WEBSITE QUALITY ELEMENTS AND ONLINE SHOPPER

BEHAVIOUR: ADAPTING THE UNIFIED THEORY OF

ACCEPTANCE AND USE OF TECHNOLOGY TO FASHION
RETAILERS’ WEBSITES

Luisa Cavallero

Dissertation submitted as partial requirement for the conferral of

Master in Marketing

Supervisor:

Prof. Doutor Sandra Maria Correia Loureiro, Prof. Auxiliar com agregação, ISCTE Business
School, Departamento de Marketing, Operações e Gestão Geral

September 2016
Acknowledgements

Working on this dissertation was a great experience and I would like to express my gratitude to
all of those who continuously supported me during this research.

First, this dissertation wouldn’t be possible without the help of Prof. Sandra Loureiro, my
supervisor, whom I thank for her availability, patience and assistance throughout this process.
Thanks to her vast experience and knowledge, as well as to her accessibility and openness, the
elaboration of this thesis contributed immeasurably to my knowledge and personal development.
Her advice always gave me the inspiration that enabled me to build my research ideas and to
continue with my dissertation. I really appreciated the opportunity to work under her direction.

Also, I would like to thank my family, especially my parents and my brothers, and my friends,
who have accompanied me in my academic journey and have given me the comfort and courage
to pursue my goals. Lastly, I would like to thank my Portuguese friends for helping with the
translation of my questionnaire and all those who answered it. Your help was extremely
appreciated.

i
Abstract

Companies in the fashion industry are increasingly establishing their presence online, in the form
of online platforms, such as websites. However, some brands are more successful than others in
selling fashion clothing online. In this perspective, the two main objectives of the dissertation are:
(i) to adapt and extend the UTAUT model in the context of online shopping for fashion clothing;
(ii) to compare the adapted model in two regions of Portugal and Italy. By testing the adapted
model in two regions of Portugal and Italy, the study gives insights into online consumer
behaviour when using fashion websites. Results are based on a survey of 312 Internet users who
are familiar with fashion websites. Findings reveal that perceived website quality is the strongest
predictor of online behaviour, having a positive impact on customer satisfaction, website trust
and word-of-mouth. Besides, the present study explores whether online shoppers are or are not
influenced by other customers’ reviews and finds evidence that online customer reviews do not
impact consumers’ perceptions of online shopping usefulness.

Key-words: Website quality, Unified theory of acceptance and use of technology

(UTAUT), Trust, Online customer reviews, Satisfaction, Word-of-mouth.

JEL: M31 Marketing

JEL: L81 Retail and Wholesale Trade; e-Commerce

ii
Resumo

As empresas na indústria da moda estão cada vez mais a estabelecer a sua presença online, sob a
forma de plataformas online, tais como websites. No entanto, algumas marcas são melhor
sucedidas do que outros na venda de roupas de moda. Nesta perspetiva, os objetivos principais da
dissertação são: adaptar e estender o modelo UTAUT ao contexto de compras online para a roupa
da moda; comparar o modelo adaptado em duas regiões de Portugal e de Itália. Ao testar o
modelo adaptado em duas regiões de Portugal e Itália, o estudo permite obter informação sobre o
comportamento do consumidor online ao usar sites de moda. Os resultados são baseados em uma
pesquisa com 312 usuários de Internet que estão familiarizados com websites de moda. Os
resultados mostram que a qualidade percebida do website é o mais forte determinante do
comportamento online, tendo um efeito positivo na satisfação do cliente, na confiança e no passa-
palavra. Além disso, o presente estudo investiga se os compradores online são ou não
influenciados pelos comentários de outros clientes e encontra evidência que os comentários de
clientes não têm efeito significativo na perceção dos consumidores online.

Palavras-chave: Qualidade do Website, Teoria Unificada de aceitação e uso de tecnologia

(UTAUT), Confiança, Comentários online dos clientes, Satisfação Passa-palavra.

JEL: M31 Marketing

JEL: L81 Retail and Wholesale Trade; e-Commerce

iii
Table of contents
Acknowledgements ................................................................................................................................... i
Abstract ................................................................................................................................................... ii
List of tables ............................................................................................................................................ vi
List of figures ......................................................................................................................................... vii
1. Introduction ......................................................................................................................................... 1
1.1 Theme of the dissertation .............................................................................................................. 1
1.2 Objectives of the dissertation ......................................................................................................... 2
1.3 Structure of the dissertation........................................................................................................... 3
2. Literature review.................................................................................................................................. 3
2.1 Fashion consumption ..................................................................................................................... 3
2.1.1 Consumer motivations for fashion consumption....................................................................... 4
2.1.2 Behavioural mechanisms of fashion consumption .................................................................... 5
2.1.3 Models of fashion consumption ............................................................................................. 6
2.2 Shopping and the consumer decision-making process .................................................................... 7
2.3 Online shopping ............................................................................................................................. 9
2.3.1 Advantages of online shopping ................................................................................................ 9
2.3.2 Disadvantages and perceived risks of online shopping ........................................................... 12
2.4 Website quality dimensions: technical quality and information quality ......................................... 14
2.5 Online customer reviews .............................................................................................................. 16
2.6 The role of performance expectancy ............................................................................................ 19
2.7 The moderating impact of previous online experience ................................................................. 21
2.8 Relationship outcomes of website quality .................................................................................... 22
2.8.1 Customer satisfaction, website trust & recommendation adoption ........................................ 23
2.8.2 Word-of-mouth behaviour ..................................................................................................... 25
3. Methodology ..................................................................................................................................... 27
3.1 Research context: Italy and Portugal ............................................................................................ 27
3.2 Data collection ............................................................................................................................. 29
3.3 Sample profile .............................................................................................................................. 32
4. Data treatment and Results ............................................................................................................... 33

iv
 4.1 Descriptive statistics ..................................................................................................................... 33
4.2 Regression Analysis results ........................................................................................................... 39
4.2.1 Determinants of Performance Expectancy.............................................................................. 40
4.2.2 The moderating impact of previous online experience on perceived website quality ............... 44
4.2.3 Relationship outcomes of website quality .............................................................................. 46
4.4 One-way ANOVA and non-parametric tests .................................................................................. 54
5. Conclusions ........................................................................................................................................ 56
5.1 Discussion of results ..................................................................................................................... 56
5.2 Managerial implications ............................................................................................................... 59
5.3 Limitations and future research .................................................................................................... 62
6. References ......................................................................................................................................... 63
7. Appendixes ........................................................................................................................................ 71
Appendix A: Sample profile & Descriptive statistics ............................................................................ 71
Appendix B: Regression analysis results.............................................................................................. 77
Appendix C: Regression analysis results for the Portuguese sample.................................................... 86
Appendix D: Regression analysis results for the Italian sample ........................................................... 94
Appendix E: One-way ANOVA and non-parametric tests .................................................................. 102
Appendix F: Questionnaire ............................................................................................................... 106

v
List of tables
Table 1. Companies in the online environment

Table 2. Pros and cons of e-commerce

Table 3. Items and scales employed in the questionnaire

Table 4. Socio-demographic characteristics of the respondents

Table 5. Descriptive statistics for Technical Quality (TQ)

Table 6. Descriptive statistics for Product Information (PI)

Table 7. Descriptive statistics for Information Quality of customer reviews (IQ)

Table 8. Descriptive statistics for Source Credibility (SC)

Table 9. Descriptive statistics for Customer Satisfaction (S)

Table 10. Descriptive statistics for Website Trust (T)

Table 11. Descriptive statistics for Social Influence (SI)

Table 12. Descriptive statistics for Performance Expectancy (PE)

Table 13. Descriptive statistics for Word-of-mouth Behaviour (WM)

Table 14. Descriptive statistics for Recommendation Adoption (REC)

Table 15. Descriptive statistics for Internet Experience (IEX)

Table 16. Descriptive statistics for User Online Experience (UEX)

Table 17: Determinants of Performance Expectancy

Table 18: Influence of experience on Performance Expectancy

Table 19. Determinants of Performance Expectancy together with experience

Table 20. Experience influences Technical Quality

Table 21. Experience influences Product Information

vi
Table 22. Experience influences Information Quality of customer reviews

Table 23. Performance Expectancy and Customer Satisfaction

Table 24. Performance Expectancy and Website Trust

Table 25. Performance Expectancy and Recommendation Adoption

Table 26. Performance Expectancy and WOM Behaviour

Table 27. Perceived Website Quality and Customer Satisfaction

Table 28. Perceived Website Quality and Website Trust

Table 29. Perceived Website Quality and WOM Behaviour

Table 30. Perceived Website Quality and Recommendation Adoption

Table 31. Online drivers of Website Trust

Table 32. Website Trust and Recommendation Adoption

Table 33. Performance Expectancy, Customer Satisfaction, Website Trust and Recommendation
Adoption as drivers of WOM Behaviour

List of figures
Figure 1. Stages in consumer decision-making

Figure 2. Conceptual model

Figure 3. Hofstede dimensions for Italy and Portugal

Figure 4. Determinants of Performance Expectancy

Figure 5. The impact of experience on perceived website quality

Figure 6. Relationship outcomes

Figure 7. Means plot for Customer Satisfaction

Figure 8. Means plot for WOM Behaviour

vii
Figure 9. Means plot for Product Information

viii
1. Introduction
1.1 Theme of the dissertation

Fashion consumption is without any doubt one of the oldest phenomena in human history; it has
been theorized and analysed by art historians, sociologists, anthropologists as well as
philosophers such as Immanuel Kant, who identified novelty as the main driver of fashion (Fang
et al., 2012). In more practical terms, fashion is an enormous business and has an economic
impact of global proportions. The apparel market size is expected to grow even more rapidly in
the coming years than it has in the recent past. Since the fashion industry is highly competitive
and requires companies to be flexible and react quickly to new trends, marketing is necessary to
satisfy the needs of contemporary consumers.

The way fashion is consumed has radically changed over the last two decades, with the
introduction of new technologies and the Internet. When applied to the retailing industry, the
Internet has been defined as the latest disruptive innovation (Christensen & Tedlow, 2000). As a
consequence, e-commerce, which uses websites to transact the sale of products and services
online (Kotler & Keller, 2012), has skyrocketed with global retail sales projected at 1.5 trillion
U.S. dollars for 2018 (Statista, 2016). Table 1 shows the two main types of online retailers,
according to Kotler and Keller (p: 439-440).

Table 1. Companies in the online environment.

Company Description Example

Brick-and-click A company that has an established physical Procter&Gamble,
company presence and adds the internet as a new Zara
distribution channel. Requires a multi-
channel strategy.
Pure click company A company that only has an online presence, Amazon.com,
without any physical store. Requires an Asos.com
online strategy.
Source: adapted from Kotler and Keller, 2012

Despite the challenges of selling consumer goods such as apparel online, clothing and shoes
represent one of the most popular online shopping category in Europe, particularly in Germany,
Great Britain, Finland and Denmark (PR Newswire, 2012).

1
A significant amount of marketing research highlights the importance of website design and
product information in fashion e-commerce websites (Sung & Jeon, 2009; McCormick & Livett,
2012; Seock & Norton, 2007; Vila & Kuster, 2011). Like brick-and-mortar retailers, pure click
retailers include store attributes such as product information and website layout to attract
customers. Recently, several online retailers in other industries have attempted to offer unbiased
and honest information in the form of online customer reviews. Because today’s fashion
consumer seeks information more from their peers than from other sources (Salonen et al., 2014),
hosting customer reviews can act as a way of building competitive advantage in the online
environment for fashion online retailers (e-tailers), so increasing the relevance of their websites.
Filieri et al. (2015) empirically show that customer reviews positively influence trust and
encourage customers to spread positive word-of-mouth (WOM) opinion about the website
hosting the reviews. However, there is a lack of research on the usefulness of customer reviews in
the context of fashion-and-apparel online stores.

This study considers the information contained in customer reviews as a parameter of overall
perceived quality. Among the many approaches available for understanding the impact of website
quality on online consumer behaviour, the Unified Theory of Acceptance and Use of Technology
(UTAUT) has been chosen here, similarly to a study conducted by Al-Qeisi et al. (2014) on
online banking usage.

This study intends to contribute to the body of literature on website design by answering the
question: what are the elements of a website that really explain online consumer behaviour, in the
context of fashion apparel?

1.2 Objectives of the dissertation

Empirical research on online consumer behaviour has been conducted extensively from different
perspectives. Many frameworks have been developed to explain the forces motivating consumers
to shop online. However, the majority of research about online fashion clothing consumption has
been done in the United States (Liu et al., 2013; Yoo & Kim, 2012; Kim & Lennon, 2010; Park
& Lennon, 2009; Seock & Norton, 2007; Park & Stoel, 2005; Goldsmith & Goldsmith, 2002);
thus, there is a research gap in Europe. To this end, this study is focused on European consumers

2
and their perceptions of fashion brands’ online stores. The study aims at accomplishing the
following objectives:

• To adapt and extend the UTAUT model in the context of online shopping for fashion
clothing.
• To compare the adapted model in two regions of Portugal and Italy.
In particular, this dissertation intends:
• To analyse the role of perceived website quality in generating customer satisfaction,
website trust and word-of-mouth behaviour.
• To explore whether online shoppers are or are not influenced by peer customer reviews.

1.3 Structure of the dissertation

In addition to the Introduction (Section 1), the dissertation is structured as it follows:

• Section 2 clarifies the theoretical background of this study in the form of a literature
review of earlier research. Firstly, fashion consumption is approached from a consumer-
oriented perspective. Secondly, online shopping is examined and website quality is
explained in its multi-dimensional nature. Lastly, the literature review ends with a
discussion of relationship-based outcomes such as satisfaction and trust in the online
environment. In this section the hypotheses are presented and a conceptual model is
illustrated.
• Sections 3 and 4 exhibit the research method and the results. Specifically, Section 3
focuses on the methodology and contextualizes the research. Section 4 consists of a
verification of the hypotheses, using multiple regression analysis.
• Section 5 concludes with a discussion of the results, providing managerial implications
along with the limitations of the study and future research.

2. Literature review
2.1 Fashion consumption

Fashion is an enormous entity encompassing not only clothing, but fine art in all its expressions,
architecture, poetry, and history. It even has an impact on the morals of a society (Fang et al.,
2012). More concisely, “the fashion system consists of all those people and organizations

3
involved in creating symbolic meanings and transferring these meanings to cultural
goods.”(Solomon et al., 2006:543). Davis (1994) claims that fashion clothing constitutes a code,
able to convey a range of meanings. These meanings are highly context-dependent and they vary
across cultures and communities. For the purpose of this study, fashion is intended as “the entire
spectrum of attractive clothes styles at any given time” (Fang et al., 2012:85) rather than a
generic phenomenon.

According to O’Cass (2004), individuals often define themselves on the basis of what they
possess: following this logic, the possession of fashionable clothes defines one’s status in society.
Consumers attach a certain importance to the fashion goods they possess, establishing a level of
involvement with fashion products. The term “fashion involvement” is used to indicate interest
with the clothing product category (Kim, 2005). Hourigan and Bougoure (2012) identify
materialism and gender as the main drivers of fashion clothing involvement, but also indicate
some outcomes of this construct: an interesting one is product information search. In other words,
consumers highly involved in fashion clothing consumption are more willing to dedicate time
and effort in search of information about fashion products. Additionally, deeply-involved
customers are very likely to gather information not only for themselves but also for sharing with
other consumers.

2.1.1 Consumer motivations for fashion consumption

Human behaviour is goal-driven. Similarly, consumer behaviour is performed as a means towards
an end. Consumption is the result of the discrepancy between a current state and a desired state,
whereby consumers act to achieve different types of goals (Barbopoulos & Johansson, 2016).
What motivates consumers to engage in an act of consumption is, according to Barbopoulos and
Johansson (2016), the desire to achieve specific goals. Firstly, the “gain goal” includes
consumers who are more sensitive to cost and value and want to use their money in a cost-
effective way; secondly, the “hedonic goal” implies instant gratification motivation, whereby
consumers are more sensitive to pleasure and moods and less worried about price and objective
quality. Lastly, the “normative goal” is associated with social norms and consumers perform
consumption in order to feel accepted by others or, alternatively, to make statements about status,
prestige and social class. These three goals can be seen as motivators for consuming goods in a
variety of consumer product categories, including fashion clothing. Usually, two or more

4
different goals are active at the same time and consumers may adopt those products and services
that are able to satisfy normative, gain and hedonic goals simultaneously.

Similarly, Fang et al. (2012) propose that psychological motives in fashion consumption can be
divided into rational motives and perceptual motives. Specifically, rational motives refer to the
need for high quality, good price and convenience, while perceptual motives include the search
for novelty, uniqueness and ‘showing-off’. However, it is hard to draw a line between rational
and perceptual motives and often consumption is a result of both factors.

In their research on luxury consumers, Eastman and Liu (2012) explain the concept of “status
consumption”, which is related to the need to consume products that confer a particular status or
prestige on the owner. Individuals who engage in status consumption strive to improve their
social standing and want to display their presumed high status to others, thus evoking envy and
respect. However, some groups of consumers purchase luxury products for personal reasons such
as self-reward, without developing public display and ostentatious behaviours. There is evidence
that status consumption, also referred to as conspicuous consumption, acts as a motivational
factor in purchasing luxury goods among Generation Y, or millennial consumers (Giovannini et
al., 2015).

2.1.2 Behavioural mechanisms of fashion consumption

Consumer behaviour is context-dependent, as the decision-making process is influenced by
several factors. A relevant factor influencing shoppers in the purchase of fashion clothing is the
judgement of peers or, more generally, of society. According to Solomon et al. (2006) it is
possible to distinguish between two different personality types in terms of behaviour, each
sensitive in one way or another to the opinion of those in their own social circle. In particular:
“The high self-monitors stress the brand of a consumer good (specifically clothing) more than
low self-monitors, who are on the other hand more positive to functional product attributes.”
(Solomon et al., 2006:545). Giovannini et al. (2015) stress that Generation Y shoppers show
high levels of self-monitoring and are highly brand conscious but not as brand loyal as older
consumers; they are consumers with high public self-consciousness who make purchasing
decisions based on the influence and opinions of their peers.

5
To reinforce the pressure of social norms on consumers, Rajagopal (2011:288) states that “one of
the principal drivers of consumer behaviour toward fashion apparel is the dominance of social
interactions.” In fact, individuals strive to fulfil two competing needs: i) the need to be approved
of and included by their peers and family, and ii) the need for autonomy and differentiation.
Thus, consumers purchase fashion clothing both to ‘fit in’ and to ‘stand out’ from the rest. The
need to be unique is particularly evident today, since individuals seek to create personalized
styles, purchasing a mix of specific clothing items (Gabrielli et al., 2013).

2.1.3 Models of fashion consumption

Fashion is characterized by continuous change. One of the most influential approaches to
understanding fashion change is the trickle-down theory, introduced by Georg Simmel in 1904
(Solomon et al., 2006). According to this model, there are two conflicting forces that drive
fashion change. Firstly, lower class individuals try to establish a better status for themselves by
imitating the fashions of higher social groups, in order to climb up the ladder of social mobility.
Simultaneously, the upper classes constantly adopt new styles to differentiate themselves from
those they consider below them on the social ladder to ensure that they are not currently imitated.
Following this pattern, new styles are created by the upper classes but eventually permeate to
those below, contributing to a continuous process of fashion change.

Nowadays, this approach is being modified to account for new developments in mass culture.
Today’s consumers tend to be more influenced by opinion leaders who are similar to themselves.
As a result each social group has its own fashion innovators who determine fashion trends. It is
often more accurate to speak of a “trickle-across effect, where fashions diffuse horizontally
among members of the same social group.” (Solomon et al., 2006:547)

Because of the “easy availability of media and magazines covering fashion news, catwalk styles,
and celebrity looks, consumers are increasingly interested in fashion and appearance, desire
newness and variety, and shop frequently” (Joung, 2014:689). The contemporary fashion
consumer is always up-to-date on the latest fashion trends and demands a wide assortment of
clothing, at affordable price. As the attention is moving away from the quality of a product to the
experience of it (Gabrielli et al., 2013), the life cycle of fashion clothing is shortening and fast-
fashion retailing has grown significantly.

