminat

ing some tasks (e.g., collapsing the design process from

two to one design

level, i.e., doing only one usability test) in order to

reduce the costs.

A Cross-Cultural Product Information Web Site

This example is based on a hypothetical scenario given in a

Forrester report
called “Get ROI From Design” (Souza, 2001). It involves an
automobile

manufacturing company, who has put up a web site that

allows customers to

get information about the features of the different models

of cars they offer,

and options available on those cars. It allows users to

configure a base model

with options of their choice and get sticker price

information. Users cannot

purchase a car online through this site; it is meant to

generate leads and

point users to dealerships and salespeople in their area.

Start With a Usability Engineering Plan

In this example, we again start with the same assumed plan

as in the

e-commerce site example (see Table 8.1).

Establish Analysis Parameters

Sample parameters are given here for this example. Again,

we assume

there is an existing site with known traffic and sales

statistics, and the pro

ject involves a redesign.

Base Country Parameters:

• The site is a product information site (vs. an e-business

site, or a site based on an advertising model).

• The current site gets an average of 500,000 visitors per

month (vs., for example, 125,000 visitors or 10,000
visitors).

• Currently 1% of visitors result in a concrete lead.

• Currently 10% of leads generate a sale.

• The profit on a sale averages $300.

• Users are paid to participate in usability engineering

tasks at an hourly rate of $40.

• UE and Developer fully loaded hourly wage is $150.

International Country Parameters:

• The site is a product information site (vs. an e-business

site, or a site based on an advertising model).

• The current site gets an average of 250,000 visitors per

month (vs., for example, 125,000 visitors or 10,000
visitors).

• Currently .5% of visitors result in a concrete lead.

• Currently 5% of leads generate a sale.

• The profit on a sale averages $300.

• Users are paid to participate in usability engineering

tasks at an hourly rate of $20.

• UE and Developer fully loaded hourly wage is $150.

Calculate the Costs of the Usability Engineering Plan

We can use the same cost calculations as before, shown in

Table 8.1.

Select Relevant Benefit Categories

Because this is a product information site, only certain

benefit categories are

of relevance to the business goals of this redesign

project. The project us

ability engineer decides to include only the following

benefit category: in

creased lead generation.

The usability engineer selected this benefit category

because it is of most

relevance to the audience for the analysis, the business

sponsors of the site.
There may be other very real potential benefits of the
usability engineering

plan, but only this one was chosen, for simplicity and to
make a conservative

estimate of benefits.

As compared to the existing site design, the usability

engineer antici

pates that in the course of redesign, the usability

engineering effort will in

crease leads by insuring that visitors can find basic

information and

successfully configure models with options. Accomplishing

this depends on

conducting the requirements analysis and testing activities

in the proposed

plan, as well as on applying general user interface design

expertise.

Quantify/Estimate Benefits

Next, the project usability engineer estimates the

magnitude of the benefit

that would be realized (relative to the current site, which

is being redesigned)

if the usability engineering plan (with its associated

costs) were implemented.

Thus, in this case, she or he estimates how much higher the

lead generation

rate would be on the site if it were reengineered for

usability as compared to

the existing site. Table 8.3 summarizes these benefits

predictions.

The project usability engineer calculated the estimated

benefits as follows:
Increased Lead Generation Rate: Base Country .5% more
visitors will generate a lead 500,000 visitors per month ×
.5% = 2,500 more leads 10% of these new leads will result
in a sale = 250 more sales 250 more sales @ profit margin
of $300 = $75,000 per month International Country: 1% more
visitors will generate a lead 250,000 visitors per month 1%
= 2,500 more leads 5% of these new leads will result in a
sale = 125 more sales 125 more sales @ profit margin of
$300 = $37,500 per month

Compare Costs to Benefits

Next, the usability engineer compares benefits and costs to

determine

the pay-off period: TABLE 8.3 Expected Monthly Benefits for

a Product Information Site

BENEFIT CATEGORY VALUE in Base Country— PER MONTH VALUE in

International Country— PER MONTH

Increased lead generation $75,000 $37,500

TOTAL MONTHLY BENEFITS $75,000 $37,500 Base Country:

Benefits per month = $75,000 One-time cost = $160,605
PAY-OFF PERIOD = 2.14 months International Country:
Benefits per month = $37,500 One-time cost = $252,880
PAY-OFF PERIOD = 6.74 months

Again, our project usability engineer’s initial usability

engineering plan

appears to be well justified. It was a fairly aggressive

plan, in that it included

all lifecycle tasks and moderate to very rigorous

techniques for each task,

and the benefit assumptions were fairly conservative. Given

the relatively

short estimated pay-off period and the expectation of a

site lifetime of much

longer than 7 months, it would be wise to stick with this

aggressive plan and

submit it to project management for approval. In fact,

based on the very
modest assumptions regarding increases in lead generation,
the project us

ability engineer might consider redesigning the usability

engineering plan

to be even more thorough and aggressive because the

increased benefits

that might be realized by a more rigorous approach will

likely be more than

compensated for. In this case, it might be advisable to

increase the level of

effort of the requirements analysis tasks, which usually

have a high pay-off.

SUMMARY AND CONCLUSIONS

The previous examples are based on a simple subset of all

actual costs and

potential benefits, and very simple and basic assumptions

regarding the

value of money. More complex and sophisticated analyses can

be calculated

(see Bias & Mayhew, 1994; Bias et al., 2003.) However,

usually a simple and

straightforward analysis of the type offered in the

previous examples is suf

ficient for the purpose of winning funding for usability

engineering invest

ments in general, or planning appropriate usability

engineering programs

for specific development projects.

One key to applying cost justification to planning

usability engineering

programs for web development projects is the generation of

web traffic and

sales statistics that are relevant and useful. In the

Forrester Report called

“Measuring Web Success” (Schmitt, 1999), it is pointed out

that typical web

traffic statistics do not really tell you what you want to

know. For example,

knowing how many visitors to a product information site

there were this

month, and how many of them viewed each product, does not
tell you how

many customers were satisfied with the information

provided. Similarly,

knowing how many visitors to an e-commerce site made a

purchase this

month and what the average purchase price was does not tell
you why more

shoppers did not buy. And, knowing how many visits a

customer service site

received and how many e-mails were received and responded

to by live cus

tomer service representatives does not tell you how many

customer support

issues were resolved online. When we have traffic software

that can generate

more specific and integrated statistics that allow these

kinds of inferences, it

will become much easier to validate cost justification

estimates after the fact,

and make future cost–benefit analyses much more credible.

Ideally, says this report, “Managers will have self-service

access to relevant

information: Human resources will get reports on job

seekers, web designers

will see detail on failed navigations, and marketing will

receive customer re

tention metrics” (Schmitt, 1999, p. 11). And, “For example,

a marketing di

rector trying to determine why look-to-buy ratios plummet

after 6:00 pm

might discover that more evening users connect through

dial-up modems

that produce a slower and more frustrating shopping

experience” (p. 11).

This report also points out that cross-channel statistics

that track a given

customer are just as important, if not more so, than simple

web traffic statis

tics. For example, if it could be captured that a

particular user visited a cus

tomer service web site, but then shortly thereafter called

the customer service hot

line, this might point to some flaw or omission in the

information in the on

line customer service channel. In addition, if the

customer’s “tracks”

through the web site, as well as the nature of the phone

conversation, could

also be captured and integrated, inferences might be made

about exactly how

the web site failed to address the customer’s need.

This kind of very specific traffic tracking starts to sound

very much like a

sort of ongoing automated usability test, which could

generate data that

would point very directly to specific user interface design

problems and so

lutions, making the case for usability engineering even

stronger.

One might look at typical web development time frames

during the early

years of web development, and notice that while the average

web develop

ment project back then might have taken all of 8 to 12

weeks in total from

planning to launch, the usability engineering program laid

out in the sam

ple analyses in this chapter would itself take an absolute

minimum of 5

months in the base country, and 6 months in the

international country.

How can it make sense to propose a usability engineering

program—most

of which would occur during requirements analysis and

design of the over

all development effort—that takes significantly longer to

carry out than the

typical whole development effort?

Initially, web sites were functionally very simple compared

to most tradi

tional software applications, and so the fact that they

typically took 8 to 12

weeks to develop, as compared to months or even years for

traditional soft

ware applications, made some sense. In addition, very few

were explicitly

intended to be used internationally. Now, however, web

sites and applica

tions have gotten more and more complex and are, in many
cases, much

like traditional applications that happen to be implemented

on a browser

platform. In addition, web applications aimed at the

international market

have gotten much more common. The industry needs to adapt

its notion of

reasonable, feasible, and effective time frames (and

budgets) for develop

ing complex web-based applications for use in multiple

other cultures,

which simply are not the same as simple content-only web

sites intended

only for one’s own culture. This includes adapting its

notion of what kind of

usability engineering techniques should be invested in.

In a report by Forrester Research, Inc., called “Scenario

Design” (their

term for usability engineering), it is pointed out that:

Executives Must Buy Into Realistic Development Time Lines

and Budgets

The mad Internet rush of the late 1990’s produced the

slipshod experiences

that we see today. As firms move forward, they must shed

their misplaced fas

cination with first-mover advantage in favor of lasting

strategies that lean on

quality of experience. • Even single-channel initiatives

will take eight to 12 months. The time required to conduct
field research, interpret the gathered information, and
formulate implementation specs for a new web-based
application will take four to six months. To prototype,
build, and launch the effort will take another four to six
months. This period will lengthen as the number of
scenarios involved rises. • These projects will cost at
least $1.5 million in outside help. Firms will turn to
eCommerce integrators and user experience specialists for
the hard-to-find-experts, technical expertise, and
collaborative methodolo gies required to conduct Scenario
Design. Hiring these outside resources can be costly, with
run rates from $150K to $200K per month. This ex penditure
is in addition to the cost of internal resources, such as
project owners responsible for the effort’s overall success
and IT resources han dling integrations with legacy
systems. (Sonderegger, 2000, p. 12)

We agree that 8 to 12 months is a more realistic time frame

(than 8 to 12

weeks) to develop a usable web site or application that

will provide a decent

ROI—in one’s home country. Sites which must be localized to

multiple other

cultures obviously should expect to take longer. And, if

this is the overall

project time frame, there is ample time in the overall

schedule to carry out a

usability engineering program such as the one laid out

earlier in this chap

ter. The total cost of the sample usability engineering

program offered here

is also well within Sonderegger’s estimate of the costs of

“hard-to-find-ex

perts,” which include “user experience specialists.” The

sample cost justifi

cation analyses offered here, as well as others offered by

a Forrester report

called “Get ROI From Design” (Souza, 2001), suggest that it

is usually fairly

easy to justify a significant time and money investment in

usability engi

neering during the development of web sites. And the

framework and ex

amples presented in this chapter should help you

demonstrate that this is
the case for your international web project.

ACKNOWLEDGMENTS

Portions of this chapter are excerpted or adapted from The

Usability Engi

neering Lifecycle, by D. J. Mayhew, 1999, San Francisco:

Morgan Kaufmann.

Copyright 1999 by D. J. Mayhew. Used with permission.

Portions of this chapter are excerpted or adapted from

Cost-Justifying Us

ability, by R. G. Bias and D. J. Mayhew (Eds.), 1994,

Boston: Academic press.

Portions of this chapter are excerpted or adapted from

“Discount Us

ability vs. Usability gurus: A Middle ground” by D. J.

Mayhew that first ap

peared on

in September, 2001.

Portions of this chapter also appeared in “Cost

Justification” by G. Bias,

D. J. Mayhew, and D. Upmanyu, 2003, in J. Jacko and A.

Sears (Eds.), The

Handbook of Human–Computer Interaction (pp. 1202–1212).

Mahwah, NJ:

Lawrence Erlbaum Associates.

Portions of this chapter also appeared in “Cost Justifying

Web Usabil

ity” by D. J. Mayhew and R. G. Bias, 2003, in J. Ratner

(Ed.), Human Factors

and Web Development (2nd ed., pp. 63–87). Mahwah, NJ:

Lawrence

Erlbaum Associates.
Bias, R. G., & Mayhew, D. J. (Eds.). (1994).
Cost-justifying usability. Boston: Academic

Press.

Bias, R. G., Mayhew, D. J., & Upmanyu, D. (2003). Cost

justification. In J. Jacko & A.

Sears (Eds), Handbook of human–computer interaction (pp.

1202–1212). Mahwah,

NJ: Lawrence Erlbaum Associates.

Karat, C. M. (1989). Cost-benefit analysis of osability

engineering techniques. Pro

ceedings of the Human Factors Society 34th Annual Meeting,

839–843.

Lee, A. Y., & Bowers, A. N. (1997). The effect of

multimedia components on learn

ing. Proceedings of the Human Factors and Ergonomics

Society, 340–344.

Mayhew, D. J., & Bias, R. G. (2003). Cost justifying web

usability. In J. Ratner (Ed.),

Human factors and web development (2nd ed., pp. 63–87).

Mahwah, NJ: Lawrence

Erlbaum Associates.

Nielsen, J. (Ed.). (1990a). Designing user interfaces for

international use. Amsterdam:

Elsevier.

Nielsen, J. (1993). Usability engineering. Morristown, NJ:

Academic Press.

Russo, P., & Boor, S. (1993). How fluent is your interface:

Designing for interna

tional users. Proceedings of INTERCHI, 50–51.

Schmitt, E. (1999). Measuring web success. Cambridge, MA:

Forrester Research.

Sonderegger, P. (1998). The age of net pragmatism.

Cambridge, MA: Forrester Re

search.

Sonderegger, P. (2000). Scenario design. Cambridge, MA:

Forrester Research.

Souza, R. K. (2001). Get ROI From Design. Cambridge, MA:

Forrester Research.

Whiteside, J., Bennet, J., & Holtzblatt, K. (1988).

Usability engineering: Our experi

ence and evolution. In M. Helander (Ed.), Handbook of

human–computer interaction

(pp. 791–817). Amsterdam: North-Holland.

Wixon, D., & Wilson, C. (1997). The usability engineering

framework for product

design and evaluation. In M. Helander, T. K. Landauer, & P.

Prabhu (Eds.), Hand

book of human–computer interaction (2nd ed.). Englewood

Cliffs, NJ: Prentice Hall.
CASE STUDIES

CyberAtlas Staff. (2002, March 21). The world’s online

populations. Available from

http//www.cyberatlas.internet.com/big_picture/geographics/print/0,,5911_

151151,00.html

Greenspan, R. (2002, May 24). Youth and upscale consumers

want an m-lifestyle. Avail

able from

1144611,00.html

Pastore, M. (2001, January 11). Global Internet population

moves away from US. Avail

able from

558061,00.html

Plummer-Andrews, T. (1997, March 1). Children’s animation,

a universal language?

Available from
http//mag.awn.com/index.php3?ltype=all&sort=date&arti

cle_no=809

SpectraCom. (2001, December). Future of the Internet

according to kids: A longitudinal

study. Milwaukee, WI: Author.

SpectraCom. (2001, July). The role of children in

e-commerce. Milwaukee, WI: Author. Chapter

Intercultural Human–Machine

Systems: Empirical Study of User

Requirements in Mainland China

Kerstin Röse

University of Kaiserslautern, Center for Human–Machine

Interaction,
Kaiserslautern, Germany

INTRODUCTION

The globalization trend of our industrial era has largely

intensified

cross-cultural communication among different nations. With

the increasing

exchange of technological products all around the world,

the cultural con

sideration has become an important aspect in human–machine

system de

sign. “There is no denying that culture influences

human-product

interaction” (Hoft, 1996, p. 41). This means that at

present when user-ori

ented machine design is concerned, the cultural background

of the target

users should be well addressed. In the globalization

environment, user-ori

ented design has also become culture-oriented design. This

argument is ac

cepted by both industry and academia.

Germany is one country with obvious export orientation. For

German

machine producers, cultural consciousness has a great

importance for their

durable business success in the international market. In

2001, more than

60% of the total German machinery production was oriented

to export. The

export destinations covered a wide range of international

markets, includ

ing Western Europe, North America, Asia, Latin America,

Eastern Europe, 277
and so on. The non-European market occupied 39.9% of the
total German

machinery export (VDMA, 2002). This quite various machine

export struc

ture motivated a research project at the Center for

Human–machine Inter

action (ZMMI) in 1996, which aimed to find out the

culture-specific user

requirements of non-European markets. This project was

supported by the

German BMBF funds and named INTOPS-1: requirements of the

non-Eu

ropean market on machine design. Through an extensive

on-the-spot in

vestigation in some Asian and North and South American

countries, many

special user requirements were revealed. Interesting

significant results

from this project were published (Zühlke, Romberg, & Röse,

1998).

From 2000 on, based on the observation of the quick

economical de

velopment in mainland China and its implication on the

German ma

chinery industry, one continuous project (INTOPS-2 project)

has been

carried out in the same center to elicit the

culture-specific user require

ments of the Chinese market. The previous experience from

the

INTOPS-1 was taken into consideration in the new project:

On the one

hand, the previous project revealed that especially in

China, the ma

chine users show significantly different requirements on

machines,

which still need to be further researched (Zühlke, Romberg,

& Röse,

1998); on the other hand, the experience from the INTOPS-1

project re

vealed that the participation of an investigator from the

target culture is

important to well analyze the information and interpret the

investiga

tion data, which has been actually fulfilled in the

INTOPS-2 project by

incorporating a Chinese researcher in the project group.

In the following sections of this paper, the most important

aspects of the

INTOPS-2 project, including the approach and interesting or

significant

results, are presented in the form of an overview.

APPROACH TO INTERCULTURAL

HUMAN–MACHINE SYSTEM DESIGN

Culture Research

Cultural influence on human–machine system design can be

found in vari

ous design aspects. The early research in this area

concentrated on the an

thropological and physical characteristics of target users

from non-Western

countries to provide ergonomic design suggestions for

technology trans

fers. In the 1990s, with the quick development of computer

science research
it focused on the cultural influence on computer user
interface design,

mainly on the language issues, information presentation

format, different

information coding, graphical symbols, metaphor, and so

forth. Many

guidelines were formulated (e.g., del Galdo & Nielsen,

1996; Fernandes,

1995; Russo & Boor, 1993), but the quite invisible cultural
influence on hu

man–machine interaction was not identified.

Since mid-1990s, research has concentrated on the quite

invisible cul

tural influences, such as thinking patterns on user

interface design issues,

menu structure, and information presentation mode (e.g.,

Choong, 1996;

Choong & Salvendy, 1998; Dong & Salvendy, 1999; Shih &
Goonetilleke,

1998). Efforts have also been made to deal with the

intercultural user inter

face design in a less pragmatic but more scientific way

(Honold, 1999).

More and more researchers are trying to connect

cross-cultural psychology

with user interface design theories to propose theoretical

and methodologi

cal guidance to culture-specific user interface design

(e.g., Plocher, Garg, &

Chestnut, 1999).

Up to now, there have been plenty of studies on

intercultural human–
machine system design, and there have been a lot of
applicable results. In

the computer software industry, many culture-specific

design features

have been incorporated in many recent software products for

the target

markets. However, in the area of production automation

industry, the lo

calization of machine user interfaces for the target market

is much slower.

In the case of machine export to Asian countries, many

Western compa

nies have only taken into account basic considerations on

user interfaces

(e.g., translation for the target market). Other user

interface design issues

have been neglected. Often, the user interface design

standards for West

ern countries are simply applied to the machines and

actually used by

other user groups with quite different cultural

characteristics. This can re

sult in many problems in human–machine interactions or even

cause seri

ous industrial accidents.

As Honold (1999) pointed out, there are many research

communities

conducting research in the intercultural usability

engineering area.

Therefore, the results are marked by a wide heterogeneity.

No generally

accepted design approaches exist, and there is still a

large gap between
the applicable research results and the practical design
requirements.

“What is the theoretical foundation that allows one not

only to integrate

these rather scattered research results, but also to

predict where cultural

differences will be found in other cultures and products,

and how they

might be addressed through design?” (Plocher et al., 1999,

p. 803). It is

clear that a good understanding of cultural aspects and

their influence on

human–machine system design could help to establish the

theoretical

foundation in this area.

Essentially, three questions about the intercultural

human–machine sys

tem design must be answered:

1. Which aspects of culture should be considered?

2. Which machine design issues should be addressed?

3. How can the culture-specific design features be elicited

through the analysis of target culture?

Culture Model

Developing a culture model by using some cultural variables

is the widely

accepted way to study the culture. However, because the

existent culture

models are mainly targeted at the interpersonal level of

cross-cultural com

munication, their application in intercultural

human–machine system de

sign is limited. In his approach of web user interface

design, Marcus (2001)

used the culture model of Hofstede to connect the five

cultural variables

(power distance, uncertainty avoidance, individualism vs.

collectivism, mas

culinity vs. femininity, longvs. short-term orientation) to

user interface de

sign issues (metaphor, mental model, navigation,

interaction, and

presentation; see also Marcus, chap. 3, this volume). This

has provided one

possibility to connect cultural aspects to user interface

design. However,

due to the significant differences between the web user

interface and the

user interface of production systems, the applicability of

this culture model

in the production system area should be questioned.

A culture model was then developed in the INTOPS-2 project

to organize

the related cultural data in a way that could be easily

applied in machine user

interface design. In reference to the culture metamodel of

Stewart and Bennett

(1991), which classifies culture into two layers—the

objective culture and the

subjective culture—this INTOPS-culture model classifies the

cultural aspects

(which should be well addressed in human–machine

interaction design) into

two groups: cultural mentality and cultural environment

(see Fig. 10.1).

FIG. 10.1. Cultural mentality and cultural environment.

Cultural mentality refers to the psychological and
behavioral characteris

tics of the user group whereas cultural environment refers

to the physical and

organizational portion of the culture that has significant

influence on ma

chine design.

Design Issues

Issues that are affected by cultural conventions should be

well addressed in

culture-oriented machine design. To date, these issues have

been mainly di

rected toward design issues within machine user interface.

They include in

formation presentation issues such as color, icon, layout,

format, and

language. Practical experience has, however, proven that in

the area of pro

duction automation, especially for machine design for Asian

markets, those

design issues “beyond machine user interface” (Bannon &

Bodker, 1991), such

as machine functionality, technological features, service

organization, are

critical for the user’s acceptance of machines (Zühlke,

Romberg, & Röse,

1998). Therefore, in the INTOPS-2 project, issues not only

within user inter

face, but also beyond user interface, have been considered

(see Table 10.1).

Research Approach

One research approach is the description of the way to

elicit culture-specific

design features. Figure 10.2 shows the research approach of

the INTOPS-2

project, which is based on the analysis of Chinese culture.

Two phases are

characterized in the approach.

In the first phase, the cultural characteristics of the

target culture (both

in cultural mentality and in cultural environment) are

formulated through

the analysis of present cultural studies and existent

machine design cases.

In the second phase, the hypotheses of different design

features are gener

ated according to the analyzed cultural characteristics,

and these hypothe TABLE 10.1 Intercultural Human–Machine
System Design Issues

Issues Within User Interfaces Issues Beyond User Interfaces

Information presentation Machine functionality

Dialogue and structure Technological feature

User support Technical documentation

Interface physical design Service

Language General machine design

ses are then verified through investigative methods (e.g.,

test, interview,

and questionnaire) to obtain the targeted culture–specific

human–machine

system design features.

INTOPS-2: INTERCULTURAL HUMAN–MACHINE

SYSTEM DESIGN FOR THE CHINESE MARKET

From 2000 on, the project INTOPS-2 has been conducted at
the Center for

Human–Machine Interaction of the University of

Kaiserslautern, in coop

eration with some German machine producers, with the

purpose of helping

them to localize their machines for the Chinese market more

effectively. In

order that the final results be used as a practical design

aid for the industrial

partners in suitably and effectively adapting their

machines to the Chinese

market, three categories of research topics in the INTOP-2

project were

formulated, covering a wide and complete range of localized

machine de

sign issues for the Chinese market (see Fig. 10.3).

The approach, which was proposed in the last section, was

actually ap

plied in the INTOPS-2 project. In correspondence to the two

characterized

research phases, the following eight research steps were

implemented in

the INTOPS-2 project to elicit the machine design features

for the Chinese

market in practice:

Collection of information material: The information

materials regarding

Chinese cultural studies and existing machine designs in

China were col

lected, which include cultural study literature, Chinese

standards, design

FIG. 10.2. Approach to culture-specific human–machine

system design.

guidelines, machine operation manuals, machine samples,

Chinese

market research reports, et cetera.

Analysis of information material: according to the related

machine design

issues for the Chinese market, the information materials

were analyzed

to get the possible machine design features that are

related to the Chi

nese cultural specifics;

Generation of hypotheses: for those design issues which

have at present very

few directly applicable results from the cultural studies

and existing ma

chine designs, the hypotheses on possible design features

for the Chi

nese market were generated, based on the analysis result of

the

information material;

Choice of investigation methods: three investigation

methods, namely the

test, interview, and questionnaire, have been chosen to

verify the hy

potheses and to elicit the Chinese user’s requirements on

the machine

localization design;

Preparation for the investigation: Cities and organizations

to be visited were

selected, initial contacts were made, investigation

materials (test mate

rial, questionnaire, and interview checklist) were prepared

and trans

lated, and travel was organized;

FIG. 10.3. Topics of INTOPS-2 project.

Implementation of investigation: the on-the-spot

investigation in China was

conducted. The realistic requirements of the Chinese

machine users

were directly investigated through test, interview, and

questionnaire. In

vestigation data were collected in the selected regions in

China;

Analysis of the investigation data: the collected

investigation data were ana

lyzed through some qualitative and quantitative methods, to

verify the

previous generated hypotheses and to obtain more meaningful

localiza

tion design features for the Chinese market;

Interpretation of investigation data: investigation data

are further inter

preted and summarized to formulate the applicable design

suggestions

regarding the corresponding machine localization design

issues for the

Chinese market.

Collection and Analysis of Information Material

Information materials were collected through many possible

ways in either

Germany or China. The meaningful sorts of information

materials include:

Cultural study literature: articles, reports,

dissertations, presentations,
etc. on cross-cultural communication, methodology and
results of cul

ture-specific machine design, Chinese cultural specifics

and cognition,

Chinese machine user interface features, case studies of

machine design

practice in China;

Chinese market research reports: statistic reports and

government regu

lations on Chinese economic development trends,

technological devel

opment of Chinese industry, Chinese machine market demands,

import

characters and regulations;

Chinese machine design guidelines and standards: Chinese

national

standards, standardization administration for imported

machines, espe

cially the mandatory standards for machine design and

Chinese inter

face design, comparison of Chinese standards to ISO or DIN;

Chinese machine design cases: machine samples and user

operation

manuals with realistic reflection to general Chinese

machine design fea

tures and user interface design features (information

coding, layout,

structure, flow, etc.), typical Chinese documentation for

machines, work

ing conditions, working structure.

The information materials were then analyzed to get

meaningful in
formation about the Chinese cultural characteristics and
the machine

design features and interface design features in China. In

the analysis

phase, all of the collected information materials were

structured and

classified according to the research topics of INTOPS-2 and

their contri

butions to the project were extracted and analyzed in a

comprehensive

way. Information materials from different sources were

compared to ver

ify the analysis results.

The design features of some superficial culture-specific

design issues can

be directly obtained from the analysis of the collected

information materi

als. These issues are usually format issues related to

information presenta

tion. Other design features could only be obtained through

more compre

hensive processes, including generating hypotheses and

verifying them

through methods such as tests, questionnaires, and/or

interviews.

Hypotheses on Intercultural Human–Machine System Design for

China

Especially for those culture-specific design issues related

to the invisible

cultural conventions, such as information structure,

dialogue design, user

support, screen layout, and general machine design

features, it is necessary
to generate and verify hypotheses. The generation of
hypotheses is based

on the analysis of the information material. In particular,

the following cul

tural specifics in China have been taken into consideration:

• According to Hofstede’s (1997) cultural variables, the

Chinese culture is characterized as high uncertainty
avoidance, high power distance, obvious long-term
orientation, and collectivism.

• The problem-solving strategy of the Chinese can be

characterized as systematic and synthetic, wherein
understanding of the general problem at the initial stage
is very important.

• Chinese culture is polychronic, having the tendency to do

many things at once and often changing among them.

• The Chinese language is unique. It is essentially

composed of pic torial characters and is more complex in
form than Western alpha betic languages.

• Obedience to teachers and elders is much emphasized by

the Chi nese education system. Students tend to follow
teachers’ instruc tions rather than actively solve
problems.

• The Chinese education system emphasizes knowledge more

than methodology, intensifying the students’ tendency to
solve a prob lem using a well-accepted method rather than
trial and error.

• Abstract visual art is not popular in China. The Chinese

prefer art work with realistic figures. Traditional Chinese
art is characterized by symmetry.

• Technical standards in China still differ in many ways

from inter national standards, including the application of
colors, symbols, and so on.

• Interpersonal communication between the Chinese is

considered high-context communication, with most meaning
contained within the context. What is said is often not as
important as what is not said.