6
2.2 Shopping and the consumer decision -making process

The usual aim of shopping is to acquire needed products or services. Consumers shop because
they experience a need and believe that shopping activities can satisfy that need (Davis, 2013).
Shopping is an activity that can be performed for either utilitarian (functional or tangible) or
hedonistic (pleasurable or intangible) reasons (Solomon et al., 2006). In their research on
shopping for fashion items, Carpenter and Fairhurst (2005) found that both hedonistic and
utilitarian shopping benefits are important for shoppers and are the drivers of customer
satisfaction. Cai and Xu (2006) define “shopping value” as consumers’ evaluations of the
shopping process, in a retailing environment. Shopping value is the result of several sub-
dimensions, such as: acquisition value (economic value of products), choice value (a broad
choice of merchandise), efficiency value (the efficiency of shopping trips), exploration value
(finding novelty goods), social interaction value (socializing with family and friends), self-
gratification value (improving personal well-being), aesthetic value (enjoying pleasant retail
visual displays), and transaction value (gaining pleasure from finding a bargain) (Davis, 2013).

A purchase is the result of a series of stages, referred to as the consumer decision-making process
(Solomon et al., 2006). Generally, shoppers go through five different stages (figure 1); first a
problem is identified as a need or a desire. Tension forms due to the gap between the current state
and the desired state. Next, the consumer searches for product information; the third stage is the
evaluation of alternatives and finally the purchase decision is made and a product is chosen. At
the end of the process the consumer evaluates the outcomes of the purchase, forming an opinion
of the retailing environment where he or she completed the purchase. However, consumers do not
go through this elaborate sequence every time they buy something, and they sometimes act
irrationally. For example, involved fashion consumers often engage in impulse buying, “in which
the rapidity of an impulse decision process precludes thoughtful and deliberate consideration of
alternative information and choices” (Park et al., 2006:435).

7
 Figure 1. Stages in consumer decision-making

Source: adapted from Solomon et al., 2006

The physical environment in which shopping activities take place is said to influence
consumption. Specifically in the context of fashion clothing retailing, environmental or
atmospheric characteristics such as merchandise colour, fragrances, materials and lighting affect
consumers’ decisions. Ha et al. (2007) suggest that coordinated merchandise colour leads to more
purchasing than uncoordinated merchandise colour. However, if a store doesn’t have clear aisles
and walkways, shoppers experience difficulty in moving around, and this impacts negatively on
perceived product quality.

Since the early 1990’s, a lot has changed in the retailing landscape, due to the introduction of the
Internet as a new way of acquiring products. In 1995, the arrival of Amazon promised to
revolutionize retailing, by being the first online bookstore to enable buyers to purchase a wide
variety of books with just one click (Kotler & Keller, 2012). As Javadi et al. (2012:81) point out:
“over the past few decades, the Internet has developed into a vast global market place for the
exchange of goods and services.”. Today, online shopping is as important as its offline
counterpart and has gained considerable attention from the marketing community. Because many

8
consumers choose the Internet to shop around in a variety of product categories, marketers are
evaluating whether or not e-commerce is going to replace traditional retailing (Solomon et al.,
2006).

2.3 Online shopping

Online shopping is a worldwide and growing phenomenon, especially in countries with a well-
developed online marketing infrastructure (Kau et al., 2003).
Kukar-Kinney and Close (2010) suggest that, in a similar manner to offline shoppers, online
shoppers form a need or want, then they search, consider alternatives, evaluate them, and finally
make a decision. In the online environment, the consumer decision-making process is less rigid
and sequential; shoppers change their mind easily and frequently revert back to the information
search stage, or abruptly interrupt a transaction and abandon their virtual shopping carts. Javadi et
al. (2012) point out that when potential consumers recognize a need for some article of
merchandise or a service, they go to the Internet and search for need-related information. When
the transaction is complete, the consumer evaluates the overall shopping experience at the
particular online store and forms an attitude towards it. However, online shoppers have to wait
some days until they actually have the product in their hands, therefore the performance of offline
activities such as delivery is a crucial factor (Kim et al., 2009).

The popularity and growth of online shopping is explained by the fact that consumers’ perception
is one of having received multiple benefits from this activity. In the next section, some of the
advantages of online shopping are presented.

2.3.1 Advantages of online shopping

Online shopping has become an everyday activity for tech-savvy consumers because it has
several advantages compared to traditional shopping. There is an ample supply of literature
devoted to the many advantages of online Internet shopping, with consumer convenience being
the most common motivational attribute. Other advantages include the amount of product and
brand related information, variety, and price comparison.

Convenience. From the consumer’s perspective, e-commerce has increased convenience by

removing many of the barriers caused by time and location (Solomon et al., 2006). Javadi et al.
(2012) claim that online stores are convenient and time-savers because consumers are no longer

9
obliged to physically travel to a store. Besides, online stores are open 24 hours a day and they are
accessible from anywhere. Consumers can save time and shop without the constrictions of going
to a physical store, parking or waiting in line. Reduction in shopping time is an important type of
convenience for at-home shoppers (Kau et al., 2003), particularly in contemporary societies
where people suffer from “time poverty”, that is the feeling of being pressed for time more than
ever before (Solomon et al., 2006). As a consequence, consumers are more than willing to adopt
any marketing innovation that allow them to save time.

Burke (1997), who analysed online shopping motivations using focus groups, reported that
shoppers appreciate the possibility of being able to perform other activities like cooking or
exercising while shopping online. Also, they can shop even if transportation is unavailable and
avoid crowded parking lots or bad weather. Online shopping also gives them the possibility to
access distant stores and not having to handle heavy and bulky packages, thereby putting an end
to order-size and weight constraints.

In their study on business-to-consumer (B2C) e-commerce, Jiang et al. (2013) identify five
convenience parameters: access, search, evaluation, transaction and post-purchase. Access
convenience is related to the ability to shop at any time and from any place; additionally,
consumers enjoy the benefits of accessing products that are unavailable in the location where
they live. Search convenience is concerned with the ability to navigate a website easily and
finding the wanted products quickly. Evaluation convenience is associated with the amount of
detailed and easy-to-understand product descriptions, which make the evaluation process easier.
Transaction convenience relates to check-out functions, which should be seamless and easy to
follow. Finally, post-purchase convenience indicates consumers’ perceptions of the time and
effort expended in possessing what they wish; it includes the effectiveness and conditions of the
delivered goods and the possibility to easy return unwanted items.

Further, empirical research on online shopper-motivation describes convenience as the possibility

to shop from home, one-stop shopping, completing shopping tasks quickly, avoiding regular
shopping issues such as having to deal with salespeople, the stressful experience of queuing,
traveling from store to store and avoiding crowds (Ganesh et al., 2010). Liu et al. (2013) refer
that convenience is valued by luxury shoppers. When purchasing luxury goods, such consumers
may buy online in order to save time and find desired items quickly. Convenience is therefore a

10
relevant shopping motive for varying fashion consumer segments. Besides convenience, other
attributes may influence the use of Internet shopping, including the amount of product and brand-
related information and variety of merchandising.

Amount of product and brand related information. A key feature of the Internet is its role in
facilitating information search for consumers (Brown et al., 2003). Information online is easy to
access, available at any time and usually for free. Consumers can access considerable amounts of
information quickly and effortlessly, without visiting a physical store. Online information can
include video, images, photos, text, audio, tables and graphs. Elliott and Speck (2005) claim that
the right amount of product information is positively linked to shopper’s attitude towards a
website and appears to increase the amount of online shopping, spending and satisfaction with
online purchases. Online stores selling fashion clothing are sometimes more capable of
successfully and innovatively conveying to the customer up to date trends and sharing cutting
edge fashion advice than physical stores (McCormick & Livett, 2012).

Some websites offer a reliable way of retrieving product information, in the form of online
customer reviews. As De Maeyer (2012) points out, consumers rely increasingly on online
information for their purchase decisions and they regularly consult online consumer reviews or
product ratings before making an important purchase. Beyond product-related information, online
information also includes contact information, general company information, consumer-oriented
policies, and customer support. Consumers who are eager to have timely information on
promotions, new product launches and events related to brands, are now also able to find it on
social media platforms (Rohm et al., 2013).

Variety. The maturing of Internet technology has led to an increased variety of product offerings
(Jeong et al., 2009). Variety seeking is “the desire to choose new alternatives over more familiar
ones” (Solomon et al., 2006:267). Since searching on the Internet is easier and faster than
traditional search, online shoppers search more extensively for a variety of brands and products
than their offline counterparts (Brashear et al., 2009). On the Internet, consumers are likely to
find a sufficient variety of products to choose from (Rajamma et al., 2007) without the hassle of
physically visiting multiple stores. Sung and Jeon (2009) suggest that variety of merchandise is
considered important by shoppers, particularly those driven by hedonistic and recreational

11
motivation. A consequence of product variety is the possibility of finding different prices for the
same product, making price comparison an additional incentive to shop online.

Price Comparison. According to Javadi et al. (2012: 82) “the Internet shifted the balance of
power in favour of consumers as it became very easy for them to make shopping comparisons
and evaluate alternatives without being pressured by salespeople.” Internet shoppers usually
search for many alternatives because it is easier and less costly to compare products online than
offline. As Brashear et al. (2009) point out, since prices are more transparent on the Internet,
consumers can search for the lowest price available for the specific product they want.

Because online retailers provide many sales promotions, such as free gifts, discounts, or free
shipping, shoppers tend to believe that product prices in online stores are usually lower than in
brick-and-mortar stores. Price promotions positively influence customer estimates in regard to the
fair price of a product and enhance the perceived value of the deal (Park & Lennon, 2009).
Furthermore, research shows that luxury consumers like to shop online because they can compare
prices and look for the best deals (Liu et al., 2013).

2.3.2 Disadvantages and perceived risks of online shopping

Despite the many advantages, online shopping has its drawbacks. Consumers are sometimes
frustrated with Internet shopping, especially when they are not familiar with the online
environment. With online shopping, consumers may develop low trust and perceive a greater
degree of risk because of the lack of face-to-face, shopper-to-salesperson communication (Javadi
et al., 2012). Levin et al. (2005) affirm that consumers still prefer offline shopping for products
like clothing and health goods, due to the consumers’ need to touch, smell or try on the item. On
the other hand, consumer behaviour is changing fast, and consumers are now more comfortable
with purchasing different kinds of products online. Two relevant aspects regarding online
shopping are security issues and the shopping experience itself.

Security and privacy issues. These aspects of online shopping are, according to Solomon et al.
(2006), related to the security surrounding credit card payments and the privacy of shared
information. In the online environment, security and privacy refer to “a shopping site taking
appropriate measures to protect consumers’ personal information from being misused.”(Hsu et
al., 2014:336). Online shoppers may perceive additional risks when a website requires them to

12
reveal a great amount of personal and financial information, especially on the checkout pages
where they are asked to input personal information such as credit card details and address
(Rajamma et al., 2009). Levin et al. (2005) state that the perceived security of an online
transaction is decisive in the last stages of the decision-making process.

Moreover, consumer insecurity issues when shopping online are considered one of the most
important obstacles to the growth of e-commerce. Brick-and-mortar store shoppers perceive a
higher sense of privacy protecting any exchange of financial information (Flavián et al., 2006).
Similarly, online shoppers’ frustration with websites, due to the lack of perceived security, may
explain why high levels of online search activity do not always translate into high purchase levels
(Seock & Norton, 2007).

Security and protection of consumer information (such as credit card number, personal details, e-
mail address) are mentioned by O’Cass and Carlson (2012) as antecedents of website service
quality because they heavily influence the overall perceived quality of a shopping website. Vila
and Kuster (2011) report that the majority of consumers would not purchase from a website that
did not display a trust mark or security policy, even if they find a good deal. On the contrary, if
the perceived level of security meets the consumer’s expectations, he or she may be willing to
disclose his or her personal information with an increased level of trust (Kim et al., 2009).
According to Yoon (2002), reassuring shoppers about online security in the form of control of
information is an important stage in developing trust, and symbols and trust marks such as
VeriSign or Visa logos support the build-up of trust.

Unsatisfactory shopping experience. Another possible inconvenience of online shopping is the

actual shopping experience. McCormick and Livett (2012) indicate that consumers are
demanding online shopping experiences that provide high levels of interaction and entertainment.
This is particularly relevant for younger consumers, seeking original site environments with
innovative designs and features.

Compared to brick-and-mortar stores, websites cannot offer as many sensory stimuli (e.g., touch,
taste, smell, vision, sound). While it may be satisfactory to buy a computer or a book on the
Internet, buying clothing and other products where touching the item or trying it on is essential
could be less attractive (Solomon et al., 2006). Levin et al. (2005) classify products into “low

13
touch” and “high touch”. The former are products that are more suitable for online shopping
because of the special importance placed on shopping quickly: they include products like airline
tickets and computer software, which don’t have a sensory component. The latter are products
that consumers feel they need to touch, smell or try on, like clothing, sporting goods, and health
and grooming products. Bae and Lee (2011) mention that for “experience products” quality can
be evaluated only after trying them out, as opposed to “search products”, whose information
about attributes can be easily acquired before purchasing. According to this distinction,
experience products are less likely to be purchased online, since their characteristics cannot be
evaluated completely by consumers. Research by Rajamma et al. (2007) show that consumers
perceive services to be more congruent with online stores and tangible products with brick-and-
mortar stores. Solomon et al. (2006:319) report that “consumers with a higher need for tactile
inputs tended not to use the Internet so much for product purchase”. Table 2 summarizes the main
pros and cons of e-commerce, from the consumer’s perspective.

Table 2. Pros and cons of e-commerce

Benefits of e-commerce Limitations of e-commerce

Shop 24 hours a day Lack of security
Less travelling Fraud
Can receive relevant information in seconds Can’t touch items
from any location Exact colours may not reproduce on
More choice of products computer monitors
Greater price information Expensive to order and then return
Lower prices so that less affluent can purchase Potential breakdown of human
Fast delivery relationships
Electronic communities
Source: adapted from Solomon et al., 2006

2.4 Website quality dimensions: technical quality and information quality

Website quality has been the object of substantial empirical research in the field of online
marketing. Aladwani (2006) proposes a model whereby perceived website quality is the result of
four sub-dimensions: technical quality (ease of use, security, well-organized links, speed of page

14
loading, interactivity and ease of access) general content, specific content (content clarity,
currency, completeness, usefulness and company information, product details, customer support)
and appearance.
Websites can be seen as the interface mediating the encounter and interaction between consumers
and companies, in an environment characterized by any level of technology (O’Cass & Carlson,
2012). Brick-and-mortar retailers plan store layouts to help consumers find merchandise (Ha et
al., 2007). In the same way, online retailers design their websites so that shoppers are able to find
the wanted products quickly. Therefore, technical quality is essential: the website must be
intuitive, well organized, and easy to follow and include appropriate navigation structures for the
customer to interact seamlessly during the shopping process. A study by Feng et al. (2014) shows
that website quality elements like usability and technical adequacy positively affect perceived
usefulness, regardless of cultural differences. Design communicates value to the customer
(McCormick & Livett, 2012) and a well-designed website can provide users with a higher level
of perceived control and self-efficacy, so satisfying their utilitarian needs. Issues associated with
poor design include complex navigation, long download times, confusing return policies and low
levels of interactivity (Siddiqui et al., 2003). Seock and Norton (2007) report that navigation
factors such as uncluttered screens and easy search paths and links are important for shoppers to
evaluate the quality of a fashion online store.

Another dimension of perceived website quality is content in the form of information. Kim et al.
(2010:79) define digital content as “information and experience that provide value to users”.
Content is not only a mere combination of text, images and music but also includes information
that will be valuable and helpful for website users. Elliott and Speck (2005:41) claim that
“product information includes the amount, accuracy and form of information about the products
and services offered by a website.”

Several researchers have emphasized the importance of information quality in a website. Feng et
al. (2014) claim that information quality has a positive impact on perceived usefulness and
acceptance of online technology. Hasanov and Khalid (2015) mention the quality of information
offered by a brand in online shopping websites as an important factor in assessing overall website
quality and actually say that information quality has the highest influence on customer
satisfaction among all website quality dimensions. Bai et al. (2008) argue that an information

15
system which meets the needs of its users will reinforce customer satisfaction. By contrast, if the
system does not provide the needed information, its users will be dissatisfied and subsequently
look elsewhere for information needs. Aladwani (2006) explains how information search is one
of the main stages in the purchasing process and inaccurate and ambiguous information may have
a negative impact on consumer trust.

The degree and type of information available on a website is crucial to those particular products
which are normally personally examined by the shopper, i.e. fashion clothing. In the context of
online retailing, the inability to physically examine the product prior to purchase could result in
increased perceived risk (Kim & Lennon, 2010). Because fashion clothing are experience goods,
their quality can only be fully determined after purchase, and thus, the inability to try on clothing
before purchase is a major risk when shopping online (Ha et al., 2007). In online apparel
retailing, people rely on the product information provided on the website without being able to
physically evaluate a product (Yoo & Kim, 2012).

According to a study by McCormick and Livett (2012), online fashion shoppers appreciate
practical information such as size, fit and washing instructions to satisfy their utilitarian needs.
The same study suggests that, along with utilitarian product information, shoppers seek fashion
information that provides hedonic value, such as fashion trends and style advice. Hourigan and
Bougoure (2012) argue that the proliferation of the Internet has radically changed the way
consumers access product-related information and it has never been so easy for them to follow
current trends and styles. Seock and Norton (2007) state that product information has a major
influence in shaping consumers’ perceptions of clothing websites.

Information search is one of the main fashion consumption practices (Salonen et al., 2014). An
interesting activity in this respect is reading product reviews. Consumers search for product
reviews relating to items of interest because they do not entirely trust traditional advertising.
Thus, from a consumer’s perspective, customer reviews can enhance the overall quality of a
website.

2.5 Online customer reviews

The development of the Internet and information technology has resulted in a new market
phenomenon: the proliferation of online customer reviews (Zhang et al., 2013). User reviews

16
have emerged as a valuable source of information for shoppers, complementing and often
substituting offline word-of-mouth (WOM). Compared to offline word-of-mouth, “online
consumer reviews can reach far beyond the local community, because consumers all over the
world can easily access a review via the Internet.” (Zhang et al., 2013:1116). As of 2015, 90% of
consumers read online reviews as part of their pre-purchase research and 88% of them trust
online reviews as much as personal recommendations (Rudolph, 2015). For instance, e-tailers
such as Amazon have become a popular site-destination for consumers (De Maeyer, 2012), who
can find a copious amount of customer reviews for any product category, from books to clothing.

As Chen and Xie (2008) point out, information created by retailers is more product-oriented, and
often only includes the product’s technical specifications. On the other hand, information coming
from customers is user-oriented and gives insights on the holistic experience of acquiring and
using a product. According to De Maeyer (2012), customer reviews can increase post-purchase
satisfaction, by educating consumers about important product attributes. Moreover, consumers
who make informed decisions are more willing to pay a premium price, be satisfied, and become
promoters of the particular brand online.

However, relatively little work exists to explain how product category impacts the behaviour of
consumers related to customer reviews (De Maeyer, 2012). For example, a study from Cheema
and Papatla (2010) shows that consumers rely on online information more heavily for utilitarian
products, while they want to try hedonic products for themselves or obtain specific information
from friends or acquaintances. Brown et al. (2003) claim that experience products need to be
personally inspected or tried out. However, Bae and Lee (2011) refer that online customer
reviews can assist shoppers in evaluating an experience product, since the information is
provided by other consumers who have already purchased and experienced that product.
Accordingly, online customer reviews are said to boost the sales of experience goods (Zhang et
al., 2013).

As reported by Purnawirawan et al. (2015), product type affects perceived usefulness of online
reviews, since experience products contain more subjective elements that shoppers cannot
evaluate prior to purchase. Therefore, as consumers cannot assess the quality of products directly,
they rely on customer recommendations. Also, online reviews have been found to be perceived as

17
useful when the brand is unfamiliar; the informative effect of online reviews is stronger for
unknown brands than for well-known brands (Purnawirawan et al., 2015).

Filieri et al. (2015) found evidence that the quality of information of customer reviews is a strong
predictor of website trust, which in turn impacts WOM behaviour. Besides, Aladwani (2006)
proposes that a website with a higher specific information content can attract more consumers
than one with less specific content. From this perspective, consumer-generated information can
be considered as being as important as official product information in helping shoppers evaluate
the quality of an e-commerce website.