Based on these cultural specifics in China, several

hypotheses categorized here

in 7 groups were generated. A brief list of these

hypotheses is listed below:

Hypotheses on general machine design Chinese machine users

prefer familiar general machine design fea tures (in form,
color) and would reject design features that are for eign
to them; that is, they accept machines with mature and
familiar technologies rather than totally new technologies
with which they are not familiar;

Hypotheses on general interaction concept Chinese machine

operators prefer a standard configuration of machine user
interface with low flexibility for configuring on their
own;

Hypotheses on menu structure The Chinese machine user

interface has a wider but less deep menu structure with
more menu items appearing on the high (surface) level of
the structure; The menu structure of the Chinese machine
user interface should facilitate the parallel processing of
more than one task more conveniently;

Hypotheses on interaction navigation The navigation of the

Chinese user interface should provide very strict and
definite ways to follow rather than to allow the operators
to decide how to proceed further;

Hypotheses on user interface layout All the menu items of

the main menu and second menu layer should better be laid
out on the first display mask of the Chinese machine user
interface; The user interface should provide visualization
methods to show the general menu structure on display;
Chinese machine operators prefer the vertical menu layout
of the Chinese interface to the horizontal one; Chinese
users prefer the symmetrical layout of information on a
user interface;

Hypotheses on information coding The color coding of the

Chinese machine user interface is different from that of
the Western standard. Western color coding application can
not be correctly interpreted by the Chinese machine
operators; The use of the Chinese textual characters,
together with graphical symbols, is typical of the Chinese
machine user interface;

Hypotheses on user support information Support information

is required by Chinese operators for basic and sim ple
machine functions. For complicated problems, they tend to
search for help from colleagues instead of from the machine
user interface; Help information is better organized in the
form of a tutorial pro gram to facilitate the learning of
the operator in this process; Graphical presentation of the
help information is dominant over tex tual presentation in
presenting the operation process; Help information
concentrates more on the presentation of the operation
process than on the explanation of technical principles of
the machine.

Investigation Methods

The following three investigation methods are applied in

the INTOPS-2

project: test, questionnaire, and interview. These methods

have been

proven by the previous project INTOPS-1 to be quite

effective in investigat

ing user requirements on culture-specific machine design

(see also Zühlke,

Romberg, & Röse, 1998). However, due to the different

purpose of these

two projects (demonstrating cultural influence on machine

design for

INTOPS-1 and finding out definite design features on the

Chinese market

for INTOPS-2), wider and more detailed requirements of the

Chinese ma

chine users on machine design have been concerned.

Therefore, a few new

tests and more detailed questionnaire and interview

checklist have been de

veloped for the INTOPS-2 project. An overview of all the

implemented

tests is presented in Table 10.2. The main information

categories within the

questionnaire and the interview checklist are shown in

Table 10.3
Furthermore, another quite critical change for the INTOPS-2
project is

the incorporation of a native-speaking Chinese researcher

in the project

group. This has proven to be an indispensable measure,

ensuring the effec

tiveness of the on-the-spot investigation.

The three investigation methods in this project complement

each other.

The tests are mainly formulated to verify the hypotheses

concerning user in

terface design. The questionnaire is mainly applied to

elicit user require

288

1 0

. 2 O v e r v i e w o f T e s t s U s e d i n I N T O P S 2
N o . T e s t A i m M a t e r i a l S u b j e c t A n a l y
s i s 1 P r e f e r e n c e t o c o l o r c o m p o s i t i
o n f o r m a c h i n e t o o l s E l i c i t p r e f e r r
e d c o l o r c o m p o s i t i o n a n d d i f f e r e n c
e t o G e r m a n o n e 1 0 c a r d s w i t h d i f f e r e
n t l y c o l o r e d m a c h i n e t o o l s N o s p e c i
a l r e q u i r e m e n t A v e r a g e p r e f e r e n c e
d e g r e e f o r e a c h c o m p o s i t i o n 2 R e c a l
l p e r f o r m a n c e f o r g r a p h i c a l i n f o r m
a t i o n v s . T e x t u a l i n f o r m a t i o n T e s t
i n f o r m a t i o n p r o c e s s i n g a b i l i t y f o
r d i f f e r e n t i n f o r m a t i o n p r e s e n t a t
i o n m o d e s 3 p i e c e s o f p a p e r w i t h d i f f
e r e n t m o d e s o f p r e s e n t a t i o n : • t e x t
o n l y • p i c t u r e o n l y • t e x t & p i c t u r e N
o s p e c i a l r e q u i r e m e n t A v e r a g e r e c a
l l r a t e f o r e a c h m o d e ; C h a r a c t e r s f o
r b e t t e r i n f o r m a t i o n r e c a l l 3 U n d e r
s t a n d i n g o f c o l o r c o d i n g T e s t u n d e r
s t a n d i n g o f s t a n d a r d c o l o r c o d i n g a
n d d i f f e r e n c e t o G e r m a n o n e 7 s t a n d a
r d c o l o r s o f I E C ; 3 g r o u p s o f c o n c e p t
s i n d a i l y l i f e a n d a t w o r k ( 5 i n e a c h o
n e ) M a t c h i n g f o r c o n c e p t s a t w o r k o n
l y f o r m a c h i n e o p e r a t o r s T h e c o l o r a
s s o c i a t i o n r a t e f o r e a c h c o n c e p t 4 S
y m b o l u n d e r s t a n d i n g T e s t u n d e r s t a
n d i n g o f s t a n d a r d I S O s y m b o l s a n d e l
i c i t t h e p r e f e r r e d s y m b o l c h a r a c t e
r i s t i c s f o r i n f o r m a t i o n c o d i n g I c o
n s f r o m I S O a n d W i n d o w s ; 2 k i n d s o f m a
t e r i a l s : 1 8 i c o n s , e a c h w i t h 3 p o s s i
b l e m e a n i n g s ; 1 4 m e a n i n g s , e a c h w i t
h 3 p o s s i b l e i c o n s M a c h i n e o p e r a t o r
s A v e r a g e r e c o g n i t i o n r a t e f o r e a c h
i c o n ; C h a r a c t e r f o r b e t t e r m a t c h e d
i c o n 289

F a

i l i

a r

i t y

w i

t h W i n d o w s i n t e r f a c e

T e

s t

f a

i l i
a r

i t y

i t h

t h

e W i n d o w s i n t e r f a c e

I n

t e

g r

a t

e d

i t h

T e

s t

a c

h i

n e

o p

e r

a t

o r

s
R

e c

o g

n i

t i o

r a

t e

f o

r W i n d o w s i c o n s 6 C o n c e p t o f g r o u p i n
g E l i c i t t h e g r o u p i n g r u l e a n d t h e d i
f f e r e n c e t o G e r m a n o n e 7 4 c a r d s w i t h
d i f f e r e n t C N C ( c o m p u t e r n u m e r i c c o
n t r o l l e d ) m a c h i n e f u n c t i o n s O n l y w
i t h e x p e r i e n c e d C N C m a c h i n e o p e r a t
o r s P r e f e r r e d s t r u c t u r e f o r g r o u p i
n g 7 P r e f e r e n c e t o s c r e e n l a y o u t E l i
c i t f a m i l i a r s c r e e n l a y o u t c h a r a c t
e r s a n d d i f f e r e n c e t o G e r m a n o n e O v e
r 2 0 d i f f e r e n t c a r d s i n f o r m a n d s i z e
r e p r e s e n t i n g t h e s c r e e n e l e m e n t s C
N C m a c h i n e o p e r a t o r s P r e f e r r e d l a y
o u t o f d i f f e r e n t s c r e e n e l e m e n t s 8 U
n d e r s t a n d i n g o f E n g l i s h t e r m s T e s t
E n g l i s h c o m p r e h e n s i o n O n e t a b l e w i
t h 5 4 E n g l i s h t e c h n i c a l t e r m s M a c h i
n e o p e r a t o r s A v e r a g e u n d e r s t a n d i n
g r a t e ; C h a r a c t e r f o r b e t t e r u n d e r s
t a n d i n g

ments on general machine design features and functional

features. These

two methods are both highly structured and can be well

controlled by the in

vestigator in practice, but at the same time they have the

disadvantage of re

stricting the investigation within the definite areas that

have been formulated
by the investigators. Therefore, they are not capable of
eliciting those design

features that are not imagined by the investigator. In

comparison, the inter

view could, in turn, provide the interviewees with more

freedom to express

their requirements in wide range, which could help the

investigator to obtain

knowledge of wider and unexpected requirements of the

target market. The

combination of these three investigation methods was

successful and effec

tive. The application of these investigation methods in

INTOPS-2 to elicit of

design features for different design issues is listed in

Table 10.4.

Before the on-the-spot investigation, different

investigation materials

were prepared in Germany, including the test materials for

a total of eight

tests (see Table 10.2), one questionnaire, and one

interview checklist. All of

these materials were translated into Chinese.

On-the-Spot Investigation

The on-the-spot investigation in China was carried out at

the end of 2000

until early 2001. During three months, 32 Chinese

organizations in

Shanghai, Beijing, and Chongqing were visited, of which 26

were Chinese

industrial enterprises (including Chinese machine tool

producers and TABLE 10.3 Categories of the Questionnaire
and Interview Checklist
Questionnaire Interview Checklist

Basic information about the visited

company Basic information about the visited company

Information about machine purchasing Information about

machine purchasing

Requirements on machine service Application situation of

imported ma chines in practice

Requirements on machine user interface Requirements on

machine user interface

Requirements on machine technical

documentation Requirements on technical documenta tion

Requirements on service Information about work organization
and training

some machine users). The other 6 organizations included

governmental

organizations for import administration and research

institutes for ma

chine user interface design in China.

Even though the investigation material had been carefully

checked in

Germany before the investigation, in practice, a few

investigation materials

caused problems due to the format or expression of some

terms. One exam

ple was with the printing of symbols on separate pieces of

paper in the recall

test, causing subjects to turn pages to remember the

symbols. This inconve

nience had not been found in the pretest phase because at

that time, the test

material had not yet been bound. Following discovery of

this problem, the
test was printed with all symbols appearing on one piece of
paper. Other

small adjustments were made to the questionnaire and

interview checklist.

The tests were usually conducted in the break room. During

the test pro

cess, the presence of other persons in the test area was,

if possible, avoided.

The subjects were all Chinese machine operators. A total of

42 subjects par

ticipated in the tests, most of them in more than one test.

The total test span

was controlled within 30 minutes, which has proven in

practice as almost

the maximum acceptable time for a subject to willingly

participate. Due to

this constraint, none of the subjects participated in more

than four tests. TABLE 10.4 Application of Different
Investigation Methods for Different Design Issues Design
Issues Investigation Methods Interview Questionnaire Test
Machine functionality ++ ++ Technological features ++ ++
Technical documentation + ++ + Service ++ + General machine
design features + ++ + Information presentation + + ++
Dialog and structure ++ + ++ User support + + + Interface
physical design + + Language issues + + +

+ Indicates that these methods are well suitable. B e y o n

d U I W i t h o u t U I

The interviews were mainly conducted in the office of the

interviewees.

Sometimes also other persons were present with the

interviewees to accom

pany the investigator. Sometimes more than one interviewee

with different

responsibilities in one firm were separately involved.

Since the investigator
was a native speaker of Chinese, there was no problem in
language commu

nication. All interviews were conducted in Chinese.

A total of 35 main interviewees were involved. Half of the

interviews were

conducted in Shanghai and the others in Beijing and

Chongqing; 58% of

the interviewees were machine tool users in China (in

automobile, motorcy

cle, and machine tool industry sectors). The interviewees

were either re

sponsible for the machine purchasing decision (the chief

manager, chief

engineer, and factory/workshop director) or responsible for

the actual ma

chine application in practice (the chief engineer,

factory/workshop direc

tor, equipment engineer ,and technician).

Most of the questionnaires were filled out by the same

interviewees directly

after the interviews or by other responsible persons in the

firms. A few ques

tionnaires were left to interviewees to fill out afterward

and then were sent

to the investigator via fax. In comparison, the former way

is much better

than the latter because any lack of clarity in the

questions could directly be

explained by the investigator. A total of 19 questionnaires

were obtained in

the investigation. More than half of the questionnaires

came from Shang

hai. About 63% questionnaires were filled out by the

state-owned firms in

China; 47% of the questionnaires were filled out by the

machine tool users

in China (in automobile, motorcycle, and machine tool

industry sectors).

SELECTED RESULTS OF THE INVESTIGATION IN CHINA

Plenty of investigation data were obtained in China. The

analysis of these data

was conducted using both qualitative and quantitative

methods. Some really

important intercultural human–machine design features were

revealed by the

investigation. The interpretation of these investigation

results would lead to

several important localization design features of use to

German machine pro

ducers adapting their machines to the requirements of the

Chinese market.

Requirements on Design Issues Beyond User Interface

As was mentioned, some user requirements on design issues

beyond user in

terface were elicited through the use of questionnaire and

interview. In the

project INTOPS-2, the user requirements on these issues are

concentrated

in the following aspects:

Quality and Technology. High machine quality and advanced

technol

ogy in imported machines are still critical characteristics

for attracting Chi

nese machine users. In many cases, these characteristics

are the most
important decisive factor influencing the machine purchase.
Very often,

the requirements on machine quality and technology are

largely influenced

by Chinese import policies or other administrative

regulations.

In comparison to the domestic machines in China, the

imported machines

have, at present, a technological advantage, especially in

the following areas,

which could facilitate their competition capability within

the Chinese market:

• The ability to provide machines with significantly higher

working accuracy, higher running reliability, wider working
range, and higher productivity than the other competitors.

• The technical collaboration ability to provide the

complete set of production equipment for large-scale
construction projects.

• The application of information technology in the machines

to rein force their communication between each other and to
connect to local area networks of users.

It should be emphasized that the quality and technology

advantage of im

ported machines might be largely influenced by

environmental factors in

China. Therefore, special measures should be taken to

ensure that the ma

chine can also reach the expected technological

specifications, even under

the following specific physical environment conditions,

which are character

istic of Chinese factories: (a) high humidity at the

workplace, (b) high concen

tration of acid substance in the air, (c) high

concentration of particulates in
the air, (d) insufficient air conditioning, (e) large
temperature differences be

tween different seasons, (f) high fluctuation in voltage

and frequency in elec

tricity supply, and (g) high frequency of electricity

blackout.

Functionality. User type, scale, product range, work

organization, and

personnel qualification vary greatly in China. Requirements

on machine func

tions are also very different. However, there are still two
common require

ments that are characteristic of Chinese machine users.

First, the requirements

on machine functionalities are pragmatic. In most cases,

only those machine

functions applicable to present production needs are

required. Second, Chi

nese machine users emphasize stability and continuity,

which implies that they

would not change the machine too often but would upgrade it
when necessary.

The all-in-one machine design concept will make the

machines too com

plex and too expensive. The specialized design of machine

functions will

cause a too wide product spectrum and leave little room for
functionality

promotion; therefore, neither is suitable for the Chinese

market. The best

way is the modularization design. This way could provide

user require

ments on flexibility, to configure machine functionality

according to their

pragmatic needs, and the possibility to promote machine

functions accord

ing to their future application changes.

Furthermore, considering the high requirement on machine

working

quality and the low qualification of Chinese machine

operators for effective

machine control, the Chinese machine users require that the

machine con

trol functions should be accomplished by automatic systems

without much

interference of the operator. This implies that the machine

control func

tions that are critical to the final machine working

quality comprise the ba

sis of the machine modularization design. On the other

hand, because of

the inexpensive work force in China, many Chinese machine

users require

that many material handling functions (which could be well

accomplished

through manual work) should be allocated to their workers

instead of to the

automatic machine system. These functions, mainly the

auxiliary functions

in material handling, comprise the most significant part of

the functions to

be modularized.

Appearance Features. The imported machines have one special

sym

bolic meaning for the Chinese machine users and their

customers. Therefore
they have mainly the requirements on machine appearance
features that

they should be easily recognized as imported machines

through differentiat

ing appearance to domestic machines. The visual appearance

features such

as form, color composition, machine brand name

visualization, etc. should

have special characteristics. In some cases, the

consideration on these aspects

could to some extent influence the user’s decision to

purchase imported ma

chines. The collected information material from China shows

that the Chi

nese machines have their own appearance features. For

example, the colors

blue and green are dominant but red and purple are seldom,
the brand

names have ascertain and the same with the model numbers.
(e.g., 914: they

will die, 014: you will die, 740 have bad associations in
Chinese).

In most of the cases, the requirement on machine appearance

feature is

not so important as that on machine quality, service and

price, etc. and is

not a decisive factor in machine purchasing. The

translation of machine

brand name in Chinese, the sound of the machine model

names, etc. should

be well considered to reinforce the positive feeling of the

machine and

avoid some negative associations. Especially the

application of some unfor

tunate names in China should be avoided.

Technical Documentation. Technical documents of imported ma

chines are a basic and important support for the Chinese

users to effectively

apply the machine. However, the actual situation of

technical documenta

tion of imported machines in China has not met the user’s

requirements in

many aspects. Very often the machine producers provide only

the original

technical documents (usually in English) for their Chinese

customers. For

those who have already provided the Chinese technical

documents, the

translation quality is very often unsatisfactory.

The users required that the machine producers should well

organize the lo

calization of technical documents into one important aspect

of their product

development. Even in the least scale, the most basic

requirement of the target

users on the technical documentation, e.g. language, should

be well fulfilled.

Document Language. The language requirements are the most

basic

requirements for the technical documentations. Depending on

different

machines and different machine users, the language

requirements of tech

nical documentations can be fulfilled in different scales.

But no matter how,
the following aspects regarding language should be well
considered:

• In principle, all the descriptions except some

internationally ac knowledged technical terms and
abbreviations on technical docu ments should be translated
into Chinese. However, because of the high translation
cost, sometimes only a part of the whole technical
documents is translated. In this case, it should be ensured
that the machine operation part should be translated at
first. This requirement is a little different from that of
machine user interface, because the operation manual
involves mainly the textual description which can not be
well understood by the operators;

• Especially important for the foreign machine producers,

is that there are significant differences in language
habits of different Chinese spoken areas, such as Hong
Kong, Taiwan, Singapore, and Mainland China which would
largely influence the understanding of the translation;

• No matter who has accomplished the translation work, the

translation should be checked in practical application for
some possible misunderstanding of the operators. The
machine producers should also check the usability of their
technical documents in practice and then to revise them if
necessary;

• The language character of textual description of the

operation manual and maintenance instruction should well
correspond with understanding ability of the machine
operators and maintenance technicians. For the operation
manual the language should be more clear and short. The
description of some important opera tions should be
supplied with a brief description of the operation effect.
For the maintenance instruction the description should be
more detailed and complete. The description of some
technical principles should be supplied.

Document Form and Content. The document characteristics on

its

type, content, structure, format, and so on, should be well

addressed accord

ing to the target user’s learning habit and working

organization. Usually, the

machine producers should still do some further

investigation regarding
these detailed requirements on technical documents for
definite types of ma

chines on the Chinese market. Some general suggestions for

well preparing

the technical documents for use in China are summarized

below:

• A complete set of technical documentation should be

supplied to the Chinese machine users. For the purpose of
maintaining ma chines by themselves, many users in China
emphasize the need of some technical drawings, including
electrical connection plan and mechanical outfit drawing of
some machine parts;

• For the sake of good documentation management, measures

should be taken to provide documents which could be more
easily preserved by the Chinese users. The traditional
paper document form is proven to be easily lost in China
because of ineffective management organization. Other
alternatives, especially the electrical forms of documents,
or some back-up forms of documents for example on the
internet, should be considered to supply in China in the
future;

• Considering that the Chinese machine operators are always

trained on the job at the machines, it is suggested that
the operation manual should be supplied in combination with
the machine help information on the user interface to
facilitate the operators learning ma chine operations more
conveniently during their work;

• Realistic pictorial illustrations to describe the machine

operation and maintenance procedures are much preferred by
the Chinese users. But the pictorial presentations should
be well arranged with the textual descriptions in the
documents. At the same time the for mat and layout
information to differentiate various kinds of opera tions
and their structure should be well utilized.

Machine Service. Because of the general lack of high

qualified mainte

nance technicians in many Chinese companies and the

critical importance

of imported machines to their production (the high

investigation correlates
with very high exploitation), the Chinese machine users
have very high re

quirements on service for imported machines. The service

quality of the

foreign machine producers influences on a large scale the

final machine us

ability in the user’s factory.

In China the present service quality of many foreign

machine produc

ers is very problematic. Many services provided in China

can not com

pared to the services the producers provided elsewhere. The

user’s

interest in service is often neglected. However, with the

increase of more

foreign machines on the Chinese market, the present lack of

service

could be fatal to their further business in China for some

foreign ma

chine producers. The following requirements on machine

service are

characterized in China:

• With the presentation of machines on the Chinese market

the for eign machine producers should establish a good
local service de partment which covers the service in
before-sale, sale, and after-sale loops. Especially the
before-sale service for machine choice consultation and the
after-sale service for machine problem solving is a lack in
China;

• The most important requirements of Chinese users are

after-sale service including the quick and effective
response to service re quests and prompt spare parts
supply. Most Chinese users empha sized that there should be
one service center in Mainland China (not in Hong Kong or
Taiwan) with qualified service engineers. The best way for
spare parts supply would be a regular spare part bank in
Mainland China. If this is impossible, other measures in
machine design, in service organization etc. should be
taken to reduce the delivery time effectively;

• Considering the geographical constrain to effective

machine service, some tele-service alternatives such as
through telephone, internet, etc. are practical and should
be used much more by the foreign machine producers in the
future;

• Training for machine operators as well as for maintenance

techni cians is much required by the Chinese machine users,
especially the effective training for maintenance
technicians is of high impor tance for them. More complete
training programs should be pro vided in China, including
some regular or irregular training for machine operators or
maintenance technicians after some time of machine
application.

Requirements on Design Issues Within User Interface

The user’s requirements on human–machine system issues

within user in

terface have been elicited through either tests, or

interviews and question

naires. The investigation results have indicated that

although the present

competitive aspects of imported machines on the Chinese

market are still

mainly in machine quality, price, service, etc., in the

near future, the com

petition in user-oriented machine design and user

friendliness will be a sig

nificant factor to distinguish different machine producers.

Even today, a

special approach for user interface aspects could also add

extra machine

competition capability on the market. The main requirements

on the ma

chine user interface design are summarized next.

Language. The interaction language requirements are the
most basic

and important requirements of the Chinese users for machine

user inter

face. For most of the Chinese users the machine user

interface in Chinese is

a significant sign that the machines have been localized in

China. A Chinese

user interface is usually taken as the largest convenience

provided by the

foreign machine producers, because it could give the

Chinese machine us

ers a direct confidence that the foreign machine could be

well operated

without great understanding problems.

The test of English term comprehension revealed that the

Chinese

machine operators have generally quite limited

understanding ability for

English terms applied in machine operations. The

understanding rate for all

tested English terms is 46.6%. The correct understanding

rate for some

representative English terms is shown in Table 10.5. The

best understood

English terms include: OK (95.7%), OFF (82.6%), Stop

(82.6%), Home (73.9%),

Help (73.9%), kg (73.9%), ON (69.6%), Start (69.6%), mm

(69.6%), etc. They

are all terms which are related to the basic and most often
met machine

operations (on/off, start/stop), or the terms which could

be often met in daily
life (ok, home, help, kg). The worst understood English
terms include: PIN

(8.7%), Icon (13.0%), Esc (21.7%), Undo (21.7%), Disk

(21.7%), inch (26.1%),

rpm (26.1%), Edit (30.4%), Menu (39.1), Shift (39.1%),

etc.. Most of these terms

are related to the computerized system operations (PIN,

icon, disk, edit,

menu, shift), which implies that most of the Chinese

machine operators are

still not familiar with the operation of the computerized

control systems. The

test results have demonstrated the importance of a Chinese

user interface

version for the effective and safe machine operation in

China.

However, the present language issues of the foreign machine

user inter

face in China are not satisfactory. Only a few of them are

provided with Chi

nese user interfaces. A large part of other machines are

provided only with

user interfaces in a foreign language, usually in English

(sometimes also in

German, in Japanese, etc.). TABLE 10.5 Comprehension Rate

of Selected English Terms

Terms OK OFF Stop Home Help kg ON Start mm End Enter Del

% 95.7 82.6 82.6 73.9 73.9 73.9 69.6 69.6 69.6 65.2 56.5
52.2

Terms Exit Reset Shift Menu Edit rpm inch Disk Undo Esc
Icon PIN

% 43.5 39.1 39.1 39.1 30.4 26.1 26.1 26.1 21.7 21.7 13.0 8.7

It is required by the Chinese machine users that the

language issues

should be firstly addressed for the machine with the aim to

be used in

China. However, because the user’s requirement on this

issue is different, in

practice the foreign machine producers have some

flexibility to handle this

localization issue according to their localization

capability. Some sugges

tions for them to well treat this aspect could be:

• Localization of machine user interface regarding language

can be divided into different levels. Two levels of
localization are signifi cantly differentiated: the
translation of machine labels into Chi nese and the
translating of the screen display into Chinese. The former
concerns only the reprinting of indications and text de
scriptions on the machine surface into Chinese but it could
give the first and direct impression of localization
design. The later con cerns the reediting of the software
codes is more complicated and expensive than the former
one;

• The translation of screen displays into Chinese concerns

integrating one Chinese character databank into the control
software and reediting those software parts which present
screen information. It includes mainly the work of
replacing the English text or words with Chinese or
corresponding characters. This involves two levels of work.
The first level is to translate only those parts directly
related to machine operations, which are mainly some
indications or simple text description on the screen. The
second level is to translate all the screen displays,
including the information for machine maintenance into
Chinese. Usually only the first level is fulfilled, because
the amount of work is much less that the com plete
translation. The investigation results, however, have indi
cated that for the Chinese machine users the translation of
the machine maintenance part is more important than that of
the op eration part. With this part there are more
understanding prob lems. This requirement from the
investigation is quite different to the current
localization cases;

• In many cases, the machine user interface would be used

by differ ent persons with different language backgrounds,
for example, the service engineers from Germany teach the
machine operation in China. Therefore, if it is possible,
one dual language user interface would be a good
alternative to the user interface separated for each
operator or the preselection of language and in dependence
of the individual user profile;

• It is clear from the investigation that the requirement

on interac tion language is at present only on the
information presentation, namely the display aspect. There
are some requirements on infor mation input aspects,
because the Chinese operators don’t need to input the
Chinese characters. Most of the input (e.g., program ming)
could be accomplished through some codes. This means that
the Chinese input is not yet needed for present machines
and the input could still be kept as alphabetic.

Menu Structure. It was assumed before the investigation

that the dif

ferent requirements of the Chinese users on the menu

structure would be

influenced by the different thinking patterns of the

Chinese to the German.

However, in practice, the most significantly (or most

obviously) different re

quirement on this aspect comes from the different working

organization.

One of the significant characters of working organization

in China is the

more elaborate division of operation tasks. A Chinese

operator’s task range

is more narrow than that for a German machine operator. The

effectiveness

of the German machine user interface to support the working

organization

in China is problematic. In fact, this problem has been

reflected by some

Chinese interviewees they require a self-defined user

interface to fit their
individual production tasks. Some requirements regarding
the menu struc

ture are proposed by the Chinese machine users:

• At first, the menu structure should well correspond to

the operation tasks of the Chinese machine operators. The
new menu structure should be characterized by the separated
(and hierarchical) visiting right of different machine
operators to different machine functions, with a simpler
menu structure for each operator group. Especially for the
machine operators the operation functions should be much
simpler. This could also make the operation of def inite
machine function more quickly to be reached;

• Because the actual working organization for each customer

is differ ent, it is impossible for the machine producers
to provide for each customer his one individually menu
structure. In practice, there should be a balance of
consistence in menu structure provided by the machine
producers and the flexibility of users to configure the
struc ture especially for his definite working
organization. Based on this consideration, the
modularization of the menu structure to leave room for the
further adaptation is a good alternative. Then the menu
structure for a specific customer could be configured by
the producer according to his individual need. This freedom
to define their own user interface through some openness of
the system has been required by many Chinese machine users.

Dialogue and Navigation. Some direct or indirect

information has

been obtained from the investigation which is involved in

the design of dia

logue forms and navigation system for the user interface.

From the prob

lems that the interviewees often reflected about their

operators and the

self-evaluation of the machine operators about their own

work, some re

quirements of the Chinese users on machine dialogue and

navigation sys

tem is revealed. The following points should be helpful for

the foreign

producers to design user interface for the Chinese

operators:

• The Chinese customers have very low assessment of their

own ma chine operators. The operators themselves have also
very low eval uation about themselves. It is often remarked
by both of them that the operators have quite low
qualification and could not under stand the complicated
machine operations. This reflects that the machine
operators could have potential fear to actively interact
with the machine and would prefer to follow the definite
operation instructions from the system. It is then required
that the dialogue system should provide more error
tolerance to the operation and should have a very clear
guide for the operators to let them follow the operation
process;

• Secondly, from the questionnaire results it can be found

that the main problem of the Chinese machine operators in
machine operation are based in the understanding of
different machine functions and different operation
processes. It is then required that the user interface
should provide a navigation concept which presents the
general machine functions and operation processes in an
overview way, to let the operators establish one clearer
overall image of machine operation more easily;

• Because of the general doubt of the Chinese machine users

about the qualification of their operators, it is required
that the possibility to intervene by operators in machine
control operation should be low. This means that the
machine options and settings should be more standardized
and offer minimal possibilities for the opera tors to
change. It is then suggested that those dialogue forms
should chosen which have only few options.

Interface Layout. It was hypothesized that the cultural

influences from

language (the reading direction) and aesthetic preference

should have

large influence on preference to screen layout. However,

the screen layout

test has indicated, the preference of the Chinese machine

operators to ma
chine user interface layout is largely influenced by their
actual working ex

perience with some existent machine user interfaces. This

result also means

that for the design of machine user interface layout for

the Chinese market,

the existent user interfaces should be well investigated.