Despite the success of customer reviews, Lee and Ma (2012) claim that this type of feedback can
create a degree of uncertainty and confusion, if consumers experience cognitive incongruence
because of conflicting information. Additionally, “conflicting reviews about a product, service, or
company can also hinder consumers’ information processing and increase difficulty in
determining the quality of the information” (Lee & Ma, 2012:113). Park and Lee (2008) contend
that when consumers are confronted by a surfeit of reviews they can become cognitively
overloaded trying to process such large amounts of information in a limited time. To make
matters worse, because shoppers cannot communicate face-to-face with the customers who have
written the reviews, it is harder for them to evaluate the trustworthiness of the review sources
(Filieri et al., 2015).

To summarize, perceived website quality is considered here as defined by three subordinate

parameters: technical quality (ease of use, well-designed web pages, high speed and
accessibility), content quality (satisfying and high-quality product information) and specific
content quality (information provided in the online customer reviews). In the context of the
Unified Theory of Acceptance and Use of Technology (UTAUT), “facilitating conditions” are
described as the extent to which an individual believes that a technical infrastructure exists to
support and facilitate the use of a specific system (Venkatesh et al., 2003). Al-Qeisi et al. (2014)
claim that website quality can be conceptually equal to “facilitating conditions” in the UTAUT,
because it represents the resources and support available to permit customers to behave in a
specific manner. In the following section, the UTAUT is discussed and another of its key
constructs, performance expectancy, is analysed.

18
2.6 The role of performance expectancy

“Performance expectancy is the degree to which the use of technology helps consumers perform
certain activities.” (Al-Qeisi et al., 2014:2283). Similarly, “a user becomes aware that a particular
technology is useful when the technology or system reduces the time it takes to do a job while
simultaneously increasing efficiency and accuracy” (Feng et al., 2014:29).
The construct performance expectancy is conceptually equivalent to “usefulness” in the
Technology Acceptance Model (TAM), introduced by Fred D. Davis in 1989 to explain the
adoption and use of technology (Al-Qeisi et al., 2014). Since its introduction, the TAM model
has been successfully applied in several contexts and it is considered an effective model to
predict behaviour. Even so, cross-cultural research performed by Smith et al. (2013), tested the
TAM model in three countries (Norway, Germany and United States) and the full model did not
hold true for the European samples.

Extending from the TAM, Venkatesh et al. (2003) used the Unified Theory of Acceptance and
Use of Technology (UTAUT) to predict employee adoption of information technologies.
According to the UTAUT, “four constructs play a significant role as direct determinants of user
acceptance and usage behaviour: performance expectancy, effort expectancy, social influence and
facilitating conditions.” (Venkatesh et al., 2003:447). Specifically, the performance expectancy
construct is said to be the strongest predictor of intention (Venkatesh et al., 2003). In fact, if an
individual believes that using a technology helps to achieve a task, he or she will be more likely
to adopt that technology. As a consequence, if consumers perceive online shopping as helpful and
useful they will be motivated to use the Internet to purchase products and services.
Performance expectancy is influenced by several external variables, including website quality.
Research by Feng et al. (2014) supports the idea that when a website has a degree of quality
which is appropriate, users’ perceived usefulness increases. Empirical research by Al-Qeisi et al.
(2014) shows that website design elements have a direct impact on performance expectancy,
because when people believe they will benefit from a high quality website, not only will they use
it but they will evaluate it as being useful. The UTAUT suggests that another construct impacting
an individual’s behaviour is the influence of social factors.

19
2.7 Social influence

“Social influence is defined as the degree to which an individual perceives that important others
believe he or she should use the new system.” (Venkatesh et al, 2003:451).
The UTAUT assumes that social influence has an impact on individual behaviour. Similarly,
previous research has found that behavioural intention is affected by social surroundings:
consumers are more likely to adopt a new technology if important others think they should do so
(Yang, 2010). Javadi et al. (2012) affirm that often consumers act in ways that are based on their
perception of what others think they should be doing, thus supporting the observation that the
opinion of friends and peers influences online buying behaviour.
Brown and Venkatesh (2005) propose that the adoption of technology in the household is subject
to a range of social influences: friends, family or co-workers and such secondary sources as TV
and newspapers. Specifically, consumers appreciate informal advice about products, in the form
of WOM: because “WOM is initiated by customers independent of the market, it is perceived to
be more reliable and trustworthy than firm-initiated communications.” (Zhang et al., 2013: 1116).

Nowadays, online shoppers tend to trust consumer-generated information more than seller-
created advertising (Chen & Xie, 2008). From this perspective, online customer reviews can be
seen as a form of electronic WOM, and an online consumer review acts as a social influence
(Park & Lee, 2008). Thus, online shoppers may be influenced both by offline peers (family and
friends) and by online peers (the reviewers who write online product reviews). Since online
reviews are written by unknown sources who have no previous relationship with the recipient, it
is harder for shoppers to evaluate their reliability. However more and more websites require
reviewers to provide some personal qualification information to help consumers assess the
credibility of each reviewer (Filieri et al., 2015). Thus, reviewer credibility may be important in
influencing consumers’ perceptions of online shopping.

Based on the literature review presented above, the first group of hypotheses concerns the
relationship between website quality elements and performance expectancy:
H1: Perceived website quality has a positive impact on performance expectancy.
H1a: Technical quality has a positive impact on performance expectancy.
H1b: Product information has a positive impact on performance expectancy.

20
H1c: The quality of information in customer reviews has a positive impact on performance
expectancy.

Reference groups are said to influence consumers’ purchasing behaviour (Hsu et al., 2013)
Therefore, recommendations from these groups are expected to have an impact on how online
shoppers evaluate the activity of online shopping:
H2: Social recommendations affect performance expectancy.
H2a: Family and the opinion of friends affect performance expectancy.
H2b: Reviewers’ credibility affects performance expectancy.

2.7 The moderating impact of previous online experience

Researchers report that experience and proficiency influence use-of-technology behaviour (Al-
Qeisi et al., 2014). The UTAUT conceptualizes experience as an individual variable that
moderates the key relationships between the model’s various constructs (Celik, 2016).

The outcome of an action provides people with experience that influences the likelihood of the
action being repeated in future (Lim, 2013). There is evidence that consumers who have previous
experience in shopping online for clothing are more confident in their purchasing ability
(Goldsmith & Goldsmith, 2002). Similarly, Park and Stoel (2005) claim that people with
previous experience of Internet shopping find it less risky to buy online in the context of apparel
shopping. More experience or familiarity with online shopping generally leads to greater
expertise, hence experienced online shoppers are able to perform shopping-related tasks more
efficiently (Broekhuizen & Huizingh, 2009). Thus, online shoppers’ perceptions of the usefulness
of an online store might be influenced by experience with browsing and/or using websites.
Experience can also include users’ Internet knowledge and expertise. Cheema and Papatla
(2010:981) infer that experience indicates “the extent of time that a consumer has been using the
Internet”. As consumers' experience with Internet or online shopping increases, they tend to
perceive more benefits associated with Internet shopping and show a greater predisposition to use
this means as a shopping medium (Kwon & Noh, 2010).

21
In the context of online reviews, Filieri et al. (2015) propose that the level of experience in using
online customer reviews has an impact on the behaviour towards the website. Consumers with
extensive experience of customer reviews are expected to be more confident in using them and to
easily identify fake and unreliable product reviews. In this study, experience with customer
reviews is approached as a component of overall user online experience.

Not only does experience influence performance expectancy, but also the users’ perceptions of
facilitating conditions, so that users who are more experienced with the Internet have more
positive perceptions of website quality elements (Al-Qeisi et al., 2014). Online experience plays
a moderating role in strengthening favourable attitudes towards a retail website and it also
impacts product information perceptions because experienced shoppers place higher importance
on information (Elliott & Speck, 2005).

Therefore, the following hypotheses investigate how two types of experience influence
performance expectancy and perceived website quality:
H3: Internet knowledge and user online experience affect performance expectancy.
H4: Internet knowledge and user online experience affect perceived website quality.
H4a: Internet knowledge and user online experience affect technical quality.
H4b: Internet knowledge and user online experience affect product information.
H4c: Internet knowledge and user online experience affect information quality of customer
reviews.

While this section presents the determinants of online consumer behaviour, the next part of the
literature review deals with the behavioural outcomes of perceived website quality and
performance expectancy. Particularly, the constructs of satisfaction, trust and word-of-mouth are
described.

2.8 Relationship outcomes of website quality

Several studies have investigated online consumer behaviour and attempted to discover which
elements influence online shoppers the most. Yoon (2002) found evidence that shoppers feel
satisfied when a website is functional, facilitates a high navigation speed and guarantees privacy,

22
while a company’s reputation determines trust. Other more personal variables, such as familiarity
with e-commerce, influence website satisfaction and trust. In another study (Devaraj et al., 2002),
performance expectancy was found to influence consumer attitudes and satisfaction towards the
e-commerce channel.

In this study the UTAUT is extended. According to the original model, the main outcome is the
usage of technology. Recently, along with the rise in e-commerce, questions of satisfaction, trust,
and loyalty have become increasingly important in the online context (Chung & Shin, 2010).
Thus, in the context of this dissertation, the outcomes are relationship-based constructs, including
such factors as satisfaction, trust, adopting recommendations and WOM behaviour.

2.8.1 Customer satisfaction, website trust & recommendation adoption

The most influential approach to customer satisfaction is the “disconfirmation of expectations”
model introduced by Oliver (1980), which proposes that consumers experience satisfaction when
comparing their perceptions of a product’s performance with their previous expectations. Spreng
et al. (1996) argue that feelings of satisfaction arise not only when the product itself performs
well, but also when the information about that product, usually provided by marketers, is of high
quality. Consumers can be satisfied or dissatisfied with the information provided by advertising,
salespeople or other sources in the same way as they are disappointed or delighted by the product
itself. However, if sellers promise more than they can deliver, then the consumer’s expectations
are negatively disconfirmed and feelings of dissatisfaction may arise.

While Oliver (1980) described satisfaction in the context of a single transaction, another
approach is to consider satisfaction as a cumulative construct. In relationship marketing literature,
the concept of customer satisfaction goes beyond a mere transactional exchange, and has been
defined as a holistic evaluation of the total purchase and consumption experience (Loureiro et al.,
2014). Similarly, “customer satisfaction is measured as the general satisfaction of a customer,
which is based on all cumulative experiences with a company, a product or a service” (Filieri et
al., 2015:177).

The construct of satisfaction is crucial in e-commerce because e-tailers’ success depends on

customer satisfaction (Hung et al., 2014). User satisfaction with a website is found to be
positively influenced by website elements such as information content, graphics and their style,

23
as well as good navigational guidance (Kim et al., 2009) and usability (Flavián et al., 2006).
Similar to store layout in offline retailing, website design immediately projects the first
impression of the retailer as an organization, and this factor is a major contributor in generating
satisfaction in an online environment. Hung et al. (2014) found that the quality of information
provided by a website is one of the main drivers of user satisfaction with any particular website.

Finally, Devaraj et al. (2002) refer that perceived usefulness directly influences satisfaction with
e-commerce. Since performance expectancy is conceptually equivalent to perceived usefulness,
users who believe they will benefit from shopping online might be more satisfied with their
online shopping experience. Therefore:

H5a: Performance expectancy directly impacts customer satisfaction.

In the marketing literature, trust is considered a key component in the development of marketing
relationships (Yoon, 2002). Trust occurs “when one party has confidence in an exchange
partner's reliability and integrity” (Morgan & Hunt, 1994:23). Also, trust has been defined “as the
belief that a partner’s word or promise is reliable and a party will fulfil his/her obligations in the
relationship” (Wong & Sohal, 2006:247).

In the online retailing context, trust may be a fundamental component for initiating a transaction
(Kim et al., 2009). Similar to satisfaction, website trust is influenced by such website elements as
privacy and security (Kim et al., 2009) and perceived usability (Flavián et al., 2006). Filieri et al.
(2015) report that website quality is a strong predictor of website trust. Therefore, performance
expectancy might have a mediating role between website quality elements and trust towards the
website.

H5b: Performance expectancy directly impacts website trust.

In a marketing context, it is impossible to completely detach trust from satisfaction, since

consistent satisfaction with individual transactions over time is required for trust to develop
(Chung & Shin, 2010). Yoon (2002) also finds evidence that satisfaction with previous online
transactions affects website trust. Accordingly, customer satisfaction has a direct and positive
influence on the degree of website trust (Flavián et al., 2006). Thus, if shoppers are satisfied with

24
a fashion online store they are reasonably expected to develop a feeling of trust towards the
website.

H5c: Customer satisfaction directly impacts website trust.

Liu et al. (2013) suggest that online consumer trust is related to the credibility of the sellers and
of the online reviews and ratings from other customers. Online customer reviews act as a
recommender, as they provide either a positive or negative indicator of product popularity (Park
& Lee, 2008). In the consumer’s best interest, a trustworthy website is expected to monitor the
legitimacy of the reviews submitted by its users, so avoiding the issue of fake content (Filieri et
al., 2015). Accordingly, online shoppers who consider a website trustworthy may be more
willing to adopt the recommendations it offers. Similar to blogs, where the blogger’s credibility
leads to recommendation adoption (Hsu et al., 2013), consumers are more incentivized to adopt
other customers’ recommendations when a feeling of trust has been previously established.

H5d: Website trust directly impacts recommendation adoption.

Performance expectancy is said to predict technology adoption and usage (Al-Qeisi et al., 2014).
Thus, if online shoppers evaluate shopping websites as useful and helpful in accomplishing
shopping tasks, they might actively adopt the recommendations provided by online customer
reviews.

H5e: Performance expectancy directly impacts recommendation adoption.

2.8.2 Word-of-mouth behaviour

Word-of-mouth (WOM) has been defined as a marketing phenomenon which consists of passing
product information from one person to another via spoken communications (Hung et al., 2014).
The influence of peer opinion is at times more powerful than one’s own perceptions (Solomon et
al., 2006). In online retailing, WOM can frequently have a significant impact, both positive and
negative, on the acquisition of new customers and can be one key online relationship outcome
(Chung & Shin, 2010).

O’Cass and Carlson (2012) claim that positive assessments of website quality are directly
correlated with engaging in favourable WOM behaviour. In other words, users who perceive a
website as being of high quality may spread positive word-of-mouth about that website. Thus, it

25
can be hypothesized that if users perceive online shopping as useful they will say favourable
things about the websites they use.

H5f: Performance expectancy directly impacts WOM behaviour.

Previous research states that customer loyalty and word-of-mouth are dependent on customer
satisfaction; Carpenter and Fairhurst (2005) affirm that customer satisfaction has a positive
impact on WOM in the context of apparel retailing. Moreover, “the likelihood of customers
spreading WOM will depend on their satisfaction level for at least two reasons.” (Matos & Rossi,
2008:580). First, if the product or service performance exceeds the customer’s expectations he or
she will be motivated to tell others about their positive experience. Second, if the customer’s
expectations are not fulfilled, possibly creating a negative experience, that customer will spread
WOM to express their negative feelings, including anger and frustration, in order to alleviate
their anxiety and to warn others.

H6a: There is a significant positive effect of satisfaction on WOM behaviour.

Matos and Rossi (2008) argue that a high degree of trust is associated with a greater tendency to
engage in favourable WOM. In a study of young Facebook users, Kucukemiroglu and Kara
(2015) found that users who received information from trusted sources are more likely to
participate in online word-of-mouth communication and actively make product recommendations
to their friends. Thus, if consumers trust a website they may talk positively about it with their
peers.

H6b: There is a significant positive effect of trust on WOM behaviour.

As mentioned by Filieri et al. (2015), consumers who use other consumers' recommendations are
more likely to improve their decision making. If consumers find high quality customer reviews
on a website and adopt the recommendations from other customers, this experience may motivate
them to initiate and engage in conversations with friends and family and even recommend the
website.

H6c: There is a significant positive effect of recommendation adoption on WOM behaviour.

Figure 2 shows the conceptual model developed from the hypotheses.

26
 Figure 2. Conceptual model

Source: Author’s elaboration

3. Methodology
3.1 Research context: Italy and Portugal

A consumer’s culture determines “the overall priorities he or she attaches to different activities
and products” (Solomon et al., 2006:500); thus, the success or failure of specific products and
services has to be referred to the cultural context (Solomon et al., 2006). Although the Internet is
a global space wherein information about products and brands is available to consumers from
different countries, evidence proves that online consumer behaviour is not homogenous across
the Web and may differ by national culture (Christodoulides et al., 2012).

As of 2016, Portugal is estimated to have about 6.9 million Internet users, corresponding to a
penetration rate of 67.3%, while Italy counts around 39 million Internet users, with a penetration
rate of 65.6% (Internet Live Stats, 2016). Therefore, Portuguese and Italian consumers are
expected to be similar in their familiarity with the Internet channel. However, it is useful to draw

27
some distinctions between the two countries, using the Hofstede (1994) cultural dimensions as a
background.

In the context of this study, the dimension of uncertainty avoidance is quite interesting and can
be related to the usage of Internet. This dimension is defined as “the degree to which people in a
country prefer structured over unstructured situations” (Hofstede, 1994:5). Cultures which score
high in this dimension can be called rigid, as they rely heavily on rules and traditions. Hofstede
(1994) claims that in rigid countries, the feeling “what is different is dangerous” prevails, while
in low uncertainty avoidance societies the feeling would rather be “what is different is curious”.
According to Smith et al. (2013), in high uncertainty avoidance cultures, consumers' discomfort
with ambiguity may influence perceptions of the usefulness of online shopping. Portugal scores
99 on this dimension, thus it has a very high preference for avoiding uncertainty and might resist
innovation. Italy has a high score of 75 meaning Italians are not comfortable in ambiguous
situations (The Hofstede Centre, 2016). Compared to other countries such as the United
Kingdom, which scores only 35 on uncertainty avoidance and has an Internet penetration rate of
92% (Internet Live Stats, 2016), it can be said the Portugal and Italy haven’t reached their full
potential in terms of e-commerce.

Hofstede (1994:6) defines individualism, as opposed to collectivism, in these terms: “it is the
degree to which people in a country prefer to act as individuals rather than as members of
groups.” Individualism might impact consumers’ online behaviour: in fact, consumers from more
individualistic cultures are expected to have stronger desires for personal convenience (Smith et
al., 2013). Therefore, if a country presents a high score in individualism, people might be more
willing to use online services if they are perceived as convenient and useful. Portugal, in
comparison with the rest of the European countries, presents a low score of individualism of 27
(The Hofstede Centre, 2016). This means that Portuguese people are committed to be part of a
group, be that family, extended family, or extended relationships. At a score of 76, Italy is an
individualist culture, self-centred, especially in the big and rich cities of the North (The Hofstede
Centre, 2016). Italians might be more willing to use online shopping services if they can gain
personal benefits for themselves, such as lower prices or reduced shopping time and effort.
Figure 3 shows a comparison between the two countries according to the Hofstede cultural
dimensions.

28
 Figure 3. Hofstede dimensions for Italy and Portugal.

Source: The Hofstede Centre (2016)

3.2 Data collection

The questionnaire for this study is developed based on previous studies and an online platform
was used to collect data. The web questionnaire is mainly composed of closed questions that
were measured using a 5-point Likert scale; it was administered through social media platforms
in form of a link and participants were asked to share the link with others. The questionnaire was
originally developed in English and it was translated to Italian and Portuguese; back translation
was employed to assure that all questionnaires communicate the same ideas and content. Before
launching the questionnaire, a pilot test with 10 Portuguese users and 10 Italian users was
performed in order to analyse the content validity and only few adjustments were made. The use
of online platforms seems appropriate since the current study is devoted to fashion brands who
use the online context to promote and sell their products. Even so, in order to avoid bias, and
following Filieri et al. (2015), the fashion consumers who participated to this study i) use fashion
websites that provide consumer-generated content (CGC) such as online customer reviews ii)
have purchased fashion clothing online and iii) are living in Portugal –in the Lisbon area– or in
Italy –in the regions of Piedmont and Lombardy.

The scale and items used to measure the constructs in this study are adapted from previous
studies. Particularly, Technical Quality, Information Quality of customer reviews, Source
Credibility, Customer Satisfaction, Word-of-mouth Behaviour, Recommendation Adoption and

29
User Online Experience were adapted from research on consumer-generated media on
TripAdvisor, by Filieri et al. (2015) (see table 3). The construct Product Information was
developed similarly to the one by Seock and Norton (2006), who identified some of the most
relevant attributes of clothing websites in the United States. The items for the construct Website
Trust were adopted from Filieri et al. (2015) and Liu et al. (2013). The constructs Performance
Expectancy, Social Influence, Internet Experience were developed from research on online
banking by Al-Qeisi et al. (2014) and were adapted to fashion online stores.

Table 3. Items and scales employed in the questionnaire

Construct Items Source

Technical Quality This fashion website is: Filieri et al. 2015
(5-point, completely disagree - - easy to use
completely agree) - has well-organized hyperlinks
- provides opportunities to interact with other
customers
- has high speed of page loading
- is easily accessible from different media
- guarantees users’ privacy
Product Information This fashion website: Seock & Norton
(5-point, completely disagree - - shows all the colours available for each product 2006
completely agree) - shows all the sizes available for each product
- tells the prices of products clearly
- gives up-to-date information about products and
trends
- has good quality photos of products
- truthfully shows the colours of the products
Information Quality of The information in the customer reviews is: Filieri et al.
customer reviews - timely 2015
(5-point, completely disagree - - relevant to my needs
completely agree) - complete for my needs
- valuable
- useful
- credible
Source Credibility The reviewers on this website are: Filieri et al.
(5-point, completely disagree - - credible 2015
completely agree) - experienced
- trustworthy
- reliable
Customer Satisfaction - I am satisfied with the information I have Filieri et al.
(5-point, completely disagree - received from this website 2015
completely agree) - I am satisfied with my previous experiences with
this website
Source: Author’s elaboration

30
Table 3. Items and scales employed in the questionnaire (Continuation)

Construct Items Source

Website Trust - I think that the information offered by this website Filieri et al. 2015 ,
(5-point, completely disagree - is sincere and honest Liu et al. 2013
completely agree) - I think that the advice and recommendations given
by the customer reviews are trustworthy
- I trust the online customer reviews on this website
- I trust this fashion website
Social Influence - People who are important to me think that I should Al-Qeisi et al.
(5-point, completely disagree - use online stores to shop for fashion products 2014
completely agree) - People who influence my behaviour think that I
should use online stores to shop for fashion
products
Performance Expectancy - I find fashion online stores useful Al-Qeisi et al.
(5-point, completely disagree - - Using fashion online stores enables me to get 2014
completely agree) fashion information more quickly
- Using fashion online stores increases the effective
use of my time in handling shopping tasks and
purchase
- Using fashion online stores increases the quality of
my fashion knowledge at minimal effort
Word-of-mouth Behaviour - I mentioned to others that I seek fashion Filieri et al.
(5-point, never - frequently) information from this website 2015
- I made sure that others know that I rely on this
website to purchase fashion products
- I spoke positively about this fashion website to
others
- I recommended this website to close friends
Recommendation Adoption - Online customer reviews and comments made it Filieri et al.
(5-point, completely disagree - easier for me to make a purchase decision (e.g., 2015
completely agree) purchase or not purchase)
- Online reviews have motivated me to make a
purchase decision (purchase or not purchase)
- The last time I read online fashion reviews I
adopted consumers' recommendations
- Information from customer reviews contributed to
my knowledge of fashion products and trends
Internet Experience How would you describe your: Al-Qeisi et al.
(5-point very bad - very good) - Internet knowledge 2014
- General computer knowledge
User Online Experience How would you rate: Filieri et al.
(5-point, not experienced at all - - Your level of experience in terms of using fashion 2015
very experienced) websites
- Your level of experience in terms of browsing
fashion websites
- Your level of experience in terms of online
customer reviews
Source: Author’s elaboration

The last part of the questionnaire contained socio-demographic variables.

31
3.3 Sample profile

The socio-demographic characteristics of the sample are presented in Table 4. The demographic
variables Employment, Age and Gender are split by Nationality.

The sample is composed by 312 respondents of which 251 females and 61 males (Appendix A).
The sample obtained was relatively young, since the mean age in years is 27 years old. This can
be justified by the fact that young segments are the target of most research in the fashion
consumption field, as young consumers have the interest to try new innovations (Law et al.,
2004).

The sample was primarily composed by females: the predominance of women in the sample
reflects the profile of online fashion shoppers in Italy and Portugal. Globally, females are more
fashion conscious than males and show higher levels of fashion clothing involvement (Hourigan
& Bougoure 2012).

Table 4. Socio-demographic characteristics of the respondents

Italian Portuguese
Gender Female 44.2% Female 36.2%
Male 14.4% Male 5.1%
Total 58.7% Total 41.3%
Age Mean 28.1 Mean 26.6
Standard deviation 9.810 Standard deviation 7.877
Employment status Employed 23.4% Employed 13.8%
Other 2.6% Other 1.3%
Self-employed 4.8% Self-employed 3.8%
Student 26.6% Student 19.9%
Unemployed 1.3% Unemployed 2.6%
Total 58.7% Total 41.3%
Source: Author’s elaboration based on SPSS data

Since the research is conducted between Italy and Portugal, the sample is composed of 183
Italian people, more precisely from the northern regions of Piedmont and Lombardy, and 129
Portuguese people, mostly from the region of Lisbon.

Few differences can be spotted between the Italian and Portuguese sample: the Italian group has a
more substantial percentage of males. Regarding the employment status the Italian respondents
have a higher percentage of the status Employed than the Portuguese. In both samples student is

32
the most popular employment category, suggesting that the questionnaire was answered mostly
by students of an age between 26 and 28 years old.

4. Data treatment and Results

4.1 Descriptive statistics

First each construct is analysed by calculating the mean and standard deviation of each item.
Besides, reliability is assessed for each construct with Cronbach's α coefficient. All the data was
treated using the software IBM SPSS Statistics 20.

Table 5. Descriptive statistics for Technical Quality (TQ)

Variable Mean Standard Deviation Cronbach’s Alpha

TQ1. This fashion website is easy to 4.2 0.859
use
TQ2. This fashion website has well 3.9 0.974
organized hyperlinks
TQ3. This fashion website provides 2.3 1.206
opportunities to interact with other
customers 0.796
TQ4. This fashion website has high 3.8 0.986
speed of page loading
TQ5. This fashion website is easily 4.0 0.992
accessible from different media
TQ6. This fashion 4.0 0.973
website guarantees users' privacy
Source: Author’s elaboration based on SPSS data

The variable Technical Quality shows substantially high values of the mean, suggesting that
respondents evaluate technical quality of the chosen website quite positively. Only the variable
TQ3 has a lower mean of 2.3: respondents may not find the opportunity to interact with other
customers on the preferred fashion website. The standard deviation for the construct Technical
Quality ranges from 0.973 to 1.206. The highest value for standard deviation is in WQ3. The
Cronbach’s Alpha coefficient has become by far the most popular measure of internal
consistency (Mooi & Sarstedt, 2011). It is generally agreed that values higher than 0.7 are
considered acceptable, therefore the value 0.796 for the construct of Technical Quality is
considered satisfactory, which means the items are suitable to describe the construct.

33
Table 6. Descriptive statistics for Product Information (PI)

Variable Mean Standard Deviation Cronbach’s Alpha

PI1. This fashion website shows all 3.9 1.150
the colours available for each
product
PI2. This fashion website shows all 4.0 1.125
the sizes available for each product
PI3. This fashion website tells the 4.4 0.817 0.850
price of products clearly
PI4. This fashion website gives up- 4.0 1.036
to-date information about products
and trends
PI5. This fashion website has good 4.0 1.050
quality photos of products
PI6. This fashion website truthfully 3.8 0.978
shows the colours of products
Source: Author’s elaboration based on SPSS data

The second multi-dimensional construct is Product Information. The table shows that the means
range between 3.8 and 4.0 suggesting that the participants agree with the statements. The lowest
mean belongs to PI6, with a value of 3.8. This might suggest that respondents do not trust
completely the chosen online store in terms of showing the products in an honest way. The
Cronbach’s alpha value is 0.850 which is higher than 0.7, therefore it’s considered acceptable.

Table 7. Descriptive statistics for Information Quality of customer reviews (IQ)

Variable Mean Standard Deviation Cronbach’s Alpha

IQ1. The information in the customer 3.2 1.091
reviews is timely
IQ2. The information in the customer 3.1 1.135
reviews is relevant to my needs
IQ3. The information in the customer 3.1 1.134
reviews is complete for my needs
IQ4. The information in the customer 3.1 1.133 0.945
reviews is valuable
IQ5. The information in the customer 3.3 1.130
reviews is useful
IQ6. The information in the customer 3.0 1.079
reviews is credible
Source: Author’s elaboration based on SPSS data

The third multi-dimensional construct (IQ) examines the quality of information provided by
customer reviews. The mean values are approximately 3, lower than in the previous two
constructs, thus respondents are more neutral to the statements. Particularly, IQ6 has the lowest

34
mean, suggesting that respondents do not think that the information provided by customer
reviews is credible. The standard deviation values are all higher than 1, while the Cronbach’s
Alpha is 0.945.

Table 8. Descriptive statistics for Source Credibility (SC)

Variable Mean Standard Deviation Cronbach’s Alpha

SC1. The reviewers on this fashion 3.2 1.075
website are credible
SC2. The reviewers on this fashion 2.7 1.058
website are experienced 0.935
SC3. The reviewers on this fashion 3.1 1.052
website are trustworthy
SC4. The reviewers on this fashion 3.0 1.043
website are reliable
Source: Author’s elaboration based on SPSS data

The construct Source Credibility describes the level of credibility and reliability of the sources
who write customer reviews on an online store. The credibility of reviewers is an increasingly
important matter, since fake and paid reviews have recently proliferated (Filieri et al., 2015). The
mean values vary between 2.7 and 3.2. Therefore, respondents are neutral to the statements and
do not show positive attitudes towards the sources of customer reviews. The standard deviation
values are slightly higher than 1. The Cronbach’s Alpha coefficient is 0.935, which is higher than
0.7 therefore suggesting that the items are very suitable to characterize the construct.

Table 9. Descriptive statistics for Customer Satisfaction (S)

Variable Mean Standard Deviation Cronbach’s Alpha

S1. I am satisfied with the fashion 4.0 0.935
information I have received from
this website 0.886
S2. I am satisfied with my previous 4.0 0.900
experiences with this website
Source: Author’s elaboration based on SPSS data

For the construct Customer Satisfaction two items were chosen. Both the mean values are 4,
therefore respondents judge positively the online store they chose and they are satisfied by the
information received. The standard deviation values are very similar; the reliability test results in
a Cronbach’s Alpha coefficient of 0.886, which is satisfactory.

35
Table 10. Descriptive statistics for Website Trust (T)

Variable Mean Standard Deviation Cronbach’s Alpha

T1. I think that the information 3.9 0.864
offered by this fashion website is
sincere and honest
T2. I think that the advice and 3.4 1.071
recommendations given by the 0.830
customer reviews are trustworthy
T3. I trust the online customer 3.3 1.121
reviews on this website
T4. I trust this fashion website 4.1 0.850
Source: Author’s elaboration based on SPSS data

The construct Website Trust investigates the extent to which participants trust different aspects of
the website, including customer reviews. The means of each item vary between 3.3 and 4.1. The
lowest mean values belong to the items T3 and T4, which pertain to customer reviews.
Respondents are slightly less willing to trust customer reviews than general product information,
in fashion online stores. Overall respondents trust the online store, since the mean value is 4.1
The values of standard deviation vary between 0.850 and 1.121. The reliability test shows a
Cronbach’s Alpha value of 0.830.

Table 11. Descriptive statistics for Social Influence (SI)

Variable Mean Standard Deviation Cronbach’s Alpha

SI1. People who are important to 2.7 1.171
me think that I should use online
stores to shop for fashion products
SI2. People who influence my 2.6 1.155 0.897
behaviour think I should use online
stores to shop for fashion products

Source: Author’s elaboration based on SPSS data

Social Influence measures the extent to which consumers believe that important others (e.g.,
family and friends) think they should use a particular technology; for this construct, the value of
the means is 2.7 and 2.6, suggesting that respondents tend to feel neutral or disagree with the
statements. The standard deviation values are slightly higher than one for both items. The
Cronbach’s Alpha value is 0.897 therefore it’s acceptable.

36
Table 12. Descriptive statistics for Performance Expectancy (PE)

Variable Mean Standard Deviation Cronbach’s Alpha

PE1. I find fashion online stores 4.0 1.016
useful
PE2. Using fashion online stores 4.0 1.032
enables me to get fashion
information more quickly 0.884
PE3. Using fashion online stores 3.8 1.125
increases the effective use of my
time in handling my shopping tasks
and purchase
PE4. Using fashion online stores 3.7 1.102
increases the quality of my fashion
knowledge at minimal effort
Source: Author’s elaboration based on SPSS data

For the construct Performance Expectancy (PE) four items were introduced. The values of the
means for this construct range between 3.7 and 4, therefore respondents, on average, agree that
online stores are useful and enable them to retrieve fashion information quickly. Fashion online
stores are in general considered helpful in performing shopping tasks by respondents. The
standard deviation values are higher than 1 for all four items, while the Cronbach’s Alpha
coefficient is 0.884.

Table 13. Descriptive statistics for Word-of-mouth Behaviour (WM)

Variable Mean Standard Deviation Cronbach’s Alpha

WM1. I mentioned to others that I 3.2 1.289
seek fashion information from
this website
WM2. I made sure that others know 2.9 1.258
that I rely on this website to 0.867
purchase fashion products
WM3. I spoke positively about this 3.7 1.070
fashion website to others
WM4. I recommended this website 3.6 1.178
to close friends
Source: Author’s elaboration based on SPSS data

For the construct Word-of-mouth Behaviour, the mean values are included between 2.9 and 3.7.
For this construct a value of 1 indicates Never, while 5 indicates Frequently. Therefore
respondents are slightly willing to spread positive word-of-mouth about fashion websites,
however, on average, they don’t do it frequently. The standard deviation values are higher than 1
for all four items, while the Cronbach’s Alpha coefficient is higher than 0.7.

37
Table 14. Descriptive statistics for Recommendation Adoption (REC)

Variable Mean Standard Deviation Cronbach’s Alpha

REC1. Online customer reviews and 2.9 1.226
comments made it easier for me to
make a purchase decision (e.g.,
purchase or not purchase)
REC2. Online reviews have motivated 2.8 1.161
me to make a purchase decision
(purchase or not purchase) 0.918
REC3. The last time I read online 2.7 1.170
fashion reviews I adopted consumers'
recommendations
REC4. Information from customer 2.8 1.164
reviews contributed to my knowledge
of fashion products and trends
Source: Author’s elaboration based on SPSS data

For the construct Recommendation Adoption 4 items are considered. The values of each item’s
mean are quite similar, ranging from 2.7 to 2.9. The mean suggests that respondents slightly
disagree with the statements, therefore they usually do not follow the recommendations from
other customers on fashion online stores. The standard deviation values are higher than 1 for all
items. The Cronbach’s Alpha coefficient is 0.918 and therefore acceptable.

Table 15. Descriptive statistics for Internet Experience (IEX)

Variable Mean Standard Deviation Cronbach’s Alpha

IEX1. How would you describe 4.1 0.883
your Internet knowledge
IEX2. How would you describe 3.9 0.955 0.904
your general computer knowledge
Source: Author’s elaboration based on SPSS data

For the construct Internet Experience respondents could rate their Internet and computer skills
from 1 to 5, 1 being very bad and 5 being very good. The mean values are approximately 4 for
both items, thus, on average, respondents have a quite good knowledge of the Internet and
computers. Specifically, they seem more experienced in using the Internet than computers in
general. The standard deviation values are lower than 1, being 0.883 and 0.955, while the
reliability test results in a coefficient of 0.904 which is satisfactory.

38
Table 16. Descriptive statistics for User Online Experience (UEX)

Variable Mean Standard Deviation Cronbach’s Alpha

UEX1. How would you rate your 3.4 1.121
level of experience in terms of using
fashion websites
UEX2. How would you rate your 3.6 1.109 0.839
level of experience in terms of
browsing fashion websites
UEX3. How would you rate your 3.0 1.100
level of experience in terms of
online customer reviews
Source: Author’s elaboration based on SPSS data

The construct User Online Experience concerns the familiarity of users with fashion online stores
and online customer reviews. The respondents could score their experience from 1-very bad to 5-
very good. The mean of UEX3 is the lowest, suggesting that respondents are not very
experienced with online customer reviews. However they are slightly more experienced with
using fashion websites to complete purchases and, on average, they rate their level of experience
higher in browsing fashion websites, without necessarily completing a purchase. The standard
deviation values are slightly higher than 1 in all three items. The Cronbach’s Alpha coefficient is
0.839, which can be considered acceptable.

4.2 Regression Analysis results

Regression analysis is one of the most frequently used tools in market research and it allows
market researchers to analyse relationships between dependent and independent variables to
predict a certain outcome (Mooi & Sarstedt, 2011). Therefore, multiple regression analyses were
conducted to test the hypotheses previously formed. The conceptual model was developed so that
the variable Performance Expectancy acts as a mediator between perceived website quality
elements and behavioural outcomes, specifically, Website Trust, Customer Satisfaction,
Recommendation Adoption and Word-of-mouth Behaviour.

Tolerance and VIF values were tested for all the following regressions. VIF values are inferior to
10, and tolerance is above 0.1, indicating that there are not collinearity issues. The results
obtained from the regressions are presented below. Autocorrelation was also checked before
interpreting the results.

39
4.2.1 Determinants of Performance Expectancy
The first part of the model investigates the influence of Technical Quality, Product Information,
Information Quality of customer reviews, Social Influence and Source Credibility on
Performance Expectancy (Appendix B).

Table 17: Determinants of Performance Expectancy

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant)
0.516 0.236 2.182 0.030
Tecquality
0.217 0.075 0.166 2.908 0.004 0.558 1.791
0.431
ProductInf
0.479 0.065 0.406 7.427 0.000 0.613 1.632
Infquality -0.058 0.065 -0.063 -0.904 0.367 0.382 2.620
Sourcecred
0.072 0.065 0.075 1.106 0.269 0.396 2.526
SocInfluence
0.225 0.038 0.270 5.983 0.000 0.896 1.116
Source: Author’s elaboration based on SPSS data

Table 17 shows that Information Quality of customer reviews and Source Credibility are not
relevant and don’t have a direct effect on Performance Expectancy (p values ˃ 0.05). The
standardized β coefficients allow us to compare the relative effect of each independent variable.
The variable with the highest absolute value is Product Information. The unstandardized effect of
Product Information suggests that the effect of a single increase in product information quality
increases perceived usefulness by an average of 0.479.

The results prove that, contrary to our predictions, the information provided in customer reviews
does not increase the degree to which participants find online fashion shopping useful. Similarly,
the credibility of the reviewers is not important to explain performance expectancy. On the other
hand, technical website quality, product-related information and social influences can be accepted
as predictors of perceived usefulness.

40
The second group of regressions investigates the role of users’ experience with the Internet and
online shopping. Research by social psychology suggests that experience with an attitude object
influences indirectly behaviour (Al-Qeisi et al., 2014). Therefore individuals who have higher
computer skills might form more positive perceptions of usefulness of a website. Similarly, the
user’s expertise in browsing websites and using customer reviews might have an impact on
perceived usefulness.

Table 18: Influence of experience on Performance Expectancy

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant) 1.731 0.201 8.608 0.000

Intexperience 0.344
0.130 0.061 0.124 2.135 0.034 0.627 1.595
Userexperience 0.483 0.055 0.506 8.728 0.000 0.627 1.595
Source: Author’s elaboration based on SPSS data

The results confirm that both types of experience are relevant to explain Performance
Expectancy. The standardized coefficients are both positive, therefore the two variables influence
positively the dependent variable. Looking at the absolute value, User Online Experience has a
higher significance in explaining the dependent variable than Internet Experience. After
establishing the significant role of experience, all the variables influencing Performance
Expectancy are considered simultaneously, as seen in figure 4.

41
 Figure 4. Determinants of Performance Expectancy

Source: Author’s elaboration.