Especially the fol

lowing two points should be considered in interface layout

design:

• The working experience of the Chinese machine operator

with the existent user interface has significant influence
on their pref erence to interface layout. This means that
principally all kinds of interface layouts could be well
adapted and accepted by the Chi nese machine operators in
the long run. It is then suggested that the widely applied
user interface layouts in the Western countries could be
applied in China without significant changes. This su
perficially “simple” way could even be a better way to
facilitate Chinese operators to adapt themselves more
quickly to the im ported machines in the future;

• The Chinese machine operators have little experience with

the op eration of Windows interfaces. Therefore the
application of the Win dows concept for machine operation
in China will meet some problems at the beginning.
Furthermore, the quite free dialog pro vided by Windows
interaction could also make the Chinese machine operators
more unsure about their operations than the guided dia log.
It is then suggested that the application of the Windows
interface in China is not encouraged for machine operation
at present.

Help Information. Chinese machine operators rely much on

help to

operate the machines. At present, this help is mainly from

their colleagues

or masters. The machine system, including the technical

documents, have

provided only little help for the Chinese machine

operators. In the
investigation, many Chinese machine operators expressed
their

requirements for more help on the machine user interface,

which includes

not only the machine trouble diagnosis help information,

but also the

help for normal machine operation or learning. The help

information on

machine user interfaces is, in some ways, even more

important than the

accompanied machine operation manuals, because the

operators are

always trained on the job at the machine, and the operation

manuals are

not allocated to every operator. It is then required that

the provision of

more effective help information on user interface should be

completed by

the machine producers.

Information Coding. Although different information coding

ways can

be found on machine user interface in China, the most

preferred way is

through Chinese text. Almost all Chinese domestic machines

are labeled

with Chinese text to indicate keys, buttons, and other

information. The for

eign machines which have been provided no Chinese labels

have also been

very often labeled by the Chinese users themselves with

Chinese text. The

interview results have also revealed that the Chinese

machine operators
prefer the Chinese text label, when they are required to
choose among dif

ferent information coding ways (e.g., color, symbol). For

them, only the text

label can express the operation information most clearly

without any mis

understanding. The Chinese text is the most reliable and

explicit informa

tion coding in China.

The results of the English comprehension test revealed that

the Chinese

machine operators have generally unsatisfactory English

comprehension

of safe and effective machine operation. Therefore, the

text coding in Eng

lish should be restricted to a very limited scope. More

detailed require

ments at this point are summarized:

• The most important information should be at any time

encoded and usable Chinese text (and with combination to
other informa tion coding ways such as color, shape and
position, etc.);

• Only when it is really necessary, the English texts which

express some general machine operations such as On, Off,
Start, Stop, etc. and are closely related to some daily use
such as Ok, Yes, Help, Home, etc. could be applied for
information coding. But the application of most English
abbreviation in text labeling should be avoided;

• In many cases, other information coding ways should be

applied together with the Chinese text to ensure that the
machine can be well maintained by foreign service
engineers.

The results of the test of understanding of color coding

are presented in

Table 10.6.
Contrary to the previous hypothesis, the results have
revealed that the

Chinese machine operators could already well interpret the

international

standard color coding in working environment (See IEC73).

Especially the

application of color red (danger), green (normal), and

yellow (caution) to

express meanings in working context could not meet any

serious problems.

This implies that the standard color coding applied in

Western countries

could also be applied in China.

The test results have also revealed that the working

experience has large

influence on the understanding of color coding in China.

From Table 10.6

it can be seen that the highest color association in daily

life is 70,4% (red

with luck) and the lowest is 25,9% (blue and black with
power), while the

highest color association in working context is 92,6% (red

with danger) and

the lowest is 40,7% (cyan with OK).

Based on the basic test results, regarding the color

application on user

interface design two points should be well considered:

• The international standard color coding is suggested to

be consis tently applied on machine user interface for the
Chinese market. For the Chinese machine operators the
working experience with machine operation is the main
influence on the interpretation of color coding. Therefore
the consistent application of an interna tional color
coding standard could intensify their understanding;
• The application of color coding should be limited within
possibly fewer colors in order to transfer the information
more clearly. The colors of red, green, and yellow are
suggested to be applied whenever it is necessary. The
application of other colors for information cod ing should
be more careful and restricted within some uncritical in
formation classes. Usually other information coding methods
should be applied together to ensure correct
interpretation.

Although symbols and icons are important information coding

ways to cross

over language barrier among different countries, the

general understanding

ability of the Chinese machine operators to different

international standard

symbols and icons is very low. The average understanding

rate for all the

tested icons and symbols is 44%; of which the average

recognition rate for TABLE 10.6 Percentage Association of
Each Color With Each Concept

Concept Color Total (%) Black Green Cyan Blue Magenta Red
Yellow Luck — 7.4 7.4 3.7 7.4 70.4 3.7 100 Fear 59.3 11.1
3.7 14.8 3.7 — 7.4 100 Hope — 40.7 11.1 3.7 7.4 7.4 29.6
100 Silence 3.7 14.8 55.6 7.4 3.7 — 14.8 100 Power 25.9 7.4
3.7 25.9 11.1 14.8 11.1 100 Danger — — — — — 92.6 7.4 100
Caution 3.7 3.7 3.7 — 11.1 3.7 74.1 100 Normal — 74.1 11.1
7.4 3.7 — 3.7 100 Regulation 44.4 — 22.2 25.9 3.7 — 3.7 100
General Info 11.1 14.8 44.4 18.5 3.7 — 7.4 100 On 7.4 74.1
3.7 — — 7.4 7.4 100 Off 7.4 3.7 — — — 85.2 3.7 100 Help —
3.7 14.8 3.7 11.1 — 66.7 100 OK 7.4 7.4 40.7 18.5 14.8 3.7
7.4 100 Exit 44.4 7.4 18.5 3.7 18.5 — 7.4 100

Total association 58 73 65 36 27 77 69 D a i l y l i f e

I n

f o

r m

a t

i o
n c l a s s O p e r a t i o n

ISO symbols is 45,7% and that for Windows icons is 38,4%.

These unsatisfac

tory recognition rates could not ensure the effective

machine operations in

China, if the information is coded only through symbols or

icons.

Figure 10.4 presents the symbol recognition results

arranged according

to their abstractness. 6 pictorial symbols and 6 abstract

symbols are pre

sented. The average recognition rate for pictorial icons is

61,3% and for ab

stract icons is 38,7%. This means that the recognition

ability of Chinese

machine operators for pictorial symbols and icons is much

better than that

for the abstract one.

Through ANOVA analysis the expected significant statistical

difference

between these two groups has not been found. This means
that the grade of

abstractness of a symbol would not be the decisive factor

to influence the

recognition ability of the Chinese machine operators. In

fact, from the test

it can be seen that some abstract symbols have a very high

recognition rate

but some pictorial symbols have very low recognition rate.

It is assumed that there should be other factors which have

significant in

fluence on recognition ability of the Chinese machine

operators. Further
FIG. 10.4. Recognition rate for pictorial and abstract
symbols.

analysis has revealed one important characteristic which

has significant in

fluence on the recognition ability of the Chinese machine

operators. This

characteristic can be defined as the context degree of a

symbol on its appli

cation areas and helping persons to be correctly

interpreted. The general

symbols have high dependence on the contexts. Their

interpretation de

pends largely on the application area and the interpreter.

The special sym

bols have low dependence on the contexts. Their

interpretation is more

definite and stabile and less dependant on the interpreters

and often used

in a specific application area.

The average recognition rate of the above selected symbols

is then calcu

lated according to this characteristic. The result is shown

in Fig. 10.5. The aver

age recognition rate for special symbols is 67.9% and for

general symbols is

32.1%. Further statistical analysis (ANOVA) has also shown

that the context de

pendence of the symbol has clearly significant influence on

the symbol recog

nition results for the Chinese machine operators. This has

revealed another

important application characteristic of symbols on machine

user interface.
Based on the above mentioned test results, for the
application of sym

bols and icons for information coding attention should be

paid to the fol

lowing points:

FIG. 10.5. Recognition rate for general and special symbols.

• The application scale of icons, symbols on machine user

interface in China should be restricted. For the most
important and critical machine information, other
information coding ways, at best the Chinese text label
should be applied together to ensure the correct
interpretation of the information;

• The application of abstract symbols should at best

avoided in China, especially those symbols which are to be
used to present im portant machine information. In the
cases that the application of abstract symbols is necessary
and unavoidable, for example accord ing to Chinese
mandatory standards, etc., the additional informa tion
coding ways, for example the Chinese text label should also
be applied to ensure the correct information
interpretation;

• The application of pictorial symbols and icons can not

always en sure the correct information interpretation. Even
the interpreta tion of realistic objects can be largely
culture dependent. Many pictorial symbols have meanings
which are too general but not spe cific and can be
differently interpreted in different contexts. Therefore,
for a specific application case, the meaning of the symbols
should be further investigated.

CONCLUSIONS ON HUMAN–MACHINE SYSTEM DESIGN

FOR THE CHINESE MARKET

As the intercultural human–machine system design is

essentially a very

complicated work for any foreign machine producers, it is

necessary to pro

vide one “general guide” which has described clearly the

importance of dif
ferent machine design issues in the whole localization
process to let the

machine producers decide on their own localization design

strategy toward

the target market (Röse, 2002).

The investigation revealed that the user’s requirements on

different ma

chine design issues have obviously different intensity.

Therefore, the local

ization of different human–machine system design issues on

the Chinese

market has different efficiency. According to the

localization cost and the

localization effect, all the previous mentioned design

issues can be orga

nized into three levels, which are shown in Table 10.7.

The localization issues on surface level concerns mainly

some cultural

conventions on surface, such as the language aspects,

information coding

standards, widely accepted formats, etc., which are

normally very explicit

and easy to be found in some standards and norms. The

elicitation of re

quired design features do not require large expenses from

the producers,

and because the design features of these issues remain

usually quite stabile

among different application cases, the whole localization

cost for these is

sues is normally the lowest. On the other hand, since these

issues are the

surface of the machine, they will very easily give the

target user the impres

sion whether the machine is specially localized for them or

not. And, be

cause these issues involve mainly the information

presentation on the

machine, they are critical for the users to well understand

the machine ap

plication. These issues have a very high localization

effect when the applica

tion productivity, safety, etc. are considered. For most of

the application

cases, the localization of the issues on the surface level

could ensure obtain

ing good localization effects with low cost, therefore

these issues have the

highest efficiency and are always localized at first.

The issues on function level reflect the main motivation of

user’s purchase

to the machine. They are the prerequisites of effective

machine application

on the foreign market. The well addressing of user’s

particular require

ments on these issues could largely promote the user’s

acceptance to the

machine, namely these issues have high localization effect.

The user’s re

quirements on these issues are not difficult to be elicited

(through some

market research methods), however, the adaptation design

issues on func

tion level could involve the major change of machine

functionality or even

the significant machine design and production, which could

increase the lo

calization cost significantly. Therefore, the localization

decision on this

level needs careful comparison of the cost and effect and

is usually imple

mented only when the functional features of the machine are

not applicable

on the target market without the adaptations. Sometimes for

some issues on

this level the localization efficiency could be very high

and it could also be

implemented in the early localization phase.

The issues on the interaction level are closely related to

the cultural con

ventions “below the surface,” namely thinking patterns,

problem solving

strategies, etc. of the users. Although they have large

influence on user’s TABLE 10.7 Three Levels of
Human–Machine System Design Issues

Design Levels Design Issues

Interaction General structure issues Dialogue design issues

User support issues, etc.

Function Machine functionality Technological features

Service issues, etc.

Surface Information presentation issues General layout

issues Language issues, etc.

effective interaction with the machine and are quite

important for the ma

chine application, however, because of the high difficulty

to elicit target

user’s characteristics on this level, the localization cost

for these issues

would be much to high. These issues are at present being

analyzed by

more and more researchers and after there are more

applicable design

guidances the producers could also gain much better from

the localization

on this level. But at present, the practical localization

on this level will

meet much difficulty.

The localization efficiency which has been explained above

can be illus

trated by Fig. 10.6. The investigation results in China

have well demon

strated this basic localization strategy. The localization

of the design issues

on the surface level (most significantly on interaction

language) and on the

function level (especially on machine quality and service)

there should be

more emphasize than other design issues.

It is quite clear that the intercultural human–machine

system design is

one good warranty for German machines to be well accepted

on the Chi

nese market. The well address of culture-specific user

requirements could

ensure the benefits of the German machine producers in

China. In this pa

per, the most important and significant user requirements

on imported

machines on the Chinese market are analyzed and summarized

in a way to

make them to be as close to the objective facts as

possible. The machine lo
calization design issues and their design features for the
Chinese market

have provided a solid basis for the German machine

producers to conduct

their machine localization design practice in China.

FIG. 10.6. Localization efficiency of different

localization levels.

All in all, the project INTOPS-2 has constructed one

general inter

cultural human–machine system design structure for the

Chinese market.

Further research is still needed to formulate more detailed

design issues

and design approaches to conduct the machine localization

design in prac

tice. In the near future, a more detailed design style

guide would be the best

way to facilitate the developers to localize their machines

more effectively

for the Chinese market.

ACKNOWLEDGMENTS

Many thanks to the sponsors of this project, including MAN

Roland

Druckmaschinen AG, IWKA AG, Bühler AG Amriswil, and Rittal

Rudolf

Loh GmbH & Co. KG. A special thanks to the PhD student Long
Liu, who

collected the data in Mainland China and worked with high

engagement on

this project and this publication.

Bannon, L. J., & Bodker, S. (1991). Beyond the interface:

Encountering artifacts in
use. In: J. M. Carroll (Ed.), Designing interaction:
Psychology at the human-computer

interaction. (pp. 227–253). Cambridge, UK: Cambridge

University Press.

Choong, Y.-Y. (1996). Design of computer interfaces for the

Chinese population. Unpub

lished doctoral dissertation, Purdue University, West

Lafayette, IN.

Choong, Y.-Y., & Salvendy, G. (1998). Design of icons for

use by Chinese in Mainland

China. Interacting with Computer, 9(4), 417–430.

Del Galdo, E. M., & Nielsen, J. (1996). International user

interfaces. New York: Wiley.

Dong, J. M., & Salvendy, G. (1999). Designing menus for the

Chinese population:

Horizontal or vertical? Behavior & Information Technology,

18(6), 467–471.

Fernandes, T. (1995). Global interface design: A guideline

to designing international user

interfaces. London: Academic Press.

Hofstede, G. (1997). Cultures and organizations: Software

of the mind. New York:

McGraw Hill, 1997.

Hoft, N. (1996). Developing a cultural model. In E. M. del

Galdo & J. Nielsen (Eds.),

International user interfaces (pp. 41–73). New York: Wiley.

Honold, P. (1999). Cross-cultural usability engineering:

Development and state of

the art. In H.-J. Bullinger & J. Ziegler (Eds.),

Proceedings of the 8th International

Conference on Human-Computer Interaction. Human–Computer

Interaction: Ergonom
ics and user interfaces (Vol. 1, pp. 1232–1236). Mahwah,
NJ: Lawrence Erlbaum As

sociates.