Table 19. Determinants of Performance Expectancy together with experience

Unstandardized Standardized t Sig. Adjusted Collinearity statistics

Coefficients Coefficients R square

B Std. Error Beta Tolerance VIF

(Constant)
0.347 0.235 1.476 0.141
Tecquality
0.161 0.071 0.123 2.263 0.024 0.534 1.874
0.507
ProductInf
0.351 0.063 0.298 5.576 0.000 0.557 1.795
Infquality
-0.051 0.061 -0.055 -0.839 0.402 0.373 2.681
Sourcecred
0.015 0.061 0.015 0.241 0.810 0.388 2.575

SocInfluence
0.187 0.036 0.224 5.258 0.000 0.871 1.148
Intexperience
0.032 0.055 0.031 0.583 0.561 0.573 1.746
Userexperience
0.301 0.053 0.316 5.716 0.000 0.521 1.921

Source: Author’s elaboration based on SPSS data

42
The adjusted R2 of this model is 0.507, which is higher than the previous two models, having the
best fit so far: 51.8 % of the variation of Performance Expectancy is explained by the variables in
the model.

The t-tests show that, as expected, Information Quality of customer reviews and Source
Credibility don’t explain the dependent variable (p values ˃ 0.05). Also, in the presence of the
other variables, Internet Experience becomes irrelevant to explain the dependent variable. The
variables that should be kept in the model are Technical Quality, Product Information, Social
Influence and User Online Experience, since they are all positively related to Performance
Expectancy. Therefore H1a and H1b are supported, while H1c, regarding the customer reviews
has to be rejected (p=0.402˃0.05). Besides, only Social Influence is relevant in terms of peers
recommendation, while the influence of reviewers is not important to explain Performance
Expectancy. Thus, H2a is supported while H2b is rejected. Concerning the role of experience, H3
can be accepted because previous experience with online shopping affects the extent to which
users find online shopping useful; however Internet and computer skills are not relevant when
other variables are present (p=0.561).

Splitting the sample by nationality we can detect some differences between the Italian and the
Portuguese respondents. For the Portuguese sample, the model has an Adjusted R2 of 0.348 (see
Appendix C). The variable Product Information is not significant for the Portuguese respondents
(p value=0.064 ˃ 0.05). Only Technical Quality, Social Influence and User Online Experience
explain Performance Expectancy. User Online Experience has the standardized β coefficient with
the highest absolute value (0.348), so it might be argued that, for these shoppers, the proficiency
in using fashion websites is the factor that determines their perception of usefulness the most.

For the Italian sample, the model explains 51.4 % of the variation of Performance Expectancy
(see Appendix D). Differently from the Portuguese sample, for these users Product Information is
relevant. On the contrary, Technical Quality becomes irrelevant (p value= 0.120˃ 0.05)
suggesting that the Italian respondents evaluate a fashion online store based on the information
provided rather than the technical quality of the website. Social Influence seems to have a slightly
stronger relevance for Portuguese respondents (β= 0.317) than for Italians (β= 0.213) .

43
4.2.2 The moderating impact of previous online experience on perceived website quality
As mentioned in the literature, experience is expected to influence not only performance
expectancy but also website quality perceptions. The following set of regressions analyses how
both types of experience affect perceived website quality, as seen in figure 5.
Figure 5. Impact of experience on perceived website quality

Source: Author’s elaboration

The results suggest that experience has a direct impact on perceived quality, specifically on
Technical Quality, Product Information and Information Quality of customer reviews.

Table 20. Experience influences Technical Quality

Unstandardized Standardized t Sig. Adjusted Collinearity
Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant)
2.303 0.171 13.474 0.000
Intexperience 0.194
0.190 0.052 0.237 3.683 0.000 0.627 1.595
Userexperience
0.191 0.047 0.261 4.053 0.000 0.627 1.595
Source: Author’s elaboration based on SPSS data

Table 20 shows that experience has a significant impact on Technical Quality perceptions.
Looking at the coefficients table (β= 0.237 for Internet Experience and β=0.261 for User Online

44
Experience), both types of experience have a positive impact on Technical Quality. Thus, users
who have more extensive experience in the use of Internet perceive higher technical quality of a
fashion retailer’s website. H4a is supported.

Table 21. Experience influences Product Information

Unstandardized Standardized t Sig. Adjusted Collinearity
Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant) 2.299 0.181 12.678 0.000

Intexperience 0.256
0.190 0.055 0.214 3.463 0.001 0.627 1.595
Userexperience
0.284 0.050 0.351 5.683 0.000 0.627 1.595
Source: Author’s elaboration based on SPSS data

Similarly, experience can affect the perceived quality of product information. The coefficient
table shows that both Internet Experience and User Online Experience (p values ˂ 0.05) are
useful to explain the dependent variable. The proficiency with browsing fashion websites and the
users’ Internet skills influence the perceptions of information quality on an online store. H4b is
supported.

Table 22. Experience influences Information Quality of customer reviews

Unstandardized Standardized t Sig. Adjusted Collinearity
Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant)
2.326 0.254 9.152 0.000
0.090
Intexperience
-0.111 0.077 -0.099 -1.442 0.150 0.627 1.595
Userexperience
0.369 0.070 0.360 5.274 0.000 0.627 1.595
Source: Author’s elaboration based on SPSS data

45
In terms of Information Quality of customer reviews, it can be expected that the more users are
experienced in using reviews the more they find this type of information useful. Similarly, users
who are not familiar with customer ratings and reviews, may not be able to establish the quality
of the information provided. As table 21 shows, only User Online Experience influences
Information Quality of customer reviews. Internet Experience is not significant in this case (p
value= 0.150˃ 0.05). Therefore H3c is partially accepted, since in this case Internet Experience
is not significant.

For the Portuguese participants, only User Online Experience seems significant in influencing
perceived website quality (Appendix C). Internet Experience is not relevant to explain Technical
Quality (p value= 0.566) nor Product Information (p value= 0.635).

Italian respondents, on the other hand, are more influenced by Internet experience and computer
skills when they evaluate overall website quality. Specifically, Internet Experience has the
strongest effect on Technical Quality (β= 0.328) and on Product Information (β= 0.327). The
variable Information Quality of customer reviews has a different relationship with experience:
only User Online Experience has a predictive power, while Internet Experience is irrelevant.

Overall, H4a H4b and H4c hold because at least one type of experience influences shoppers in
their perceptions of a website technical adequacy, product-related information and customer
reviews quality.

4.2.3 Relationship outcomes of website quality

The degree to which a shopper finds online stores useful and time saving could influence the
level of customer satisfaction, website trust and the probability of spreading word-of-mouth.
Besides, if consumers find a website useful and have the possibility to read customer reviews on
it, they might adopt the recommendations in those reviews. Figure 6 shows the remaining part of
the model and the relationships between Performance Expectancy and its behavioural outcomes.

46
 Figure 6. Relationship outcomes

Source: Author’s elaboration

Table 23. Performance Expectancy and Customer Satisfaction

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
2.015 0.177 11.356 0.000
0.308
Perfexpectancy
0.526 0.045 0.557 11.802 0.000 1.000 1.000
Source: Author’s elaboration based on SPSS data

Table 23 presents the results obtained from a simple linear regression. The findings confirm that
Performance Expectancy predicts Customer Satisfaction (p value = 0.000). Since β= 0.557 ,
perceived usefulness of fashion online stores is positively associated to users’ satisfaction. In
particular, customers are satisfied with the information provided if the website enables them to
improve their fashion knowledge at minimal effort. Thus, H5a is supported.

47
Table 24. Performance Expectancy and Website Trust

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
2.088 0.173 12.043 0.000
0.220
Perfexpectancy
0.410 0.044 0.472 9.420 0.000 1.000 1.000
Source: Author’s elaboration based on SPSS data

Similarly, Performance Expectancy predicts Website Trust. The results confirm that Performance
Expectancy is significant in explaining Website Trust. Therefore, if online shoppers believe
fashion websites are useful they might build a positive feeling of trust towards any particular
online store. H5b is confirmed.

Table 25. Performance Expectancy and Recommendation Adoption

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
1.911 0.254 7.511 0.000
0.038
Perfexpectancy
0.232 0.064 0.202 3.626 0.000 1.000 1.000
Source: Author’s elaboration based on SPSS data

Recommendation Adoption is positively influenced by Performance Expectancy, supporting H5e.

Shoppers tend to adopt customer recommendations if they positively evaluate the activity of
online shopping for fashion products.

48
Table 26. Performance Expectancy and WOM Behaviour

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
1.223 0.215 5.674 0.000
0.252
Perfexpectancy
0.556 0.054 0.504 10.274 0.000 1.000 1.000
Source: Author’s elaboration based on SPSS data

The influence of Performance Expectancy on WOM Behaviour is confirmed, as shown in the

table above. Customers are more likely to mention the fashion website they use to their peers if
they perceive online shopping as a useful activity. H5f is supported.

For the Portuguese sample the results are confirmed: Performance Expectancy predicts Customer
Satisfaction (β= 0.340), Website Trust (β= 0.273) and WOM Behaviour (β= 0.320). However,
Recommendation Adoption is not predicted by Performance Expectancy (p value=0.907˃ 0.05).

Similarly, for the Italian group, Performance Expectancy has a significant influence on Customer
Satisfaction (β= 0.613), Website Trust (β= 0.524), WOM Behaviour (β= 0.597) and
Recommendation Adoption (β=0.312). The effect of Performance Expectancy on relationship-
based outcomes is slightly stronger for the Italian sample, suggesting that Italian respondents may
be more influenced by perceived usefulness and convenience of online shopping than Portuguese
ones.

Al-Qeisi et al. (2014) found evidence that website design quality has an impact on usage
behaviour, both directly and indirectly and this impact is greater than the one of performance
expectancy. Similarly, in this context, the direct effect of perceived website quality on
relationship-based outcomes can be tested. The variable Perceived Website Quality
(PERCEIVEDQUALITY) is obtained by computing the three variables Technical Quality,
Product Information and Information Quality of customer reviews together. The results are
presented in tables 25, 26, 27.

49
There is evidence that Perceived Website Quality is directly affecting relationship-based
constructs. In fact, Perceived Website Quality has a strong positive relationship with Customer
Satisfaction, Website Trust, Recommendation Adoption and WOM Behaviour.

Table 27. Perceived Website Quality and Customer Satisfaction

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant) 0.652 0.198 3.290 0.001
0.494
PERCEIVED 0.941 0.054 0.704 17.448 0.000
1.000 1.000
QUALITY
Source: Author’s elaboration based on SPSS data

Table 28. Perceived Website Quality and Website Trust

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
0.211 0.161 1.314 0.190
0.606
PERCEIVED
0.959 0.044 0.779 21.890 0.000 1.000 1.000
QUALITY
Source: Author’s elaboration based on SPSS data

Table 29. Perceived Website Quality and WOM Behaviour

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant) 0.445 0.278 1.600 0.111 0.268

PERCEIVED 0.811 0.076 0.519 10.704 0.000

1.000 1.000
QUALITY
Source: Author’s elaboration based on SPSS data

50
Table 30. Perceived Website Quality and Recommendation Adoption

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant) 0.166 0.303 0.547 0.585
0.200
PERCEIVED 0.731 0.082 0.450 8.868 0.000
1.000 1.000
QUALITY
Source: Author’s elaboration based on SPSS data

The direct impact of Perceived Website Quality on behaviour is confirmed by the results obtained
from the Portuguese and Italian sample separately. For the Portuguese sample, Perceived Website
Quality is stronger than Performance Expectancy in predicting Customer Satisfaction, Website
Trust and WOM Behaviour; Perceived Website Quality also predicts Recommendation Adoption
(p value=0.000˂ 0.05). Similarly, for the Italian sample, Perceived Website Quality has a strong
positive impact on Customer Satisfaction, Website Trust, WOM Behaviour and Recommendation
Adoption.

Website Trust and Customer Satisfaction are said to be highly related (Yoon, 2002). Thus,
Customer Satisfaction, in addition to Performance Expectancy, is expected to influence Website
Trust, which in turn has a positive impact on Recommendation Adoption.

Table 31. Online drivers of Website Trust

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant)
0.964 0.170 5.680 0000
0.473
Perfexpectancy 0.117 0.043 0.134 2.707 0.007 0.690 1.449

Satisfaction 0.558 0.046 0.606 12.230 0.000 0.690 1.449

Source: Author’s elaboration based on SPSS data

51
Table 31 shows the results obtained. Both Customer Satisfaction and Performance Expectancy
predict Website Trust. Customer Satisfaction has the strongest effect (β= 0.606). The model has a
satisfying fit since, 47.6 % of the variation of Website Trust is explained by the explanatory
variables in the model. Therefore, if users are satisfied with the fashion information provided by
any particular website, they tend to build a feeling of trust towards it: H5c is supported.

By splitting the sample, we notice that Performance Expectancy is not as powerful as Customer
Satisfaction in explaining Website Trust. For the Portuguese sample, Customer Satisfaction
positively affects Website Trust (β = 0.504); the same can be said for the Italian subgroup where
Website Trust is strongly explained by Customer Satisfaction (β= 0.657).

Information adoption is the process by which people purposefully engage in using information
(Filieri et al., 2015). If users believe a website is reliable, they might adopt the information
provided by its customer reviews. Specifically, if the reviews are perceived as trustworthy, users
may follow the recommendations given by other customers to take more informed purchase
decisions.

Table 32. Website Trust and Recommendation Adoption

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Beta Tolerance VIF

Error
(Constant)
0.675 0.254 2.660 0.008
0.190
Trust
0.580 0.067 0.439 8.610 0.000 1.000 1.000
Source: Author’s elaboration based on SPSS data

The Coefficients table shows that Website Trust is a predictor of Recommendation Adoption,
thus H5d is accepted. The hypothesis holds for the Portuguese and Italian groups considered
separately as well.

52
Lastly, the drivers of WOM Behaviour are investigated. The results show that Performance
Expectancy, Customer Satisfaction and Website Trust are all accepted to explain the dependent
variable (Appendix B). This model has an Adjusted R2 of 0.370 and explains 37.6 % of WOM
Behaviour. However, Recommendation Adoption could be added to the model, since it is an
outcome of Website Trust. The model obtained by adding the mentioned variable has a better fit
(Adjusted R2= 0.411), and explains 41.8% of the variable WOM Behaviour (Table 33). By
adding Recommendation Adoption to the model, Website Trust becomes non-significant (p
value= 0.102˃ 0.05). Therefore, Website Trust influences WOM Behaviour positively, with the
mediation of Recommendation Adoption. In fact, if online shoppers trust a fashion website which
provides high quality customer reviews, they might be more willing to mention the website to
others. Thus, H6a, H6b and H6c are supported. H6b claims that there is a significant positive
effect of Website Trust on WOM Behaviour. This is confirmed only when the variable
Recommendation Adoption is not included.

Table 33. Performance Expectancy, Customer Satisfaction, Website Trust and Recommendation
Adoption as drivers of WOM Behaviour

Unstandardized Standardized t Sig. Adjusted Collinearity

Coefficients Coefficients R square statistics

B Std. Error Beta Tolerance VIF

(Constant)
-0.159 0.244 -0.653 0.514
0.411
Satisfaction 0.323 0.077 0.277 4.190 0.000 0.435 2.300

Perfexpectancy 0.275 0.059 0.250 4.689 0.000 0.669 1.496

Trust 0.141 0.086 0.111 1.641 0.102 0.413 2.423

Recadoption 0.228 0.048 0.238 4.748 0.000 0.755 1.325

Source: Author’s elaboration based on SPSS data

The Portuguese sample confirms the obtained results: in the presence of Recommendation
Adoption, Website Trust is not significant to explain WOM Behaviour (p value= 0.263).
Similarly to the whole sample results, the model with a better fit is the one which includes

53
Recommendation Adoption (Adjusted R2 = 0.271). For this sample, Customer Satisfaction has
strongest unstandardized effect on WOM Behaviour.
For the Italian sample, the model with the best fit is the one including Recommendation
Adoption, as it shows an Adjusted R2 = 0.494. This model explains 50.5% of the variation of
WOM Behaviour. Again, Website Trust seems to influence WOM Behaviour indirectly, through
Recommendation Adoption. For Italian participants, Performance Expectancy has the strongest
relationship with WOM Behaviour (β= 0.296).

4.4 One-way ANOVA and non-parametric tests

In the beginning of the questionnaire, respondents were asked to indicate the brand of their
favourite fashion online store. Three retailers emerged as the most popular choice among
respondents: Zara, Amazon and Asos. In this perspective, One-way ANOVA is used to determine
whether there is reason to believe that the population means of the groups determined by brand
differ significantly (Mooi & Sarstedt, 2011).
Since satisfaction has been described as a critical success factor for e-tailers, it may be useful to
understand if the three brands are perceived as equal in providing a satisfying amount of fashion
information and experience. The Levene’s test (Appendix E) shows a sig.= 0.569 ˃0.05, thus the
assumption of equality of variances is fulfilled. In the ANOVA table, the p value is lower than
0.05, therefore we reject the null hypothesis and conclude that the mean for Customer
Satisfaction is not the same for the three groups defined by e-tailer brand. In the multiple
comparisons table the p-values for Zara-Amazon and Asos-Amazon are lower than 0.05, meaning
that the equality of means is rejected. In fact, by obtaining a means plot (Figure 7) it is visible
that Amazon has the lowest mean for the variable Customer Satisfaction.

54
 Figure 7. Means plot for Customer Satisfaction

Source: Author’s elaboration based on SPSS data

Satisfaction is found to be strongly related to Website Trust, therefore it may be interesting to

understand which e-tailer website is perceived as the most trustworthy. The Levene’s test holds
the assumption of equality of variances (sig=0.551), however the ANOVA table shows a
sig.˃0.05 indicating that the means for the three defining groups do not differ significantly.
Similarly, for the variable WOM Behaviour, the ANOVA results show that there is not a
significant difference of means for the three brands (sig.= 0.053). However, looking at the means
plots for WOM Behaviour, Amazon seems to have a slightly lower score than Asos, which is the
brand with the highest mean for WOM behaviour.

Figure 8. Means plot for WOM behaviour

Source: Author’s elaboration based on SPSS data

55
The regression analyses proved that product information is an important element to consider in
establishing the quality of a website, as it has a strong positive impact on performance
expectancy. However, for the variable Product Information, One-way ANOVA could not be
carried, since the Levene’s test has a sig.˂ 0.05. Alternatively, a Kruskal-Wallis test can be
chosen to compare the equality of distribution of the variable Product Information for the three
brands. In the Test statistics table, sig.=0.000˂ 0.05. Therefore, the distribution of Product
Information is different for at least one brand. Looking at the sample mean ranks, there is
evidence that Zara has the highest quality of product information, while Amazon has the lowest
(Figure 8).

Figure 9. Means plot for Product Information

Source: Author’s elaboration based on SPSS data

For Technical Quality, the sig. level is higher than 0.05, therefore there is not a significant
difference in how the three online stores are perceived in terms of technical features such as
design, speed, accessibility, usability and security.

5. Conclusions
5.1 Discussion of results

This study was intended to accomplish two main objectives. The first objective was to adapt and
extend the UTAUT model in the context of online shopping for fashion clothing. The objective
has been accomplished, since several constructs of the UTAUT were tested. The study finds
evidence that performance expectancy is predicted both by website quality elements (facilitating

56
conditions) and social factors (social influence). Consistently with the results obtained by Al-
Qeisi et al. (2014), website quality elements influence positively performance expectancy;
however Al-Qeisi et al. (2014) found that the impact of social influences was not significant. The
relevance of social influences is supported by research from Celik (2011), who proved that social
factors have a direct effect on performance expectancy, in the context of online shopping in
Turkey. Indeed, the opinion of internal sources (e.g. family and friends) about the usefulness of
online shopping is perceived by consumers as personally meaningful, supporting the results of
this study. Nevertheless, the role of social influence can become irrelevant, according to Zhang et
al. (2006), when consumers are confident about purchasing online and become less influenced by
others. This study reveals that Portuguese and Italians are still not completely confident about
purchasing fashion clothing online and are influenced by internal sources.

The results point out that performance expectancy has a positive impact on online behaviour, in
line with findings from Yang (2010), who analysed the role of this construct in the context of
mobile shopping services. In the UTAUT, the outcome of performance expectancy is usage of
technology. This dissertation extends the concept of usage to relationship-based outcomes and
applies them to the context of online shopping for fashion clothing. In fact, according to the
results, performance expectancy has a positive impact on website trust, customer satisfaction and
WOM behaviour. Additionally, website trust leads to the adoption of customer recommendations
and to WOM behaviour, similarly to findings from Filieri et al. (2015) in the travel and tourism
industry.

Furthermore, the present study confirms the importance of experience in influencing behaviour,
providing support to the UTAUT model. Internet experience and user online experience have a
positive impact on performance expectancy and on perceived website quality. This result
reinforces previous research from Broekhuizen and Huizingh (2009), in affirming that online
shopping experience, a reflection of a consumer’s familiarity with shopping on websites,
influences online attitudes and subsequent behaviour. Moreover, another study by Kwon and Noh
(2010) found that the level of past online shopping experience predicts consumers’ perceptions of
online shopping.

A secondary objective was to compare the adapted model in two regions of Portugal and Italy. In
evaluating the usefulness of online shopping for fashion clothing, Italian respondents seem more

57
influenced by the quality of product-related information, while Portuguese respondents do not
show the same behaviour. In fact, this group is more affected by the technical performance of
fashion websites rather than by the information provided. Both groups show a quite strong
relationship between social influence and performance expectancy. Additionally, the perceived
usefulness of fashion online shopping is strongly mediated by user online experience, for both
samples.

Italian participants seem to be more influenced by performance expectancy in their online

shopping behaviour than Portuguese ones. Thus, Italians may develop positive feelings towards
fashion clothing websites if they believe the activity of online shopping is convenient and useful.
Italy scores much higher than Portugal in the Hofstede cultural dimension of individualism.
Hence, Italian shoppers might be more responsive to personal convenience and adopt innovations
that allow them to gain personal benefits. According to Solomon et al. (2006), a study of 11
European countries proved that consumers in individualistic cultures are more innovative than
consumers in collective cultures. Online shopping for clothing can be seen as a quite innovative
practice both in Italy and Portugal. The same can be said for the customer reviews phenomenon;
the results show that for Portuguese respondents, recommendation adoption is not related to
performance expectancy, but only to website trust. In other words, Portuguese shoppers might
adopt the online recommendations from other customers only if they fully trust the website
hosting consumer-generated content. In contrast, for Italians the path Performance Expectancy-
Recommendation Adoption is accepted, suggesting that Italians adopt the online
recommendations from other customers when they evaluate positively the activity of online
shopping and they do not rely on a specific website reputation. Portugal is a high uncertainty
avoidance culture, justifying the fact that Portuguese shoppers might be slightly more reluctant to
adopt recommendations from other customers, in the online environment.

Another goal of this research was to analyse the role of perceived website quality in generating
customer satisfaction, website trust and WOM behaviour. In this context, Perceived Website
Quality is considered as a multi-dimensional construct, determined by Technical Quality, Product
Information and Information Quality of customer reviews. The regression analyses conducted
show that overall website quality has a stronger impact on behaviour than performance
expectancy. Website quality is the most important driver of customer satisfaction, website trust

58
and WOM behaviour in the online environment. This result is supported by research from O’Cass
and Carlson (2012), who proved that website service quality is a driver of loyalty and word-of-
mouth. Besides, Flavián et al. (2006) refer that the usability of a specific website has a positive
influence on customer satisfaction, and that customer satisfaction leads to trust towards the same
website. Moreover, perceived website quality has a positive effect on recommendation adoption,
in line with findings from Filieri et al. (2015).

Finally, this dissertation aimed at exploring whether online shoppers are or are not influenced by
peer customer reviews, in buying experience goods such as fashion clothing. Since H1c and H2b
were rejected, online reviews do not have a significant impact on the perceived usefulness of
using the Internet to shop for fashion clothing. Similarly, those customers who write online
product reviews are not perceived as social influences by respondents. The results show that for
both Italian and Portuguese respondents, customer reviews are not directly associated with
perceived usefulness. It seems that the extent to which respondents believe that online shopping
for clothing is useful and convenient is not determined by the quality of customer reviews.
However, as a component of overall website quality, customer reviews may have a positive
impact on customer satisfaction, website trust and WOM behaviour. Moreover, if the quality of
information contained in online reviews is excellent, users may be encouraged to adopt the given
advice. The findings confirm that the adoption of other customer recommendations is very likely
to influence WOM behaviour. In fact, shoppers might be motivated to spread positive word-of-
mouth about any particular fashion website, if their experience with the customer reviews was
satisfying.

5.2 Managerial implications

Companies in the fashion industry are increasingly extending their presence from the offline to
the online environment (Salonen et al., 2014). The study findings offer important implications
for fashion businesses to develop their online stores more effectively.

First, managers of fashion brands should design their websites so that the users perceive a high
quality, both technical (speed of page loading, well-designed webpages, accessibility, security)
and in terms of content. Specifically, product information is a very valuable element for a website
selling fashion clothing. The absence of physical access to products in online store environments

59
increases risk perceptions and the anxiety levels of online shoppers (Celik, 2016). Therefore
managers need to provide consumers with rich and relevant product-related information.
Information given by sellers should satisfy the utilitarian needs of shoppers, by describing in
detail sizes, colours and other technical characteristics of each product. However, according to
McCormick and Livett (2012), e-tailers should also supply information that stimulates shoppers’
hedonic needs, in the form of visual information, videos, trend updates, style advice and fashion
inspiration.

Because of the “high touch” nature of fashion clothing (Levin et al., 2005), consumers use
different sources of information to compensate the lack of sensorial cues. For example, customer
reviews can be helpful to learn about products and to reduce uncertainty (Purnawirawan et al.,
2015). The existence of such reviews represents a challenge for marketers, as it represents a
significant loss of control over what is said about the brand and its products (De Maeyer, 2012).
According to this study, online customer reviews do not significantly enhance the usefulness of
fashion online shopping, from a consumer’s perspective. This conclusion is applicable only in the
context of Italy and Portugal, which are two countries with a high tendency of risk avoidance. In
countries with a lower level of uncertainty avoidance, customer reviews could be more suitable.

Another obstacle to implementing customer reviews for marketing managers is the quantity of
information provided. Online customer reviews could potentially have negative effects on the
decision-making process, if the amount of information presented in the reviews is too
overwhelming to process or the opinions expressed are inconsistent (Lee & Ma, 2012). Thus,
fashion websites hosting an excessive amount of customer reviews might be perceived as less
useful and helpful. In this research, the three most popular websites among participants were
Zara.com, Amazon.com and Asos.com. Of these, only Amazon provides customer reviews for
each product. Despite hosting a conspicuous amount of reviews, Amazon seems to be less
appreciated by respondents in terms of satisfaction and product information, compared to Zara
and Asos. This fact stresses the importance of monitoring the quality and quantity of customer
reviews on websites selling fashion clothing.

Before implementing customer reviews, it is useful to consider some aspects of the target
audience, such as cultural factors. Smith et al. (2013) suggest that there is not a single theoretical
approach that fully captures the richness of cultural differences. Actually, in cultures with high

60
uncertainty avoidance such as Portugal, online marketers should provide information capable of
lowering risk perceptions and offer incentives such as easy returns to reassure shoppers about
their purchases.

Furthermore, social influence is relevant to explain perceived usefulness, both for the Portuguese
and Italian sample. Thus, e-tailers should actively seek out reference groups (e.g. friends and
celebrities) to create awareness about the usefulness and quality of their websites. Specifically in
situations where consumers do not feel very confident in purchasing online, marketers could rely
on opinion leaders, such as bloggers, to create positive perceptions of online shopping.

This study implies that online marketers should monitor perceived website quality, because this
construct has the strongest effect on satisfaction, trust and word-of-mouth. Firms can contribute
to the achievement of organizational objectives through careful and creative management of
website design quality, which has a significant influence on relationship-based constructs.

In sum, we highlight three main recommendations for managers of pure click and brick-and-click
companies:

• When designing a website selling fashion clothing, managers should guarantee an

adequate technical quality and, at the same time, provide the right amount of product-
related information (e.g. sizes, colours, high quality photos) and fashion information
(trends, style advice, fashion inspiration). Product-related information is one of the main
dimensions of website quality and it can contribute to create strong relationships with
customers, by increasing website trust, customer satisfaction and word-of-mouth.
• When implementing customer reviews on their websites, managers need to carefully plan
the amount of product reviews and the quantity of information provided by each review.
In fact, if the information contained in customer reviews is overwhelming or
contradicting, this could affect negatively the decision-making process and discourage
shoppers from staying on the website.
• When selling clothing online across different countries, managers should tailor their
websites to better satisfy the local audience. Specifically in countries where online
shopping is still a quite innovative practice (e.g. Italy and Portugal), e-tailers should
create awareness about the usefulness and convenience of their websites. Marketers can

61
 also take advantage of opinion leaders, such as fashion bloggers, to reassure shoppers
about the advantages and benefits of online shopping for fashion clothing.

5.3 Limitations and future research

Like any research, findings must be considered in light of key limitations. First, the respondents
were mostly from the region of Lisbon for Portugal and Northern Italy, therefore they might not
be representative of the two respective countries, since cultural difference exist within regions of
the same country (e.g. the north of Italy and the south). Secondly, the data were collected mostly
from female participants; therefore it is harder to generalize the findings to the male
demographic. Even so, fashion e-tailers such as Asos and Zara sell fashion for men and it could
be interesting to conduct a similar study using a male sample. Finally, many respondents find it
difficult to evaluate the quality of customer reviews, since popular retailers such as Zara.com
don’t offer this type of information. Further research could draw a distinction between fast
fashion and luxury consumers, since their shopping motivation and expectations could differ
significantly.

62
6. References
Books

• Davis, F. 1994. Fashion, culture and identity. The University of Chicago Press, Chicago.
IL

• Kotler, P. & Keller, K. 2012. Marketing management (14th edition) New York: Prentice
Hall Publishers

• Mooi, E. & Sarstedt, M. (2011). A concise guide to market research. Berlin Heidelberg:
Springer-Verlag

• Solomon, M., Bamossy, G., Askegaard, S., Hogg, M. K. 2006. Consumer behaviour: A
European perspective (3rd edition) Harlow, Essex: Pearson Education Limited

Periodicals

• Aladwani, A. 2006. An empirical test of the link between web site quality and forward
enterprise integration with web customers. Business Process Management Journal
12(2): 178–190.

• Al-Qeisi, K., Dennis, C., Alamanos, E., Jayawardhena, C. 2014. Website design quality
and usage behavior: Unified Theory of Acceptance and Use of Technology. Journal of
Business Research 67: 2282–2290

• Bae, S. & Lee, T. 2011. Product type and consumers’ perception of online consumer
reviews. Electronic Markets 21(4): 255-266

• Bai, B., Law, R., Wen, I. 2008. The impact of website quality on customer satisfaction
and purchase intentions: Evidence from Chinese online visitors. International Journal of
Hospitality Management 27 (2008): 391–402

• Barbopoulos, I. & Johansson, L.O. 2016. A multi-dimensional approach to consumer

motivation: exploring economic, hedonic, and normative consumption goals. Journal of
Consumer Marketing 33(1): 75-84

• Brashear, T., Kashyap, V., Musante, M. D., Donthu, N. 2009. A profile of the internet
shopper: evidence from six countries. Journal of Marketing Theory and Practice
17(3):267-281

• Broekhuizen, T. & Huizingh, E. E.K.R. 2009. Online purchase determinants Is their effect
moderated by direct experience? Management Research News 32 (5): 440-457

63
• Brown, M., Pope, N., Voges, K. 2003. Buying or browsing? An exploration of shopping
orientations and online purchase intention. European Journal of Marketing 37, 11/12
:1666-1684

• Brown, S. & Venkatesh, V. 2005. Model of adoption of technology in households: a

baseline model test and extension incorporating household life cycle. MIS Quarterly
29(3):399-426

• Burke, R. 1997. Do you see what I see? The future of virtual shopping. Journal of the
Academy of Marketing Science 25:352-361

• Cai, S. & Xu, Y. 2006. Effects of outcome, process and shopping enjoyment on online
consumer behavior. Electronic Commerce Research and Applications 5:272–281

• Carpenter, J. M. & Fairhurst, A. 2005. Consumer shopping value, satisfaction, and loyalty
for retail apparel brands. Journal of Fashion Marketing and Management 9(3): 256-269

• Celik, H. 2016. Customer online shopping anxiety within the Unified Theory of
Acceptance and Use Technology (UTAUT) framework. Asia Pacific Journal of
Marketing and Logistics 28(2): 278-307

• Celik, H. 2011. Influence of social norms, perceived playfulness and online shopping
anxiety on customers’ adoption of online retail shopping An empirical study in the
Turkish context. International Journal of Retail & Distribution Management
39(6):390-413

• Cheema, A. & Papatla, P. 2010. Relative importance of online versus offline information
for Internet purchases: Product category and Internet experience effects. Journal of
Business Research 63 (2010: 979–985

• Chen, Y. & Xie, J. 2008. Online consumer review: word-of-mouth as a new element of
marketing communication mix. Management Science 54 (3):477-491

• Christensen, C. & Tedlow, R. S. 2000. Patterns of disruption in retailing. Harvard

Business Review 78 (1): 42–5

• Christodoulides, G., Michaelidou, N., Argyriou, E. 2012. Cross-national differences in e-

WOM influence. European Journal of Marketing 46 (11/12):1689-1707

• Chung, K.-H. & Shin, J.-I. (2010). The antecedents and consequents of relationship
quality in internet shopping. Asia Pacific Journal of Marketing and Logistics, 22(4),
473-491.

• Davis, L. 2013. Exploring chinese consumers' shopping value across retail outlets.
International Journal of China Marketing 4(1):51-64.

64
• De Maeyer, P. 2012. Impact of online consumer reviews on sales and price strategies: a
review and directions for future research. Journal of Product & Brand Management
21/2 (2012): 132–139

• Devaraj, S., Fan, M., Kohli, R. 2002. Antecedents of b2C channel satisfaction and
preference: Validation e-Commerce metrics. Information Systems Research 13(3): 316-
333

• Eastman, J. K. & Liu, J. 2012. The impact of generational cohorts on status consumption:
an exploratory look at generational cohort and demographics on status consumption.
Journal of Consumer Marketing 29 (2): 93.102

• Elliott, M. T. & Speck, P. S. 2005. Factors that affect attitude toward a retail web site.
Journal of Marketing Theory and Practice 13(1):40-51

• Fang, M. , Huijing, S., Chen, L., Luo, Y. 2012. A Theory on Fashion Consumption.
Journal of Management and Strategy 3(4):84-92

• Feng, T.-T., Tien, C., Feng, Z.-Y., Lai, P.-J. 2014. Web Site Quality and Online Trading
Influences on Customer Acceptance of Securities Brokers. Asia Pacific Management
Review 19(1): 25-45

• Filieri, R., Alguezaui S., McLeay, F. 2015. Why do travelers trust TripAdvisor?
Antecedents of trust towards consumer-generated media and its influence on
recommendation adoption and word of mouth. Tourism Management 51 (2015) 174:185

• Flavián, C., Guinalíu, M., Gurrea, R. 2006. The role played by perceived usability,
satisfaction and consumer trust on website loyalty. Information & Management 43
(2006):1–14

• Gabrielli, V., Baghi, I., Codeluppi, V. 2013. Consumption practices of fast fashion
products: a consumer-based approach. Journal of Fashion Marketing and Management
17 (2): 206-224

• Ganesh, J., Reynolds, K. E., Luckett, M., Pomirleanu, N. 2010. Online Shopper
Motivations, and e-Store Attributes: An Examination of Online Patronage Behavior and
Shopper Typologies. Journal of Retailing 86(1, 2010): 106-115

• Giovannini, S., Xu, Y., Thomas, J. 2015. Luxury fashion consumption and Generation Y
consumers. Self, brand consciousness, and consumption motivations. Journal of Fashion
Marketing and Management 19(1): 22-40

• Goldsmith, R. & Goldsmith , E. 2002. Buying apparel over the Internet. The Journal of
Product and Brand Management 11(2):89-102

65
• Ha, Y., Kwon, W. S., Lennon, S. J. 2007. Online visual merchandising (VMD) of apparel
web sites. Journal of Fashion Marketing and Management 11(4): 447-493

• Hasanov, J. & Khalid, H. 2015. The impact of Website Quality on online purchase
intention of organic food in Malaysia: A WebQual Model approach. Procedia Computer
Science 72 (2015) : 382 – 389

• Hofstede, G. 1994. Management scientists are human. Management Science 40 (1):4-13

• Hourigan, S. R. & Bougoure, U. S. 2012. Towards a better understanding of fashion

clothing involvement. Australasian Marketing Journal 20(2012):127-135

• Hsu, M. H., Chuang, L. W., Hsu, C. S. 2014. Understanding online shopping intention:
the roles of four types of trust and their antecedents. Internet Research 24(3): 332-352

• Hsu, C.-L., Lin, J. C.-C., Chiang, H.-S. 2013. The effects of blogger recommendations on
customers’ online shopping intentions. Internet Research 23 (1): 69-88

• Hung, S.-Y., Chen, C. C., Huang, N.-H. 2014. An integrative approach to understanding
customer satisfaction with e-service of online stores. Journal of Electronic Commerce
Research 15(1): 40-57.

• Javadi, M. H. M., Dolatabadi, H. R., Nourbakhsh, M., Poursaeedi, A., Asadollahi, A. R.