Marcus, A. (2001). Cross-cultural user-interface design. In

M. J. Smith, & G.

Salvendy (Eds.), Systems, social and internationalization

design aspects of human–com

puter interaction. Volume 2 of the Proceedings of HCI

International.

Plocher, T. A., Garg, C., & Chestnut, J. (1999). Connecting

culture, user characteris

tics and user interface design. In H.-J. Bullinger & J.

Ziegler (Eds.), Proceedings of

the 8th International Conference on Human–Computer

Interaction. Human–computer

interaction: Ergonomics and user interfaces (Vol. 1, pp.

803–807). Mahwah, NJ: Law

rence Erlbaum Associates.

Röse, K. (2002). Methodik zur Gestaltung interkultureller

Mensch-Maschine-Systeme in der

Produktionstechnik (Method for the design of intercultural

human-machine sys

tems in the area of production automation).

Fortschritt-Berichte pak Band 5,

Kaiserlautern, Verlag Universität Kaiserslautern.

Russo, P., & Boor, S. (1993). How fluent is your interface?

Design for international

users. In Proceedings of INTERCHI ’93 (pp. 342–347).

Amsterdam:

Shih, H. M., & Goonetilleke, R. S. (1998). Effectiveness of

menu orientation in Chi

nese. Human Factors, 40(4), 569–576.

Stewart, E., & Bennett, M. J. (1991). American cultural
patterns: A cross-cultural perspec

tive. Yarmouth, ME: International Press.

VDMA (Verband Deutscher Maschinenund Anlagenbau). (2002).

Maschinenbau in

Zahl und Bild. Frankfurt: Maschinenbau Verlag GmbH.

Zühlke, D., Romberg, M., & Röse, K. (1998). Global demands

of non-European mar

kets for the design of user-interfaces. In Proceedings of

the 7th IFAC/IFIP/IFORS/

IEA Symposium. Analysis, Design and Evaluation of

Man-Machine Systems (pp.

000–000). Kyoto, Japan: Chapter

Travel Planning on the Web:

A Cross-Cultural Case Study

Helmut Degen

User Experience Consultant, D usseldorf, Germany

Kem-Laurin Lubin

Siemens Corporate Research, Princeton, NJ

Sonja Pedell

University of Melbourne, Victoria, Australia

Ji Zheng

Technische Universität Ilmenau, Ilmenau, Germany

INTRODUCTION

The Role of Economy of Cross-Cultural User Interface Design

Siemens is a global player employing more than 417,000

people in 192

countries. To be economically viable, it is a condition to

offer products and

services that meet the needs of people and potential

clients within the

global target markets. The quality of a product is

perceived, to a consider

able extent, by its user interface, indicating that the

user interface of a prod

uct plays a key role in a successful commercialization.

Recent studies show that people in different cultures have

different re

quirements and needs regarding the design of user

interfaces (Evers,

Kukulska-Hulme, & Jones, 1999; Honold, 2000; Prabhu &

Harel, 1999). 313

From an economic perspective, there are two conceptual

extremes in in

ternational marketing strategies. One is

internationalization, 1 by which

every country and culture receives the identical user

interface. The other

extreme is localization, by which every culture receives a

culture-specific

user interface.

These theoretical extremes have a direct impact on the cost

and volume

of sales. In the case of internationalization, the

development costs are low.

The volume of sales might be low because the user interface

very often does

not have a broad user acceptance. In the case of

localization, the develop

ment costs are high and the corresponding volume of sales

is likewise
higher because the user interface has a broader user
acceptance within its

specific culture. One point of view is to attain 100%

internationalization to

keep product costs low; the other point of view is to

maximize the revenue

by 100% localization. To be economically successful, an

appropriate bal

ance of internationalization and localization is necessary.

However, the challenge is to determine how much adaptation

is neces

sary to meet the needs of the target market while

maintaining an optimal

balance between costs and revenue. To answer this question,

we conducted

the following case study.

Objective and Scope of the Case Study

In our case study, there are two questions in the context

of cross-cultural

“User Interface Design”: First, how great are the cultural

differences and

what kind of localization is needed for user interfaces in

different cultures?

Second, where, within the entire design process, should

cross-cultural as

pects be taken into consideration? To answer these

questions, the underly

ing objective of our case study is to identify differences

and similarities of

user interfaces and how the environment of the cultural

regions repre

sented in this study impacts these differences and

similarities. We propose a

set of design factors to isolate these differences and

similarities systemati

cally. The target cultures selected are those of the United

States, Germany,

and China, more specifically New Jersey, Munich, and

Beijing, respectively.

These three locations were chosen because Siemens Corporate

Technology

“user interface design” departments are located in these

three countries,

making concurrent studies and communication easier to

manage.

In selecting what market domain would best suit our study,

it became

apparent that we had to select an area that fulfilled the

following require

ments: It should (a) play a major role in each culture, (b)

be deeply an

chored in the local culture, and (c) be easy to find

participants for

1 Internationalization is defined here as the production of

only one product version for the

whole global market. Therefore, every country and culture

receives the identical user interface.

conducting the case study. Therefore, we chose the market

domain of

travel—an area that is entrenched in most cultures. More

specifically, we

chose the task of “travel planning” in a web-based travel

portal for private

consumers. The idea is that in contrast to job-relevant

topics we have a
broader communication basis that prevents misunderstandings
with the

participants as well as among the three different research

parties in

United States, Germany, and China.

Another important criterion was to select the appropriate

target user

group. The target user group selected for this study is

“double income, no

kids” (DINKs). This target user group is easily

identifiable across cultures. This

group is characterized as having a lot of freedom with

respect to time and

money. This means that the focus is on the behavior of the

users and user inter

face itself and not on the process or methods that lead to

the interface. There

fore, these commonalities enable a more standardized and

equal comparison.

In this paper, we often refer to Americans, Germans, and

Chinese. Our

try, but only the participants in this case study, as

representatives of each

country and culture.

Cultural-Dependent Differences and the Design Process

The first question investigated was the differences among

the three user in

terfaces for the respective countries. In the event that

there were differ

ences, the task was to classify and compare the results of

the different

cultures. To guide the comparison process, we use design

factors (content,
function, layout, linkage, wording, and media) that allows
us to assess the

specific elements of interface designs (Degen & Pedell,

2003). However, in

this cross-cultural study, we focus primarily on four of

six design factors:

content, function, layout, and linkage; we do not

explicitly deal with word

ing and media. The reason for omitting the design factor,
“wording,” is that

we had to complete a comparable linguistic analysis of

three languages:

German, United States English, and Chinese, in terms of

colloquialisms

and expressions used in the graphical interface. We did not

have linguistic

experts available at this time for this specific analysis.

The design factor

“media” is discussed implicitly as part of the design

factor “layout.” This was

in part due to the fact that apart from different visual

media such as text and

graphical elements visualized in the single web pages,

there were no other

media such as sound and animations.

1. Content: Which kind of content do the users expect? How

should the contents be clustered?

2. Function: Which kinds of functions do the users expect?

To which content clusters are the functions assigned?

3. Media: In which media form (e.g., text, graphics, sound)

should content and/or function be presented?

4. Wording: In which (textual, graphical, acoustical)

expression form should the content and/or function be
denoted?

5. Layout: How should the single user interface elements be

orga nized on a single web page? How should they be
designed visually?

6. Linkage: Which is the expected sequence of single web

pages?

Where in the design process do the differences appear? The

general

question is where in the design process cross cultural

design should be lo

cated. Is it only at the end of the design process

(translating wording) or is it

rather at the beginning? It is possible to locate the

differences of cross cul

tural design in the design process with the introduced

design factors. The

design factors do not identify the type of difference but

also where the infor

mation appears within the design process. We can namely

allocate each de

sign factor to a respective phase in the design process.

For this purpose, we

use the design process model based on the ISO 13407, a

standardized Hu

man-Centered Design Process for Interactive.

System. The following list shows which phase of this design

process re

fers to which design factor.

• Use Context Analysis: Content, function

• Requirements Analysis: Wording, Layout, Media, Linkage

• User Interface Design: Layout

• Evaluation
Usually within the use context analysis, user interface
designer obtains

information about the required contents and functions of

applications. The

requirements analysis uncovers expected wording, layout,

media, and

linkage information. During the user interface design, the

user interfaces

are realized in terms of interaction and visual design.

During an evaluation

phase, usability issues are identified. To answer the

question of where cross

cultural differences appear within the design process we

also can look at the

relationship between design factors and phases of the

design process the

other way around. If we identify differences regarding the

design factors,

e.g., content, this could already be identified during the

user context

analysis. Another example: If we identified differences

regarding the layout

it can already be gathered during the requirements analysis.

APPLIED METHODS

Process Overview

The entire design process, until the final design of user

interfaces was

closely observed to investigate the drawn-up questions. We

focused on a use

context analysis, requirements gathering, and user

interface design accord

ing to ISO 13407.

The use context analysis is seen as indispensable for
obtaining knowl

edge of the background conditions of each culture. These

conditions are as

sumed to have an influence on the interface design. During

the require

ments gathering phase, the interface qualities are created.

At the same

time, the results of this step bridge the gap between the
use context analysis

and the next phase: the user interface design.

The study was very carefully organized such that each of

the three coun

tries used identical methods and followed the same

procedure (including

participants, target user group, length, used material,

etc.). Therefore, we

applied methods from a design process developed for

web-based e-busi

ness applications, called JIET (Just in E-Time; Degen &

Pedell, 2003) in

all three countries.

Use Context Analysis

Objective

The use context analysis was the starting point. The goal
of this step was

to uncover the typical travel behavior of the target user

group and the role

of the Internet within the travel planning process.

Questionnaire

For gathering information, a questionnaire concerning

travel behav

ior and some demographic data of the participants was used.

This ques

tionnaire allows for the gathering of day-to-day cultural

idiosyncracies

that would otherwise be undiscovered. Factors that might

influence

travel behavior (e.g., income and leisure time) were also

investigated.

The exact content and results of the questionnaire are

shown in the re

sult section of this chapter.

Requirements Gathering

Objective

The objective of the requirements gathering is to identify

use scenarios

and user interface requirements.

Focus Group

For collecting the requirements, focus groups were

conducted. Each of

the focus groups in the different countries had four

participants who were

representatives of the target user group, DINKs. Within

each focus group,

the following method was used.

Paper Prototyping Method NOGAP

The objective of the requirements gathering is the

detection of use scenar

ios and user specific requirements for the user interfaces.

The require

ments-gathering step bridges the gap between the analysis

and the user

interface design (Wood, 1998). To this extent, it is

essential, for efficiency

reasons, to arrive as close as possible to the desired

design of the user inter

faces. Therefore, at this point, we used a specially

developed paper prototype

procedure named NOGAP (Another Method for Gathering Design

Require

ments in E-business User Interface Design Projects; Degen &

Pedell, 2003).

At the core of the procedure are the six design factors

mentioned earlier in

this chapter. This categorization is used on a form for

requirements gather

ing (Fig. 11.1) and a form on which users outline their

ideas (Fig. 11.2).

During this entire data collection step, the user only

works with the

sketch form. The user develops and sketches ideas and

wishes and is guided

by the user interface designer. The result is a direct

image of the user’s men

tal model (Pedell, 1998). Even users with little design

experience begin to

sketch their ideas after a short warm-up phase

(approximately 5 minutes).

The user interface designer records the ideas on the

requirements form,

which he or she also uses as guideline for later questions

and clarifications.

It is recommended to carry out this type of interview

procedure with two
user interface designers, where one designer asks the
questions; the other

records the requirements. In this situation, a transparency

of recording the

results is especially important. The simplest and most

efficient way is to re

cord directly with a laptop computer. During the entire

time, all partici

pants can view the records shown via projection. This

procedure increases

the confidence between the persons involved, and at the

same time it pre

vents incorrect recordings. The results of the requirements

gathering are

used as a basis for the design of the user interface. 319

FIG. 11.2. Sketch form.

FIG. 11.1. Requirements form.

For our study, we documented the click-through parts for

the scenarios.

Later on, we compared the results of two groups of one

country to identify the

major design ideas and elements for one common country

design.

User Interface Design

Based on these ideas and elements, the user interface

designers created a

web site concept including the requested content, layout,

and click

through. This concept was delivered to the visual designer

in the form of a

grid design (sketched with Microsoft Visio). In a design

workshop, we ex
plained the concept and the background information to the
visual de

signer. He or she then had the task and freedom to shape

the interface

concept into an attractive design, with two restrictions:

considering the

given results and the respective target culture.

RESULTS

Use Context Analysis

For interpreting differences in cross-cultural design, it

is very important to

know about the users’ circumstances. Next, some general

aspects of the coun

tries and participants’ lives are discussed. These aspects

were chosen for their

probable influence on travel behavior and differences in

the three countries.

Currency and Buying Power

In particular, financial background influences both

everyday life and

travel behavior.

The buying power of the currencies differs dramatically

among the three

countries (see Table 11.1). The dollar has eight times more
value than the

Chinese RMB; this means that the Americans get the value of
eight dollars

for every one dollar when traveling within China. On the

other hand, the

Chinese have to pay eight times more in the United States

than in their own

country. The value of the Euro compared to the RMB is

similar. The value

difference between the Euro and the U.S. Dollar is not as

dramatic, but still

1.4 in favor of the Americans. 2

Target User Group: DINKs

As mentioned before, we chose the target user group DINKs

(Double In

come, No Kids). All three countries share this target user

group. Addi

2 Based on exchange rates at the time of study.

tionally, this group was an appropriate selection because

of their unlimited

financial possibilities and unrestricted ways of planning

their travel.

Table 11.2 shows some fundamental similarities across

cultures, but re

lated to buying value, the Americans have more freedom in

spending their

money than do the Germans. The Chinese DINKs earn a lot

more money

than the average non-DINK Chinese class, but less than the
Americans and

Germans. The age and the status are similar across all
countries. Concern

ing their linguistic ability, the American participants

usually do not speak a

foreign language. While not of primary interest for the

purpose of this

study, the languages factor may influence their travel

behavior as well.

Participants

The participants are representative of the target user

group, DINKs,

which allows for a more reliable and comparable result of

this cultural

study. Table 11.3 shows the most important demographic data

about the

participants. Again, we are able to ascertain some

fundamental similarities

in the profiled target group which serves for

comparability. TABLE 11.1 Currency and Buying Power Germany
United States China

Currency Euro and Cent Dollar and Cent RMB and Fen 1 Euro =
100 Cents 1 US Dollar = 100 Cents 1 CNY = 100 Fen

Conversion

Rate 1 Euro = 0.9 US Dollar 1 US Dollar = 1.1 Euro 1 CNY =

0.13 Euro* 1 US Dollar = 4 CNY* 1 US Dollar = 8 CNY* 1 CNY
= 0.12 US Dollar*

Buying

Power Euro in United States: 70% US Dollar in Euroland: 1

140% RMB in Euroland: 13%* Euro in China: 400%* US Dollar
in China: 800%* RMB in US: 12%*

Sources Statistisches Bundesamt 2002 Statistisches

Bundesamt 2002 State Administration of Foreign Exchange
*Estimated *Estimated *Estimated

1 Euroland is the region in Europe with the Euro as the

valid currency. The Euroland con

sists of the following countries: Austria, Belgium,

Finland, France, Germany, Greece, Ireland,

Italy, Luxembourg, The Netherlands, Portugal, and Spain.