2012. An Analysis of Factors Affecting on Online Shopping Behavior of Consumers.
International Journal of Marketing Studies 4(5):81-98

• Jiang, L., Yang, Z., Jun, M. 2013. Measuring consumer perceptions of online shopping
convenience. Journal of Service Management 24(2): 191-214

• Jeong S. W., Fiore, A. M., Niehm, L. S., Lorenz, F. O. 2009. The role of experiential
value in online shopping. Internet Research 19 (1): 105-124

• Joung, H.-M. 2014. Fast-fashion consumers’ post-purchase behaviours. International

Journal of Retail & Distribution Management 42(8):688-697

• Kau, A. K., Tang, E. Y., Ghose, S. 2003. Typology of online shoppers. The Journal of
Consumer Marketing 20(2):139-156

• Kim, C., Oh, E., Shin, N. 2010. An empirical investigation of digital content
characteristics, value, and flow. The Journal of Computer Information Systems
(2010):79-87

• Kim, H.-S. 2005. Consumer profiles of apparel product involvement and values. Journal
of Fashion Marketing and Management 9(2):207-220

66
• Kim, J. & Lennon, S. 2010. Information available on a web site: effects on consumers’
shopping outcomes. Journal of Fashion Marketing and Management 14(2):247-262

• Kim J., Jin B., Swinney J. L. 2009. The role of etail quality, e-satisfaction and e-trust in
online loyalty development process. Journal of Retailing and Consumer Services
16(2009)239–247

• Kucukemiroglu, S. & Kara, A. 2015. Online word-of-mouth communication on social

networking sites: An empirical study of Facebook users. International Journal of
Commerce and Management 25(1): 2-20

• Kukar-Kinney, M. & Close, A. 2010. The determinants of consumers’ online shopping

cart abandonment. Journal of the Academy of Marketing Science 38:240-250

• Kwon, W.-S. & Noh, M. 2010. The influence of prior experience and age on mature
consumers’ perceptions and intentions of internet apparel shopping. Journal of Fashion
Marketing and Management 14(3): 335-349

• Law, K. M., Zhang, Z-M., Chung-Sun, L. 2004. Fashion change and fashion
consumption: the chaotic perspective. Journal of Fashion Marketing and Management
2004; 8,4: 362-374

• Lee, H.-H. & Ma, Y. J. 2012. Consumer perceptions of online consumer product and
service reviews. (Focusing on information processing confidence and susceptibility to
peer influence.) Journal of Research in Interactive Marketing 6 (2): 110-132

• Levin, A. M., Levin, I. P., Weller, J. A. 2005. A multi-attribute analysis of preferences for
online and offline shopping: differences across products, consumers, and shopping stages.
Journal of Electronic Commerce Research 6(4): 281-290

• Lim, W. M. 2013. Toward a theory of online buyer behavior using structural equation
modeling. Modern Applied Science 7(10):34-41

• Liu, X., Burns, A.C., Hou, Y. 2013. Comparing online and in-store shopping behavior
towards luxury goods. International Journal of Retail & Distribution Management.
41(11-12):885-900

• Loureiro, S, M. C., Miranda, F. C., Breazeale, M. 2014. Who needs delight? The greater
impact of value, trust and satisfaction in utilitarian, frequent-use retail. Journal of Service
Management 25(1): 101-124

• Matos, C.A. & Rossi, C. A. V. 2008. Word-of-mouth communications in marketing: A

meta-analytic review of the antecedents and moderators. Academy of Marketing Science
Journal, 36(4): 578-596

67
• McCormick, H. & Livett, C. 2012. Analysing the influence of the presentation of fashion
garments on young consumers’ online behavior. Journal of Fashion Marketing and
Management 16(1): 21-41

• Morgan, R. M. and Hunt, S. D. 1994. The commitment-trust theory of relationship

marketing. Journal of Marketing 58(3): 20-38

• O’Cass, A. & Carlson, J. 2012. An empirical assessment of consumers’ evaluations of

web site service quality: conceptualizing and testing a formative model. Journal of
Services Marketing 26(6): 419–434

• O’Cass, A. 2004. Fashion clothing consumption: Antecedents and consequences of

fashion clothing involvement. European Journal of Marketing 38(7):869-882

• Oliver, R. L. 1980. A cognitive model of the antecedents and consequences of satisfaction

decisions. Journal of Marketing Research 17(4): 460-469

• Park, D.-H. & Lee, J. 2008. eWOM overload and its effect on consumer behavioral
intention depending on consumer involvement. Electronic Commerce Research and
Applications 7 (2008): 386–398

• Park, E. J., Kim, E. Y., Forney, J. C. 2006. A structural model of fashion-oriented impulse
buying behavior. Journal of Fashion Marketing and Management 10(4):433-446

• Park, J. & Stoel, L. 2005. Effect of brand familiarity, experience and information on
online apparel purchase. International Journal of Retail & Distribution Management 33
(2): 148-160

• Park, M. & Lennon, S. J. 2009. Brand name and promotion in online shopping contexts.
Journal of Fashion Marketing and Management 13(2): 149-160

• Purnawirawan, N., Eisend, M., De Pelsmacker, P., Dens, N. 2015. A meta-analytic

investigation of the role of valence in online reviews. Journal of Interactive Marketing
31 (2015):17–27

• Rajagopal, 2011. Consumer culture and purchase intentions toward fashion apparel in
Mexico. Journal of Database Marketing & Customer Strategy Management 18(4):286-
307

• Rajamma, R. K., Paswan, A. K., Hossain, M. M. 2009. Why do shoppers abandon

shopping cart? Perceived waiting time, risk, and transaction inconvenience. Journal of
Product & Brand Management 18(3): 188-197

68
• Rajamma, R. K., Paswan, A. K., Ganesh, G. 2007. Services purchased at brick and mortar
versus online stores, and shopping motivation. Journal of Services Marketing 21(3):
200-212

• Rohm, A., Kaltcheva, V. D., Milne, G. R. 2013. A mixed-method approach to examining

brand-consumer interactions driven by social media. Journal of Research in Interactive
Marketing, 7(4) : 295-311

• Salonen S., Närvänen E., Saarijärvi, H. 2014. How do consumers consume fashion
online? A practice-theoretical inquiry. International Journal of Marketing Studies 6(3) :
87-96

• Seock, Y. K. & Norton, M. J. T. 2007. Capturing college students on the web: analysis of
clothing web site attributes. Journal of Fashion Marketing and Management 11(4):539-
552

• Siddiqui, N., O’Malley, A., McColl, J., Birtwistle, G. 2003. Retailer and consumer
perceptions of online fashion retailers: Web site design issues. Journal of Fashion
Marketing and Management 7(4): 345-355

• Smith, R., Deitz, G., Royne, M. B., Hansen, J. D., Grünhagen, M., Witte, C. 2013. Cross-
cultural examination of online shopping behavior: A comparison of Norway, Germany,
and the United States. Journal of Business Research 66 (2013): 328–335

• Spreng, R. A., MacKenzie, S., B., Olshavsky, R. W. 1996. A reexamination of the

determinants of consumer satisfaction. Journal of Marketing 60(July 1996): 15-32

• Sung, H. & Jeon, Y. 2009 A profile of Koreans: who purchases fashion goods online?
Journal of Fashion Marketing and Management 13:79-97

• Venkatesh, V., Morris, M. G., Davis, G. B., Davis, F. D. 2003. User acceptance of
information technology: Toward a unified view. MIS Quarterly 27(3): 425-478

• Vila, N. & Kuster, I. 2011. Consumer feelings and behaviours towards well designed
websites. Information & Management 48:166–177

• Wong, A. & Sohal, A. S. 2006. Understanding the quality of relationships in consumer

services: A study in a retail environment. The International Journal of Quality &
Reliability Management 23(6): 244-264

• Yang, K. 2010. Determinants of US consumer mobile shopping services adoption:

implications for designing mobile shopping services. Journal of Consumer Marketing
27(3): 262–270

69
 • Yoo, J. & Kim, M. 2012. Online product presentation: the effect of product coordination
and a model’s face. Journal of Research in Interactive Marketing 6 (1): 59-72

• Yoon, S-J. 2002. The antecedents and consequences of trust in online-purchase decisions.
Journal of Interactive Marketing 16(2): 47-63

• Zhang, L., Ma, B., Cartwright, D. K. 2013. The impact of online user reviews on cameras
sales. European Journal of Marketing 47 (7): 1115-1128

• Zhang, X., Prybutok, V. R., Koh, C. E. 2006. The Role of Impulsiveness in a TAM-Based
Online Purchasing Behavior Model. Information Resources Management Journal
19(2):54-68

Electronic documents

• Rudolph, S. 2015. The Impact of Online Reviews on Customers’ Buying Decisions

[Infographic] . Business 2 Community. Available at:
http://www.business2community.com/infographics/impact-online-reviews-customers-
buying-decisions-infographic-01280945#i5Zb5s1h0TfSB7DF.97 (accessed on June 10,
2016)

• Statista, 2016. Global retail e-commerce sales volume from 2013 to 2018 (in billion U.S.
dollars). Available at: http://www.statista.com/statistics/222128/global-e-commerce-sales-
volume-forecast/ (accessed on July 10, 2016)

• Europe B2C E-commerce report 2012, Dec 17. PR Newswire. Available at :

http://search.proquest.com/docview/1239047735?accountid=38384 (accessed on June 10,
2016)

• Internet Live Stats, 2016. Internet Users by Country. Available at:

http://www.internetlivestats.com/internet-users-by-country/ (accessed on June 10, 2016)

• The Hofstede Centre, 2016. Italy in comparison with Portugal. Available at:
https://geert-hofstede.com/italy.html (accessed on June 10, 2016)

70
7. Appendixes
Appendix A: Sample profile & Descripti ve statistics

Gender

Gender
Frequency Percent Valid Cumulative
Percent Percent
Valid Female 251 80,4 80,4 80,4
Male 61 19,6 19,6 100,0
Total 312 100,0 100,0

Nationality

Nationality
Frequency Percent Valid Percent Cumulative
Percent
Valid Italian 183 58,7 58,7 58,7
Portuguese 129 41,3 41,3 100,0
Total 312 100,0 100,0

Employment status

Employment status
Frequency Percent Valid Percent Cumulative
Percent
Valid Employed 116 37,2 37,2 37,2
Other 12 3,8 3,8 41,0
Self-employed 27 8,7 8,7 49,7
Student 145 46,5 46,5 96,2
Unemployed 12 3,8 3,8 100,0
Total 312 100,0 100,0

71
 User online experience

Crosstabulation for Gender and Nationality

Gender * Nationality Crosstabulation

Nationality Total
Italian Portuguese
Gender Female Count 138 113 251
% within Gender 55,0% 45,0% 100,0%
% within Nationality 75,4% 87,6% 80,4%
% of Total 44,2% 36,2% 80,4%
Male Count 45 16 61
% within Gender 73,8% 26,2% 100,0%
% within Nationality 24,6% 12,4% 19,6%
% of Total 14,4% 5,1% 19,6%
Total Count 183 129 312
% within Gender 58,7% 41,3% 100,0%
% within Nationality 100,0% 100,0% 100,0%
% of Total 58,7% 41,3% 100,0%

Means statistics for age in years

Group Statistics
1 = Portuguese 2 = Italian N Mean Std. Std. Error
Deviation Mean
Age in Portuguese 129 26,56 7,877 ,694
years Italian 183 28,15 9,810 ,725

72
 Descriptive statistics for Technical Quality (TQ)

Statistics
This This fashion This fashion This fashion This fashion This
fashion Website: has Website: Website: Website: is fashion
Website: is well organized provides has high easily Website:
easy to hyperlinks opportunities speed of accessible guarantees
use to interact with page from different users'
other loading media privacy
customers
N Valid 312 312 312 312 312 312
Missing 0 0 0 0 0 0
Mean 4,2212 3,8718 2,3077 3,8558 4,0128 3,9712
Std. Deviation ,85949 ,97373 1,20618 ,98626 ,99185 ,97351

Descriptive statistics for Product Information (PI)

Statistics
This fashion This fashion This fashion This fashion This This fashion
Website : Website : Website : Website : fashion Website :
shows all the shows all the tells the price gives up-to- Website : truthfully
colors sizes of products date has good shows the
available for available for clearly information quality colors of
each product each product about photos of products
products and products
trends
N Valid 312 312 312 312 312 312
Missing 0 0 0 0 0 0
Mean 3,8686 3,9679 4,4103 3,9808 4,0032 3,8558
Std. Deviation 1,15044 1,12510 ,81680 1,03614 1,05018 ,97971

Descriptive statistics for Information Quality of customer reviews (IQ)

Statistics
The The The The The The
information information in information in information in information in information in
in the the customer the customer the customer the customer the customer
customer reviews is reviews is reviews is reviews is reviews is
reviews is relevant to complete for valuable useful credible
timely my needs my needs
N Valid 312 312 312 312 312 312
Missing 0 0 0 0 0 0
Mean 3,1795 3,1442 3,0769 3,0801 3,2692 2,9872
Std. Deviation 1,09066 1,13482 1,13431 1,13267 1,13038 1,07881

73
 Descriptive statistics for Source Credibility (SC)

Statistics
The reviewers on The reviewers on The reviewers on The reviewers on
this fashion Website this fashion Website this fashion Website this fashion Website
are : credible are : experienced are : trustworthy are : reliable
N Valid 312 312 312 312
Missing 0 0 0 0
Mean 3,2179 2,7532 3,1090 3,0417
Std. Deviation 1,07467 1,05781 1,05217 1,04321

Descriptive statistics for Customer Satisfaction (S)

Statistics
I am satisfied with the fashion I am satisfied with my previous
information I have received from this experiences with this Website
Website
N Valid 312 312
Missing 0 0
Mean 4,0128 4,0929
Std. Deviation ,93511 ,90070

Descriptive statistics for Website Trust (T)

Statistics
I think that the I think that the advice I trust the online [I trust this fashion
information offered and customer reviews Website
by this fashion recommendations on this Website
Website is sincere given by the customer
and honest reviews are
trustworthy
N Valid 312 312 312 312
Missing 0 0 0 0
Mean 3,9583 3,3750 3,3141 4,0641
Std. Deviation ,86455 1,07159 1,12158 ,85004

Descriptive statistics for Social Influence (SI)

Statistics
People who are important to me think that I People who influence my behavior think I
should use online stores to shop for fashion should use online stores to shop for
products fashion products
N Valid 312 312
Missing 0 0
Mean 2,6731 2,6282
Std. Deviation 1,17127 1,15499

74
 Descriptive statistics for Performance Expectancy (PE)

Statistics
I find fashion Using fashion online Using fashion online Using fashion online
online stores stores enables me stores increases the stores increases the
useful to get fashion effective use of my time quality of my fashion
information more in handling my knowledge at
quickly shopping tasks and minimal effort
purchase
N Valid 312 312 312 312
Missing 0 0 0 0
Mean 4,0513 3,9968 3,7917 3,6635
Std. Deviation 1,01623 1,03165 1,12478 1,10198

Descriptive statistics for WOM Behaviour (WM)

Statistics
I mentioned to others I made sure that others I spoke positively I recommended this
that I seek fashion know that I rely on this about this fashion Website to
information from Website to purchase Website to others close friends
this Website fashion products
N Valid 312 312 312 312
Missing 0 0 0 0
Mean 3,2404 2,9519 3,6955 3,6186
Std. Deviation 1,28930 1,25813 1,07001 1,17806

Descriptive statistics for Recommendation Adoption (REC)

Statistics
Online customer Online reviews have The last time I read Information from
reviews and motivated me to online fashion customer reviews
comments made it make a purchase reviews I adopted contributed to my
easier for me to make decision (purchase consumers' knowledge of
a purchase decision or not purchase) recommendations fashion products
(e.g., purchase or not and trends
purchase)
N Valid 312 312 312 312
Missing 0 0 0 0
Mean 2,9327 2,8333 2,6763 2,7917
Std. Deviation 1,22617 1,16118 1,17079 1,16412

Descriptive statistics for Internet Experience (IEX)

Statistics
How would you describe your: How would you describe your:
Internet knowledge General computer knowledge
N Valid 312 312
Missing 0 0
Mean 4,1026 3,9103
Std. Deviation ,88343 ,95474

75
 Descriptive statistics for User Online Experience (UEX)

Statistics
How would you rate: your How would you rate: your How would you rate: your
level of experience in level of experience in level of experience in
terms of using fashion terms of browsing fashion terms of online customer
websites websites reviews
N Valid 312 312 312
Missing 0 0 0
Mean 3,4359 3,6058 3,0481
Std. Deviation 1,12086 1,10902 1,09995

76
 Appendix B: Regression analysis results

Determinants of Performance Expectancy

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,663a ,440 ,431 ,69510 1,961

ANOVAa
Model Sum of Df Mean F Sig.
Squares Square
b
1 Regression 116,153 5 23,231 48,081 ,000
Residual 147,847 306 ,483
Total 264,000 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Toleranc VIF
e
1 (Constant) ,516 ,236 2,182 ,030
Tecquality ,217 ,075 ,166 2,908 ,004 ,558 1,791
Productinf ,479 ,065 ,406 7,427 ,000 ,613 1,632
Infquality -,058 ,065 -,063 -,904 ,367 ,382 2,620
Sourcecred ,072 ,065 ,075 1,106 ,269 ,396 2,526
Socinfluence ,225 ,038 ,270 5,983 ,000 ,896 1,116

Influence of experience on Performance Expectancy

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,590a ,348 ,344 ,74623 1,919

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 91,930 2 45,965 82,544 ,000b
Residual 172,070 309 ,557
Total 264,000 311

a
Coefficients
Model Unstandardized Coefficients Standardized t Sig. Collinearity
Coefficients Statistics
B Std. Error Beta Toleran VIF
ce
1 (Constant) 1,731 ,201 8,608 ,000
Intexperience ,130 ,061 ,124 2,135 ,034 ,627 1,595
Userexperience ,483 ,055 ,506 8,728 ,000 ,627 1,595

77
 Determinants of Performance Expectancy together with experience

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,719 ,518 ,507 ,64722 1,904

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 136,658 7 19,523 46,606 ,000b
Residual 127,342 304 ,419
Total 264,000 311

a
Coefficients
Model Unstandardized Coefficients Standardized t Sig. Collinearity
Coefficients Statistics
B Std. Error Beta Toleran VIF
ce
1 (Constant) ,347 ,235 1,476 ,141
Tecquality ,161 ,071 ,123 2,263 ,024 ,534 1,874
Productinf ,351 ,063 ,298 5,576 ,000 ,557 1,795
Infquality -,051 ,061 -,055 -,839 ,402 ,373 2,681
Sourcecred ,015 ,061 ,015 ,241 ,810 ,388 2,575
Socinfluence ,187 ,036 ,224 5,258 ,000 ,871 1,148
Intexperience ,032 ,055 ,031 ,583 ,561 ,573 1,746
Userexperience ,301 ,053 ,316 5,716 ,000 ,521 1,921

Experience influences Technical Quality

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,446a ,199 ,194 ,63426 1,988

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 30,942 2 15,471 38,457 ,000b
Residual 124,307 309 ,402
Total 155,249 311

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,303 ,171 13,474 ,000
Intexperience ,190 ,052 ,237 3,683 ,000 ,627 1,595
Userexperience ,191 ,047 ,261 4,053 ,000 ,627 1,595

78
 Experience influences Product Information

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,511 ,261 ,256 ,67307 1,951

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 49,367 2 24,684 54,487 ,000b
Residual 139,984 309 ,453
Total 189,352 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,299 ,181 12,678 ,000
Intexperience ,190 ,055 ,214 3,463 ,001 ,627 1,595
Userexperience ,284 ,050 ,351 5,683 ,000 ,627 1,595

Experience influences Information Quality of customer reviews

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,310a ,096 ,090 ,94335 1,984

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 29,251 2 14,625 16,434 ,000b
Residual 274,984 309 ,890
Total 304,235 311

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity
Coefficients Statistics
B Std. Error Beta Toleran VIF
ce
1 (Constant) 2,326 ,254 9,152 ,000
Intexperience -,111 ,077 -,099 -1,442 ,150 ,627 1,595
Userexperience ,369 ,070 ,360 5,274 ,000 ,627 1,595

79
 Performance Expectancy and Customer Satisfaction

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,557 ,310 ,308 ,72380 2,025

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 72,974 1 72,974 139,296 ,000b
Residual 162,403 310 ,524
Total 235,377 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,015 ,177 11,356 ,000
Perfexpectancy ,526 ,045 ,557 11,802 ,000 1,000 1,000

Performance Expectancy and Website Trust

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,472a ,223 ,220 ,70734 2,108

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 44,401 1 44,401 88,745 ,000b
Residual 155,101 310 ,500
Total 199,502 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,088 ,173 12,043 ,000
Perfexpectancy ,410 ,044 ,472 9,420 ,000 1,000 1,000

80
 Performance Expectancy and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,202 ,041 ,038 1,03773 1,926

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 14,160 1 14,160 13,149 ,000b
Residual 333,835 310 1,077
Total 347,995 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,911 ,254 7,511 ,000
Perfexpectancy ,232 ,064 ,202 3,626 ,000 1,000 1,000

Performance Expectancy and WOM Behaviour

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,504a ,254 ,252 ,87889 1,963

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 81,539 1 81,539 105,559 ,000b
Residual 239,460 310 ,772
Total 320,999 311

a
Coefficients
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,223 ,215 5,674 ,000
Perfexpectancy ,556 ,054 ,504 10,274 ,000 1,000 1,000

81
 Perceived website quality and Customer Satisfaction

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,704 ,495 ,494 ,61893 2,146

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
Regression 116,623 1 116,623 304,438 ,000b
Residual 118,754 310 ,383
Total 235,377 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
(Constant) ,652 ,198 3,290 ,001
PERCEIVEDQUALITY ,941 ,054 ,704 17,448 ,000 1,000 1,000

Perceived website quality and Website Trust

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
,779 ,607 ,606 ,50279 2,040

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
Regression 121,136 1 121,136 479,185 ,000
Residual 78,367 310 ,253
Total 199,502 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
(Constant) ,211 ,161 1,314 ,190
PERCEIVEDQUALITY ,959 ,044 ,779 21,890 ,000 1,000 1,000

82
 Perceived website quality and WOM Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
,519a ,270 ,268 ,86951 1,936

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
Regression 86,624 1 86,624 114,575 ,000b
Residual 234,375 310 ,756
Total 320,999 311

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
(Constant) ,445 ,278 1,600 ,111
PERCEIVEDQUALITY ,811 ,076 ,519 10,704 ,000 1,000 1,000

Perceived website quality and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,450a ,202 ,200 ,94626 1,824

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 70,420 1 70,420 78,646 ,000b
Residual 277,575 310 ,895
Total 347,995 311

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) ,166 ,303 ,547 ,585
PERCEIVEDQUALITY ,731 ,082 ,450 8,868 ,000 1,000 1,000