Significant differences between women and men within the

target user

group DINKs could not be found. Of interest is that the

Chinese partici

pants spent a higher percentile income than either the

Germans or the U.S.
participants.

Frequency and Length of Private Travels

One study shows that one of the basic factors in choosing a

travel destina

tion is average annual leave. Available vacation time

varies significantly

among the three countries (see Table 11.4). Germany has the
longest an TABLE 11.3 Participants Germany United States
China

Number of persons 4 persons 4 persons 4 persons

Target User Group DINKs DINKs DINKs

Gender 3 male, 1 female 1 male, 3 female 2 male, 2 female

Range of Age 30–33 years old 28–34 years old 26–29 years old

Expenses for private

traveling per year

(only DINKs) for 2

persons 6,000 Euro (8% of yearly income) 8,000 USD (5% of

yearly income) 18,000 RMB (15% of yearly income)

Source Participants Participants Participants TABLE 11.2

Demographics of Target User Germany United States China

Age 25–35 28–36 25–35

Status Professionals Professionals Professionals

Income

(2 persons) 80,000 Euro 150,000 USD 120,000 RMB

Education High educated (often university degree) High

educated (often university degree) High educated (often
university degree) Multilingual (English and other
languages) Multilingual (Eng lish and other lan guages)

Source Participants Participants Participants

nual leave. Only the Germans have the possibility of
traveling for 3 to 4

weeks in a row. This might be a better incentive to travel

far away. The

American and Chinese travel time, by comparison, is much

shorter, which

may account for a “close-to-home” attitude in their travel.

Travel Behavior

Travel behavior differs considerably among the three

countries (see Ta

ble 11.5). Whereas the Germans prefer to travel

individually, the Americans

and Chinese prefer to travel in groups. The Germans seem to

have a far

greater interest in the weather than do the participants

from the other two

countries. The participants from China do not have a

concrete preferred

travel destination. Also of interest was the primary

criteria that made a des

tination desirable. While German and American participants

cited

“weather,” the Chinese chose nature and landscape. Table

11.5 shows the

responses by frequency.

Preferred Communication Channels

As further input for later design, we paid special

attention to the mode of

communication for each travel planning step. Table 11.6

shows the usual task

flow for travel planning for each country. The first step
is the selection of a
travel destination, as well as selecting arrangements
(accommodation and TABLE 11.4 Frequency and Length of
Private Travels Germany United States China

Average annual

leave (1 person) 6 weeks 2 weeks 3 weeks at least

Public holidays Depending on the region 11 days 10 days

Average: 11 days

Frequency of pri

vate traveling 2–7 times a year (mostly 3 times a year)* 2

per year* 1–3 times a year (mostly 2 times a year)*

Length of private

traveling 3 days–4 weeks (mostly 2 weeks)* 1–2 weeks* 1

week average

Source *Participants *Participants *Participants

means of transport). Following are the steps of

reservations, booking and con

firmation. Sometimes travels arrangements are changed or

rebooked, which

requires user access to flight information. In case of

waiting times, participants

want to have information about the current status of

delivery (tracking/trac

ing). For each step, the participants communicated to us

their preferred com

munication channels. The only exact commonality is the

participants’ Internet

and telephone preference for the final stage

(tracking/tracing).

However, in the United States and in China, the Internet is

one of the pre

ferred communication channels in every step. German

participants do not use
the Internet, the Americans prefer the telephone, and the
Chinese partici

pants prefer the Internet. In Germany, the main reason for

not using the TABLE 11.5 Travel Behavior Germany United
States China

Manner of travel Individual or 2 per sons With others As a

group

Destination

Criteria Weather (4) Weather (4) Nature landscape (3)

Exotic (2) Culture (4) Culture (2) Sightseeing (2) Food and
Wine (3) Nature (1) Culture (2) Beach (2) Sports: Climbing
(1) Nature (2) Recreation (1) Service (1) Beach (1) Food
and Wine (1) Languages: English, French (1) Sports,
activity (1) Recreation (1)

Preferred

Destinations Europe, America, Southeast Asia South America,

Caribbean, and Europe Main objective: recreation as a
change from the city. To see foreign countries

Booking by

Internet Yes (4) Yes (4) Yes (4)

Source Participants Participants Participants

Internet is security. There is a hesitation and aversion to

reveal their credit

card information. In China, this problem doesn’t exist to

that extent, because

credit cards are not very popular. In Shanghai, one of the

biggest cities in

China, the percentage of credit card holders was 3.8% at

the end of 2001

(Xinhua News Agency, 2001). A very popular form of payment

is “Pay on De

livery.” With “pay and deliver,” the Chinese user orders a

flight ticket via the

Internet or the telephone and gives the travel agency the

contact information.

The ticket will be delivered and paid by the user during

the handing over; this

is like the American version of collect on delivery (COD).

Besides this form of

payment, there are also other traditional payment methods

like cash, bank re

mitting, and invoice which replace the more Western forms

of online payment.

Summary

To summarize the use context information we can say that

the similari

ties within one country, among the participants, are

extremely high, but TABLE 11.6 Preferred Communication
Channels for Travel Planning Cycle

Single steps Germany United States China

Select Travel 1: In-Person 1: In-Person 1: Internet 2:

Internet 2: Internet 2: In-Person

Select Arrangement 1: In-Person 1: In-Person 1: Telephone

2: Internet 2: Internet 2: Internet

Reservation 1: In-Person 1: Internet 1: Internet 2:

Telephone 2: Telephone 2: In-Person

Booking 1: In-Person 1: Telephone 2: In-Person 2: Telephone

2: Internet 1: Internet

Confirmation 1: In-Person 1: Internet 1: Internet, 2:

Telephone 2: Telephone telephone

Change/Rebook 1: In-Person 1: Telephone 1: Telephone 2:

Telephone 2: Internet 2: Internet

Tracking/Tracing 1: Internet 1: Internet 1: Internet, 2:

Telephone 2: Telephone telephone

Source Participants Participants Participants

Note. Of specific interest was the German affinity for

person-to-person communication in
all steps but the last. 1 and 2 refer to the primary and
secondary preferred communication

channels.

326 TABLE 11.7 Requirements Based on Design Factors

Design Factor Germany United States China

Content Home Page • Information about countries and desti

nation (travel guide) • Filter Information: Activities,
weather, travel suggestions etc. • Special Offers •
Transportation and Hotel Information • Price, dates, and
quality information • Availability of offerings Required
Travel Information • Departure and arrival times • Hotel
address Home Page • Specials • Quick filters • Link to
Accommodations (hotel and car) • Useful travel resources •
Travel News • Customer Service information • Expert Advice
Required Travel Information • Departure and arrival times •
Hotel address Home Page • Categorized travel information
such as domestic/abroad, cul tural, geographic etc. •
Travel news • BBS/Clubs for in formation exchange • No
flight or hotel information, but these should be spec ified
in the travel package offers from travel agency • Select
target city and travel offers are the aims of the users
while using this web Required Travel Information •
Appointment information of the group

Function Booking system full text search engine Booking

system quick search of flight information Booking system
full text search engine

Layout • Left-hand side: Filter criteria • Middle: Map •

Right-hand side: Offerings • Left-hand side: Traveler tools
• Middle: Search field • Right-hand side: Specials Offers •
Left-hand side: Search/Information Categories • Middle:
News, Highlights, Offers • Right-hand side: special topics,
add-value service

Linkage 1. Using filter crite ria and region (or vice

versa): Alterna tive: Search Engine 2. Travel Information
3. Travel Offerings 4. Travel Booking 1. Using the search
en gine; alternative: se lecting special offers 2. Travel
Offerings 3. Travel Booking 1. Selection domestic or abroad
destina tions; alternative: Search Engine 2. Selecting
regions or special offerings 3. Travel Information 4.
Travel Offerings 5. Travel Booking

some differences among the three countries are evident.

Although the

Americans have most financial freedom for their travel, the

Germans have

the largest blocks of free time to spend abroad. The

Chinese participants

spend 15% of their yearly income on traveling, which is the

highest percent

age spent of those studied. For the Chinese, traveling

seems to equate high

quality of life. The Germans, by choice, have the lowest

Internet affinity;

they prefer the direct contact with salespeople. When

planning a trip, the

Americans prefer contacting a salesperson, but for

reserving, booking,

tracking/tracing, they prefer the telephone or the Internet

as communica

tion channels. The Chinese prefer the Internet for almost

every step. Only

for rebooking and for travel arrangements do they prefer

the telephone.

Requirements Gathering

The requirements were gathered in all three countries

equally with the

method described earlier in this chapter. Each focus group

in all three

countries was divided in two groups of two persons. These

pairs came up

with very similar ideas that were merged by the user

interface designers in

one layout in the next step. Additionally, two extra focus

groups took place

in the United States (within a tutorial of the CHI 2002

conference), and in

China (within a demonstration workshop for this method).

The same per

sons who moderated the German focus group for this study
also moderated

these two extra groups. The collected requirements cover

the results pre

sented here and support them. That means, as well, that the
differences

found among the countries derive from cultural differences

rather than

from differences in using the method.

Germany Results

Very often, the Germans do not have a fixed travel

destination in mind.

They want to find a travel destination based on personal

criteria (e.g.,

weather, culture, exotic). For the German participants,

searching for the

travel destination is the start and preparation for the

travel itself. They en

joy “browsing” around in the world. The web site is a means

to collect ideas,

find a travel destination, and find precise information

(e.g., about availabil

ity, prices, and quality).

The Germans wish to get a quick overview of the destination

possibili

ties in general (the whole world). Quick access to the

regions of the world

(globe) in the center of the screen proved to be the ideal

solution. Ac
cording to their travel expectations, they want to have
access to filter cri

teria (e.g., weather, activities, social circumstances) on

the left-hand side

of the screen. They expect concrete offerings for the

selected region and

filter criteria on the right-hand side, as shown on the

scribble (Fig. 11.3).

Furthermore, they like to have direct access to proposed

travel (e.g., last

minute offers, all-inclusive packages) on the right-hand

side as well. In

addition, they need clear access to contact information

(see Fig. 11.3:

lower right corner).

After selecting a region (in this case, Munich), the

participants expect

to see a description very saliently placed on the page (see

Fig. 11.4). On

the left-hand side of the page remains the filter area and
on the

right-hand side the access for concrete offerings (hotel,

transportation,

last minute offers).

To classify the differences among the three user interfaces

in more detail,

their characteristics were allocated to the design factors

described earlier.

We summarize the results of the German requirements using

the follow

ing four design factors of content, layout, linkage, and

function, respectively.

Content. The first screen shows the globe and the second
screen shows

the destination, in this case, Munich. Filters represent

criteria by which the

German participants choose their travel destination, such

as weather condi

tions, medical situation in the country, and cultural life.

The filters stay on

the home page and filter further within the chosen area in
more detailed

destination levels (Munich, in this case).

FIG. 11.3. Home page (Germany).

Layout. Content and visual information go very close

together. It is

very important that the layout gives a clear overview and

supports the user’s

tasks. There is a direction from left to right that allows

the user to filter the

travel information: filter criteria (left column), region

(middle column),

and offerings (right column). (See Fig. 11.3)

Linkage. As sketched in the interaction architecture in

Table 11.8, the

users mainly prefer three different ways of finding their

travel information:

(a) using the filter functions and then the regions, and
(b) using the regions

first and then the filter functions, and (c) using the
search engine. After get

ting the travel information, the users select relevant

travel offerings and

might subsequently book the travel.

For example, in navigation path 1, the user starts on the

home page,

then applies travel filter (e.g., activity travels) (area 2

on Fig. 11.3) and se

lects the travel destination by a region (area 3 on Fig.

11.3). Afterward, the

user receives travel information and can get travel

offerings about the se

lected region. Finally, travel arrangements can be booked.

FIG. 11.4. Destination (Germany).

Function. After finding the right region for travel, the

participants

want to have access to reservation tools if they choose to

use them. This is

a step toward getting a concrete and real idea of their

complete travel

plan and to see the availabilities. They still prefer to

leave the booking it

self to the agency.

U.S. Results

In contrast to the German participants, the American

participants have a

very concrete idea of their travel destination. In part,

this is due to the short

vacation times, compared to their German counterparts who

have much

longer periods for travel. It is safe to say that most of

the information gath

ering is very much in the background and in some cases

nonexistent. Our

studies show that American participants want to have a very

effective and ef

ficient tool to book their travel that takes into

consideration time con

straints. Therefore, the scribble of the home page shows

different boxes,

which have all the intention to provide concrete

information for somebody

who knows exactly what he or she needs for a certain

travel. A clearly placed

“Quick Search,” in the center of the home page, and the

main categories,

such as air, hotel, and so on, placed in the top tabs show
this intention

clearly. For users who haven’t chosen their destination,

the web site pro

vides special offers (today deals) and “destination of the

week,” which both

include ready travel packages for the intended travel.

We summarize the results of the U.S. requirements using the

same de

sign factors. TABLE 11.8 Interaction Architecture for

Germany

Web page Navigation Path 1 Navigation Path 2 Navigation

Path 3

1 Home Page Home Page Home Page

2 Travel Filter Region Search Results

3 Region Travel Filter Travel Filter

4 Travel Information Travel Information Travel Information

5 Travel Offerings Travel Offerings Travel Offerings

6 Travel Booking Travel Booking Travel Booking

Note. Reading example for navigation path 1: The user

starts on the home page (1) and

applies travel filter (e.g., activity travels) (2) and

selects the travel destination by a region (3).

Afterward, the user gets travel information (4) about the

selected region and can get travel of

ferings (5). The user can book these (6).

Content. The tab cards remain in view throughout the whole

Web site

navigation and allow for quick change from one category to

the next. All aim

directly at the easy booking. At the second screen, when

navigating from the

main screen (Fig. 11.5), a concrete flight can be booked as

shown in Fig. 11.6.

Layout. A layout is chosen, which corresponds with this

booking-oriented

content. The content is offered in portletlike packages

that contain different

kinds of information. The quick search is the central

element and takes most of

the central space (Fig. 11.5).

Linkage. As shown in the interaction in Table 11.9, the

users prefer two

different ways to get to the travel offerings. One path

uses the search engine

and the other path is through the search for special

offerings. Following, travel

offerings are displayed and the users can book their travel
arrangements.

For example, for navigation path 1: The user starts on the

home page

and proceeds to using the search engine. After getting the

search results,

travel offerings can be selected and are displayed

according to search
filters. Then selected offerings can be booked.

Function. The functions are all related to reservation,

booking, and

getting detailed information about concrete travel

destination. These con

FIG. 11.5. Home page (United States).

crete functions, in contrast to information gathering are

on a very top level

of the web site. This means that users encounter the

concrete and primary

function on arriving at the main site. Additionally, the

quick search pro

vides the possibility to choose the means of transport.

China Results

The Chinese participants seem to be easily enticed by

pictures that show

different travel destinations. They pursue a high quality

of life and their dis

posable income essentially affords for inclusion of

traveling. TABLE 11.9 Interaction Architecture for the
United States

Web page Navigation Path 1 Navigation Path 2

1 Home Page Home Page

2 Search Results Special Offerings

3 Travel Offerings Travel Offerings

4 Travel Booking Travel Booking

FIG. 11.6. Destination (United States).

Even though traveling overseas is very attractive to them,

one main focus

is on their own country. The reasons for this might be the

short holiday

time, the size of their country (China provides a large

variety of landscapes

and different cultures), and the organizational effort

(e.g., acquiring a visa)

that is needed for traveling abroad.

Content. The first scribbles (Fig. 11.7) shows the

following contents: a

search engine, the most traveled tourism cities and

cultural places, special

customs, recommendations, tourism-relevant news, forum for

information

and experience exchange, and contact. The second scribble,

destination

China (Fig. 11.8), still includes the search engine. It

also provides catego

rized information about the destination, such as geography,

history, cul

ture, sightseeing, customs, and many pictures and links to

travel offers.

Layout. In the home page China drawing (Fig. 11.7), the

information

is clearly categorized. Similar information is combined,

such as added value

service (SMS, travel clubs). In Fig. 11.7, the top

navigation remains stable

on the page. Photos and information about customs, as well

as cultural de

FIG. 11.7. Home page (China).

tails, appear in the middle of the page. Categorized

information and search

engine appear on both sides.

Linkage. On the first scribble (Fig. 11.7), users want to
select whether

they plan to travel domestically or abroad. Following,

lists or categories of

destinations are presented. The most popular tourism

countries, regions,

and cities are selected on the top navigation. Besides,

users can also select

the “Top10” to view the most traveled or recommended

places. In addition,

a powerful search engine, with advanced search categories

like history, cul

ture, geography, and so on, enables users to find

information as quickly as

possible. The idea of the second scribble (Fig. 11.8) is to

select categories

such as geography, culture, and sightseeing for detailed

explanation. The

click on “More Photos” leads to the photo gallery. If the

users are interested

in the destination, they can immediately select respective

travel offers. The

top navigation and search engine on the right side help

users to find other

destinations.

The users prefer four options (see Table 11.10) of

completing their task.

The first and the second options guide the users through
the domestic des

tinations, the third and the fourth options through

destinations abroad. In

the fifth navigation path, the users use the search engine.
In the case of nav
igation paths 1 and 3, the users prefer to select offerings
according to re

gions and in navigation paths 2 and 4, according to special

offerings.

Afterward, travel information is displayed; the user can

then see the travel

offerings and book selected offerings.

As depicted in Table 11.10, the user starts on the home

page and selects

domestic travel destinations (area 2 on Table 11.10).

Afterward, he or she

has the opportunity to select regions or special offerings

(area 3 on Table

11.10). The travel information for region or specials

offerings can be se

lected and are displayed (area 4 on Table 11.10). The user

can see available

offerings (area 5 on Fig. 11.10) and can book the travel

(6).

Functions. The Chinese appreciate a powerful and

categorized search

engine. The following functions, which are not realized in

this case study,

are of special interest: Chinese participants would like to

have a voice inter

face, for instance, they like a friendly voice to welcome

them. Additionally,

they are very interested in 3D virtual tour.

Requirements Gathering Summary

Based on the results of the requirements gathering, the

main differences

between the three countries relate to content and linkage.

While the Germans
and the Chinese require comprehensive travel information,
the Americans do 335

FIG. 11.8. Destination (China). TABLE 11.10 Interaction

Architecture for China

Web page Navigation Path 1 Navigation Path 2 Navigation

Path 3 Navigation Path 4 Navigation Path 5

1 Home Page Home Page Home Page Home Page Home Page

2 Domestic Domestic Abroad Abroad Search Results

3 Regions Special Offerings Regions Special Offerings

4 Travel Information Travel Information Travel Information

Travel Information Travel Information

5 Travel Offerings Travel Offerings Travel Offerings Travel

Offerings Travel Offerings

6 Travel Booking Travel Booking Travel Booking Travel

Booking Travel Booking

not. Also, the linkage concept is different among the three

countries: Before

Germans select travel offerings, they expect comprehensive

travel informa

tion. The Americans prefer a straightforward strategy to

travel offerings with

out travel information. The Chinese first select between a

domestic or foreign

destination; afterward, they follow a similar strategy as

the Germans. All-im

portant results of the requirements gathering are shown in

Table 11.7.

User Interfaces

On the basis of the scribbles, which include the

requirements of the partici

pants of the respective country for a travel Web site, a

visual design was cre
ated. A visual designer in each respective country did the
final visual design.

The procedure is described earlier in this chapter. The

designs show the re

quired contents and functions and basic layout.

INTERPRETATION OF CROSS-COUNTRY RESULTS

Similarities

Use Behavior: All cultures use the Web to plan and arrange
travel, and for

tracking/tracing their orders. Reserving, booking,

confirming, and chang

ing/rebooking are steps that participants of all three

cultures take, but the

Internet is not equally used for it.

Content: Travel offerings are seen as products, which are

described and

available on the Internet, even if the preferred

destinations are different.

All web sites provide special offers that range from

complete travel pack

ages to self-organized trips.

Functions: On every web site, a search engine and a log-on

are available.

Layout: The chosen basic color on all designed travel web

sites is blue,

and tab cards are used to organize the content.

Linkage: Access to different offerings (hotel, car,

flights, etc.) and desti

nations is available. The goal is either to get a detailed

overview of a pos

sible trip or to book a complete travel package.

Differences

Travel Behavior

Germany. The German participants prefer solo traveling to

foreign

countries, even to exotic places. The duration of one trip

per year is longer 337

FIG. 11.10. Visual design: Regions (Germany).

FIG. 11.9. Visual design: Home page (Germany).

Germany

338

FIG. 11.12. Visual design: Flight information (United

States).

FIG. 11.11. Visual design: Home page (United States).

United States 339

FIG. 11.14. Visual design: Destination information (China).

FIG. 11.13. Visual design: Home page (China).

China

than 2 weeks—rather 3 to 4 weeks. The Germans take their

time to prepare

and select their travel destination. Surfing on the Web for

the “perfect spot”

is fun for them. It is supposed that the fun and desire for
very detailed plan

ning beforehand goes together with the fact that the

Germans travel on

their own. Like the U.S. and Chinese participants, the

Germans use the

Internet for finding a travel destination, arranging the

travel, and track
ing/tracing, but not for reserving, booking, and paying.

United States. Our study shows that more often than not,
American

travelers already know their travel destination when they

go on the travel

web site. Their choice of destination is often based on

advertising and

posted specials. They like traveling in pairs or in small

groups with people

they know. Because of the restricted travel time, they more

likely travel near

the U.S. continent. The Internet is used for the entire

planning process ex

cept for choice of destination.

China. The Chinese travel in groups. Individual travel is

rather un

usual. Like the Americans, the Chinese enjoy domestic

travel. On the one

hand, it is easier (organizational effort to get visa and

restricted vacation

time of two weeks) and on the other hand, the country

itself has a large vari

ety of different landscapes and cultures. Another reason is

the lower income

and buying power and the linguistic inconvenience. The

Internet is used for

the entire planning process from planning to booking. Like

the Germans,

they don’t have a concrete focus on their travel

destination, but they con

sider recommendations, web ads and other travelers’

opinions to be useful

whereas Germans use personal filter criteria.

Content

Germany. The German web site provides lots of different

travel and

country information, as well as flight and accommodation

information, and

supports the user to get informed on a broad scale.

Different categories like

activities, climate, weather, and health lead to very

detailed information

that fit respective travel interests.

United States. There are no content pages to surf, but

access to con

crete flights and hotels, services on site like rental

cars, and special travel

packages are available. Other travel options like cruise

and vacation pack

ages are quite salient, but the amount of content is not as

high as seen in the

German results.

China. A lot of information relevant to the target

destination can be

found. Additionally, information not found on the German or

U.S. travel web

sites can be found here, for example, travel news, links to

other travel agencies,

and the Top 10 most traveled destinations. BBS 3 and

information exchange of

web users are also used to find the travel destination.

That means that they

don’t have a concrete destination in mind, but make their

travel choice rather

by recommendations and web ads than cruising around and

using personal

filter criteria. Flight and hotel information are not the

focus of users because,

as already mentioned, these are arranged by the travel

agency.

Layout

Germany. The overall layout triggers a desire to travel the

whole

world. Especially the home page provides a lot of

function-free space to

support unrestrained room for the users’ imagination where

to spend the

next holiday. The single function filters, content

information, and offers

are clearly distinguished in separate blocks. They always

stay in the same

place while users click through the site. To support

different ways of choos

ing a travel destination, different accesses (filters of

different granularity,

the world map, pull-down menus) are available.

United States. The U.S. home page is compact and gives

clear direc

tions of where to enter relevant information. Graphics and

color are used

to facilitate quick entry and to show relationships and

grouping of infor

mation in a structured way. The contact information is

clearly placed.

China. The Chinese web site provides very clear categories

of target

city information in rows. Regional-specific information,

such as cultural
events, are highlighted. Photographs of the target
information are also a

central feature of the Chinese web site and take up a lot

of room.

Linkage

Germany. Our study shows that Germans travel enough that

they are

constantly looking for new and exciting travel

destinations. Therefore, they

browse through the web site to get travel inspiration. The

click-through is as

follows: Regional selection or topic-oriented selection

before looking for a

suitable flight and sometimes finish by looking for the

right accommoda

tions. Often, Germans merely choose the destination at the

web site and

book it in a travel agency.

3 BBS is the abbreviation for Bulletin Board System. Web

users can put texts, pictures, and

animations on the system. These are normally shared for all

registered users. In this case study,

it is a place of information exchange, for example, users

can tell about their travel experiences

or show nice photos.

United States. They have a clear idea where they want to

go. There

fore, they type in the name of the destination in a Quick

Search, which leads

directly to a booking screen.

The procedure is similar to that in a classical Internet

shop: First, they
enter depart/destination search criteria; then, they
receive output (e.g.,

flight options) and the option to hold or buy and the

possibility to pay for

the selected option.

China. The intention of Chinese users, in visiting the web

site, is to

both browse and book. First they peruse regional selection

or topic-ori

ented selection; then they focus on abundant information

about the target

destination. They book for group travel to the destination

and trust that

the organization of accommodation and means of transport is

dealt with

by the provider.

SUMMARY AND CONCLUSIONS

Designing user interfaces for the global market raises the

question as to

whether a high level of localization based on cultural

needs is required. We

have chosen the topic “Travel Portal” and the target user
group “DINKs”

for our case study. In this case study, we analyzed

requirements and de

signed user interfaces for three cultures (Germany, the

United States, and

China) and showed where the differences appear. We used a

set of six de

sign factors to discover and to compare cultural

differences.

Based on the results of the cross-cultural design study,

summarized in

Table 11.7, we can identify major differences for the

design factors, con

tent, layout, and linkage and minor differences for the

design factor

function. Based on the close relationship between the

design factors and

the phases of the design process (see chapter 1) we propose

that

cross-cultural differences are located in the phases “use

context analy

sis,” “requirements analysis,” and “user interface design.”

Therefore,

the cross-cultural design should be considered at the

beginning of the

design process and not at the end, such as simply

translating the content

into another language. According to the extent of

differences, the over

all effort can be estimated and the level of localization

can be determined

in early stages.

According to the results, it is necessary to include user

input of different

cultures from the very beginning of the design process.

Cosmetic changes

such as translating the language or changing the colors,

instituted late in

the process of design or even after the fact, are

insufficient. Managers and

user interface designers can see the trend of cultural

differences in an early

stage of the design process (after applying the method

NOGAP, require

ments phase). The effort of a final user interface design

to evaluate differ

ences is not necessary, since differences are identified

early enough to war

rant varying user interfaces. Here the values of

localization are illustrated

with success. Rather than assuming that one user interface,

with final alter

ations, responds to the specific needs of all cultures,

this case study allows

for the inclusion of significant cultural factors that show

how users perceive

their tasks in the context of travel planning on the Web.

While most compa

nies do not consider the rudimentary cultural differences

that are needed

to ensure market success, an understanding of the locale

idiosyncrasies un

doubtedly proves useful in this study.

ACKNOWLEDGMENTS

We would like to thank Heinz Bergmeier, Nuo Tang, and

Jui-Lin Dai for the

visual design of the web pages. Many thanks to Pia Honold

for the nice and

constructive discussion. We thank Stefan Schoen for his

interest and providing

the required resources. Special thanks goes to Nuray Aykin

for her patience.

Degen, H., & Pedell, S. (2003). JIET design process for

e-business applications. In

D. Diaper & N. Stanton (Eds.), Handbook of task analysis

for human-computer interac
tion (pp. 193–220). Mahwah, NJ: Lawrence Erlbaum Associates.

Evers, V., Kukulska-Hulme, A., & Jones, A. (1999).

Cross-cultural understanding of

interface design: A cross-cultural analysis of icon

recognition. In E. del Galdo &

G. Prahbu (Eds.), International Workshop on

Internationalization of Products and Sys

tems: Vol. 1. Designing for global markets (pp. 173–182).

Rochester, NY:

Honold, P. (2000). Interkulturelles Usability Engineering.

Eine Untersuchung zu

kulturellen Einflüssen auf die Gestaltung und Nutzung

technischer Produkte.

Düsseldorf, Germany: VDI-Verlag.

ISO/IEC 13407. Human-Centered Design Processes for

Interactive Systems, ISO/IEC

13407: 1999.

Pedell, S. (1998). Die Bedeutung der Gestaltgesetze

Ähnlichkeit und Nähe für die

Gestaltung von Benutzeroberflächen. Unpublished master’s

thesis, Technical Uni

versity of Berlin, Germany.

Prabhu, G., & Harel, D. (1999). GUI design preference

validation for Japan and

China—A case for KANSEI engineering? In H.-J. Bullinger &

J. Ziegler (Eds.),

Proceedings 8th International Conference on Human-Computer

Interaction: Vol. 1. Hu

man-computer interaction: Ergonomics and user interfaces

(pp. 521–522). Mahwah,

NJ: Lawrence Erlbaum Associates.

Statistisches Bundesamt. (2002). Kaufkraft des Euro im
Ausland. In http://www.

destatis.de/basis/d/ausl/auslkkr1.htm, Statistisches
Bundesamt, 13th of May 2002.

Wood, L. E. (Ed.). (1998). User interface design. Bridging

the gap from user requirements to

Design. Boca Raton, FL: CRC Press.

Xinhua News Agency. (2001).

http://www.xinhua.org/fortune/2002-03/18/con

tent_321301.htm