83
 Online drivers of Website Trust
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,690a ,476 ,473 ,58157 2,194

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 94,992 2 47,496 140,429 ,000b
Residual 104,510 309 ,338
Total 199,502 311

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,964 ,170 5,680 ,000
Perfexpectancy ,117 ,043 ,134 2,707 ,007 ,690 1,449
Satisfaction ,558 ,046 ,606 12,230 ,000 ,690 1,449

Website trust and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,439a ,193 ,190 ,95180 1,871

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 67,162 1 67,162 74,137 ,000
Residual 280,833 310 ,906
Total 347,995 311

a
Coefficients
Model Unstandardized Coefficients Standardized t Sig. Collinearity
Coefficients Statistics
B Std. Error Beta Tolerance VIF
1 (Constant) ,675 ,254 2,660 ,008
Trust ,580 ,067 ,439 8,610 ,000 1,000 1,000

84
 Performance Expectancy, Customer Satisfaction and Website Trust as drivers of WOM
Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,613 ,376 ,370 ,80659 1,928

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 120,618 3 40,206 61,799 ,000
Residual 200,382 308 ,651
Total 320,999 311

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,072 ,247 ,292 ,770
Satisfaction ,230 ,077 ,197 2,981 ,003 ,465 2,151
Perfexpectancy ,300 ,060 ,272 4,961 ,000 ,674 1,484
Trust ,329 ,079 ,259 4,170 ,000 ,524 1,909

Performance Expectancy, Customer Satisfaction, Website Trust and Recommendation

Adoption as drivers of WOM Behaviour

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,647 ,418 ,411 ,77978 1,991

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 134,328 4 33,582 55,229 ,000b
Residual 186,671 307 ,608
Total 320,999 311

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) -,159 ,244 -,653 ,514
Satisfaction ,323 ,077 ,277 4,190 ,000 ,435 2,300
Perfexpectancy ,275 ,059 ,250 4,689 ,000 ,669 1,496
Trust ,141 ,086 ,111 1,641 ,102 ,413 2,423
Recadoption ,228 ,048 ,238 4,748 ,000 ,755 1,325

85
 Appendix C: Regression analysis results for the Portuguese sample

Determinants of Performance Expectancy together with experience

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,619a ,383 ,348 ,55784 1,937

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 23,397 7 3,342 10,741 ,000b
Residual 37,653 121 ,311
Total 61,050 128

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity
Coefficients Statistics
B Std. Error Beta Toleran VIF
ce
1 (Constant) 1,525 ,476 3,206 ,002
Tecquality ,240 ,097 ,224 2,477 ,015 ,621 1,610
Productinf ,191 ,102 ,155 1,870 ,064 ,739 1,353
Infquality -,137 ,084 -,194 -1,623 ,107 ,358 2,797
Sourcecred ,064 ,082 ,084 ,784 ,435 ,444 2,253
Socinfluence ,191 ,044 ,317 4,291 ,000 ,936 1,069
Intexperience -,092 ,080 -,096 -1,149 ,253 ,730 1,370
Userexperience ,286 ,074 ,348 3,856 ,000 ,625 1,601

Experience influences Technical Quality

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,301a ,091 ,076 ,62027 2,052

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 4,830 2 2,415 6,277 ,003b
Residual 48,476 126 ,385
Total 53,306 128

86
 a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity
Coefficients Coefficients Statistics
B Std. Error Beta Toleran VIF
ce
1 (Constant) 2,813 ,339 8,309 ,000
Intexperience ,050 ,087 ,056 ,575 ,566 ,769 1,301
Userexperience ,207 ,074 ,270 2,789 ,006 ,769 1,301

Experience influences Product Information

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,353 ,124 ,110 ,52842 1,679

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 4,997 2 2,499 8,948 ,000b
Residual 35,183 126 ,279
Total 40,180 128

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 3,408 ,288 11,815 ,000
Intexperience ,035 ,074 ,045 ,475 ,635 ,769 1,301
Userexperience ,219 ,063 ,329 3,457 ,001 ,769 1,301

Experience influences Information Quality of customer reviews

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,326a ,106 ,092 ,93188 1,855

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 12,975 2 6,488 7,471 ,001
Residual 109,419 126 ,868
Total 122,394 128

87
 a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,727 ,509 5,361 ,000
Intexperience -,259 ,130 -,191 -1,986 ,049 ,769 1,301
Userexperience ,431 ,112 ,371 3,863 ,000 ,769 1,301

Performance Expectancy and Customer Satisfaction

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,340 ,115 ,109 ,66492 1,985

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 7,331 1 7,331 16,582 ,000b
Residual 56,149 127 ,442
Total 63,481 128

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,755 ,366 7,537 ,000
Perfexpectancy ,347 ,085 ,340 4,072 ,000 1,000 1,000

Performance Expectancy and Website Trust

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,273 ,075 ,067 ,68168 1,975

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 4,756 1 4,756 10,234 ,002b
Residual 59,015 127 ,465
Total 63,770 128

88
 a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,648 ,375 7,065 ,000
Perfexpectancy ,279 ,087 ,273 3,199 ,002 1,000 1,000

Performance Expectancy and Recommendation Adoption

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,010a ,000 -,008 1,04354 2,080

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression ,015 1 ,015 ,014 ,907
Residual 138,300 127 1,089
Total 138,315 128

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,749 ,574 4,792 ,000
Perfexpectancy ,016 ,134 ,010 ,117 ,907 1,000 1,000

Performance Expectancy and WOM Behaviour

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,320a ,103 ,096 ,98568 1,877

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 14,111 1 14,111 14,524 ,000
Residual 123,390 127 ,972
Total 137,501 128

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,492 ,542 2,754 ,007
Perfexpectancy ,481 ,126 ,320 3,811 ,000 1,000 1,000

89
 Perceived website quality and Customer Satisfaction
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,527a ,277 ,272 ,60102 2,112

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 17,605 1 17,605 48,737 ,000b
Residual 45,876 127 ,361
Total 63,481 128

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) 1,651 ,372 4,434 ,000
PERCEIVEDQUALITY ,676 ,097 ,527 6,981 ,000 1,000 1,000

Perceived website quality and Website Trust

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,664 ,441 ,437 ,52977 1,971

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 28,126 1 28,126 100,215 ,000
Residual 35,644 127 ,281
Total 63,770 128

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,578 ,328 1,762 ,080
PERCEIVEDQUALITY ,855 ,085 ,664 10,011 ,000 1,000 1,000

90
 Perceived website quality and WOM Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,389a ,151 ,144 ,95875 1,795

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 20,762 1 20,762 22,587 ,000
Residual 116,739 127 ,919
Total 137,501 128

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) ,736 ,594 1,239 ,218
PERCEIVEDQUALITY ,734 ,155 ,389 4,753 ,000 1,000 1,000

Perceived website quality and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,380 ,144 ,137 ,96549 2,021

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 19,930 1 19,930 21,381 ,000
Residual 118,385 127 ,932
Total 138,315 128

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) ,078 ,598 ,130 ,897
PERCEIVEDQUALITY ,719 ,156 ,380 4,624 ,000 1,000 1,000

91
 Online drivers of Website Trust
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,547a ,299 ,288 ,59556 2,131

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 19,080 2 9,540 26,896 ,000b
Residual 44,691 126 ,355
Total 63,770 128

Coefficientsa
Model Unstandardized Coefficients Standardiz t Sig. Collinearity Statistics
ed
Coefficient
s
B Std. Error Beta Tolerance VIF
1 (Constant) 1,256 ,394 3,189 ,002
Perfexpectancy ,104 ,081 ,102 1,284 ,201 ,885 1,131
Satisfaction ,505 ,079 ,504 6,355 ,000 ,885 1,131

Website trust and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,302 ,091 ,084 ,99484 2,113

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 12,622 1 12,622 12,753 ,001b
Residual 125,693 127 ,990
Total 138,315 128

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,111 ,485 2,290 ,024
Trust ,445 ,125 ,302 3,571 ,001 1,000 1,000

92
 Performance Expectancy, Customer Satisfaction and Website Trust as drivers of WOM
Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,474 ,225 ,206 ,92336 1,873

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 30,927 3 10,309 12,091 ,000
Residual 106,574 125 ,853
Total 137,501 128

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) -,188 ,635 -,296 ,768
Satisfaction ,279 ,142 ,190 1,970 ,051 ,670 1,493
Perfexpectancy ,288 ,126 ,192 2,277 ,024 ,873 1,145
Trust ,344 ,138 ,234 2,493 ,014 ,701 1,427

Performance Expectancy, Customer Satisfaction, Website Trust and Recommendation

Adoption as drivers of WOM Behaviour

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,542 ,294 ,271 ,88484 1,975

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 40,415 4 10,104 12,905 ,000b
Residual 97,086 124 ,783
Total 137,501 128

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) -,762 ,630 -1,209 ,229
Satisfaction ,383 ,139 ,260 2,754 ,007 ,639 1,565
Perfexpectancy ,299 ,121 ,199 2,467 ,015 ,872 1,146
Trust ,160 ,143 ,109 1,123 ,263 ,604 1,655
Recadoption ,282 ,081 ,283 3,481 ,001 ,862 1,161

93
 Appendix D: Regression analysis results for the Italian sample

Determinants of Performance Expectancy together with experience

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,717a ,514 ,495 ,69443 1,954

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 89,337 7 12,762 26,465 ,000b
Residual 84,391 175 ,482
Total 173,727 182

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,185 ,291 ,636 ,526
Tecquality ,160 ,103 ,122 1,561 ,120 ,456 2,192
Productinf ,318 ,091 ,266 3,487 ,001 ,478 2,091
Infquality ,021 ,085 ,021 ,245 ,807 ,374 2,675
Sourcecred -,061 ,087 -,061 -,698 ,486 ,366 2,732
Socinfluence ,193 ,054 ,213 3,586 ,000 ,787 1,270
Intexperience ,099 ,076 ,097 1,303 ,194 ,496 2,016
Userexperience ,283 ,076 ,280 3,745 ,000 ,497 2,010

Experience influences Technical Quality

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,511a ,261 ,253 ,64164 1,926

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 26,139 2 13,069 31,744 ,000
Residual 74,107 180 ,412
Total 100,246 182

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,105 ,202 10,440 ,000
Intexperience ,254 ,066 ,328 3,871 ,000 ,572 1,747
Userexperience ,178 ,065 ,232 2,736 ,007 ,572 1,747

94
 Experience influences Product Information

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,504 ,254 ,246 ,70864 2,261

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 30,848 2 15,424 30,715 ,000
Residual 90,391 180 ,502
Total 121,239 182

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,089 ,223 9,380 ,000
Intexperience ,279 ,072 ,327 3,848 ,000 ,572 1,747
Userexperience ,190 ,072 ,225 2,648 ,009 ,572 1,747

Experience influences Information Quality of customer reviews

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,284a ,081 ,070 ,95381 2,060

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 14,346 2 7,173 7,885 ,001
Residual 163,756 180 ,910
Total 178,102 182

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 2,187 ,300 7,294 ,000
Intexperience -,030 ,097 -,029 -,303 ,762 ,572 1,747
Userexperience ,309 ,097 ,302 3,193 ,002 ,572 1,747

95
 Performance Expectancy and Customer Satisfaction

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,613 ,375 ,372 ,75548 2,006

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 62,092 1 62,092 108,791 ,000b
Residual 103,304 181 ,571
Total 165,396 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,768 ,215 8,232 ,000
Perfexpectancy ,598 ,057 ,613 10,430 ,000 1,000 1,000

Performance Expectancy and Website Trust

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,524a ,275 ,271 ,72331 2,156

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 35,853 1 35,853 68,529 ,000b
Residual 94,696 181 ,523
Total 130,549 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,926 ,206 9,363 ,000
Perfexpectancy ,454 ,055 ,524 8,278 ,000 1,000 1,000

96
 Performance Expectancy and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,312 ,097 ,093 1,02247 1,814

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 20,441 1 20,441 19,552 ,000b
Residual 189,227 181 1,045
Total 209,668 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,562 ,291 5,373 ,000
Perfexpectancy ,343 ,078 ,312 4,422 ,000 1,000 1,000

Performance Expectancy and WOM Behaviour

b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,597a ,356 ,352 ,79642 2,040

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 63,448 1 63,448 100,029 ,000b
Residual 114,807 181 ,634
Total 178,255 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) 1,081 ,226 4,773 ,000
Perfexpectancy ,604 ,060 ,597 10,001 ,000 1,000 1,000

97
 Perceived website quality and Customer Satisfaction

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,764 ,584 ,581 ,61679 2,147

ANOVAa
Model Sum of Df Mean F Sig.
Squares Square
1 Regression 96,539 1 96,539 253,766 ,000b
Residual 68,857 181 ,380
Total 165,396 182

a
Coefficients
Model Unstandardized Standardize t Sig. Collinearity Statistics
Coefficients d
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,226 ,237 ,956 ,341
PERCEIVEDQUALITY 1,065 ,067 ,764 15,930 ,000 1,000 1,000

Perceived website quality and Website Trust

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,825a ,680 ,678 ,48042 2,087

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 88,774 1 88,774 384,636 ,000b
Residual 41,775 181 ,231
Total 130,549 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) ,017 ,185 ,090 ,928
PERCEIVEDQUALITY 1,021 ,052 ,825 19,612 ,000 1,000 1,000

98
 Perceived website quality and WOM Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,585a ,342 ,338 ,80504 2,074

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 60,950 1 60,950 94,046 ,000b
Residual 117,304 181 ,648
Total 178,255 182

a
Coefficients
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,324 ,309 1,046 ,297
PERCEIVEDQUALITY ,846 ,087 ,585 9,698 ,000 1,000 1,000

Perceived website quality and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,511a ,261 ,257 ,92506 1,715

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 54,781 1 54,781 64,017 ,000b
Residual 154,887 181 ,856
Total 209,668 182

a
Coefficients
Model Unstandardized Standardize t Sig. Collinearity Statistics
Coefficients d
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,012 ,355 ,034 ,973
PERCEIVEDQUALITY ,802 ,100 ,511 8,001 ,000 1,000 1,000

99
 Online drivers of Website Trust
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,738a ,544 ,539 ,57485 2,226

a
ANOVA
Model Sum of df Mean F Sig.
Squares Square
1 Regression 71,067 2 35,533 107,527 ,000b
Residual 59,482 180 ,330
Total 130,549 182

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,893 ,192 4,662 ,000
Perfexpectancy ,105 ,055 ,121 1,907 ,058 ,625 1,601
Satisfaction ,584 ,057 ,657 10,323 ,000 ,625 1,601

Website Trust and Recommendation Adoption

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,527a ,277 ,273 ,91492 1,681

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 58,156 1 58,156 69,474 ,000b
Residual 151,512 181 ,837
Total 209,668 182

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,421 ,294 1,432 ,154
Trust ,667 ,080 ,527 8,335 ,000 1,000 1,000

100
 Performance Expectancy, Customer Satisfaction and Website Trust as drivers of WOM
Behaviour
b
Model Summary
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
a
1 ,692 ,478 ,470 ,72077 1,986

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
b
1 Regression 85,263 3 28,421 54,707 ,000
Residual 92,992 179 ,520
Total 178,255 182

Coefficientsa
Model Unstandardized Coefficients Standardized t Sig. Collinearity Statistics
Coefficients
B Std. Error Beta Tolerance VIF
1 (Constant) ,105 ,254 ,412 ,681
Satisfaction ,199 ,089 ,191 2,219 ,028 ,392 2,549
Perfexpectancy ,338 ,070 ,334 4,839 ,000 ,612 1,633
Trust ,325 ,093 ,278 3,472 ,001 ,456 2,195

Performance Expectancy, Customer Satisfaction, Website Trust and Recommendation

Adoption as drivers of WOM Behaviour

Model Summaryb
Model R R Square Adjusted R Std. Error of Durbin-Watson
Square the Estimate
1 ,711a ,505 ,494 ,70382 2,018

ANOVAa
Model Sum of df Mean F Sig.
Squares Square
1 Regression 90,080 4 22,520 45,462 ,000b
Residual 88,175 178 ,495
Total 178,255 182

Coefficientsa
Model Unstandardized Standardized t Sig. Collinearity Statistics
Coefficients Coefficients
B Std. Beta Tolerance VIF
Error
1 (Constant) -,009 ,251 -,038 ,970
Satisfaction ,291 ,092 ,280 3,154 ,002 ,352 2,843
Perfexpectancy ,299 ,069 ,296 4,315 ,000 ,592 1,688
Trust ,146 ,108 ,125 1,355 ,177 ,327 3,058
Recadoption ,189 ,060 ,204 3,118 ,002 ,646 1,547

101
Appendix E: One-way ANOVA and non-parametric tests

Customer Satisfaction

Test of Homogeneity of Variances

Satisfaction
Levene Statistic df1 df2 Sig.
,566 2 111 ,569

ANOVA
Satisfaction
Sum of df Mean Square F Sig.
Squares
Between Groups 5,815 2 2,907 5,144 ,007
Within Groups 62,731 111 ,565
Total 68,546 113

Multiple Comparisons
Dependent Variable: Satisfaction
Scheffe
(I) (J) Mean Std. Sig. 95% Confidence Interval
website website Difference (I- Error Lower Upper
J) Bound Bound
*
ZARA AMAZON ,48571 ,17205 ,021 ,0588 ,9126
ASOS ,03591 ,17726 ,980 -,4039 ,4757
AMAZON ZARA -,48571* ,17205 ,021 -,9126 -,0588
*
ASOS -,44981 ,16950 ,033 -,8704 -,0293
ASOS ZARA -,03591 ,17726 ,980 -,4757 ,4039
*
AMAZON ,44981 ,16950 ,033 ,0293 ,8704
*. The mean difference is significant at the 0.05 level.

Website Trust

Test of Homogeneity of Variances

Trust
Levene Statistic df1 df2 Sig.
,599 2 111 ,551

ANOVA
Trust
Sum of df Mean Square F Sig.
Squares
Between Groups 1,503 2 ,752 1,281 ,282
Within Groups 65,140 111 ,587
Total 66,643 113

102
 Multiple Comparisons
Dependent Variable: Trust
Scheffe
(I) (J) Mean Std. Sig. 95% Confidence Interval
website website Difference (I- Error Lower Upper
J) Bound Bound
ZARA AMAZON ,27976 ,17533 ,284 -,1553 ,7148
ASOS ,17181 ,18063 ,637 -,2764 ,6200
AMAZON ZARA -,27976 ,17533 ,284 -,7148 ,1553
ASOS -,10795 ,17272 ,823 -,5365 ,3206
ASOS ZARA -,17181 ,18063 ,637 -,6200 ,2764
AMAZON ,10795 ,17272 ,823 -,3206 ,5365

WOM Behaviour

Test of Homogeneity of Variances

WoM
Levene Statistic df1 df2 Sig.
,943 2 111 ,393

ANOVA
WoM
Sum of df Mean Square F Sig.
Squares
Between Groups 5,357 2 2,678 3,019 ,053
Within Groups 98,483 111 ,887
Total 103,840 113

Multiple Comparisons
Dependent Variable: WoM
Scheffe
(I) (J) Mean Std. Sig. 95% Confidence Interval
website website Difference (I- Error Lower Upper
J) Bound Bound
ZARA AMAZON ,20714 ,21558 ,631 -,3277 ,7420
ASOS -,31313 ,22210 ,373 -,8642 ,2379
AMAZON ZARA -,20714 ,21558 ,631 -,7420 ,3277
ASOS -,52027 ,21238 ,054 -1,0472 ,0067
ASOS ZARA ,31313 ,22210 ,373 -,2379 ,8642
AMAZON ,52027 ,21238 ,054 -,0067 1,0472

103
 Product Information

Test of Homogeneity of Variances

Productinf
Levene Statistic df1 df2 Sig.
7,282 2 111 ,001

ANOVA
Productinf
Sum of df Mean Square F Sig.
Squares
Between Groups 18,779 2 9,389 26,282 ,000
Within Groups 39,655 111 ,357
Total 58,434 113

Multiple Comparisons
Dependent Variable: Productinf
Scheffe
(I) (J) Mean Std. Sig. 95% Confidence Interval
website website Difference (I- Error Lower Upper
J) Bound Bound
*
ZARA AMAZON ,87063 ,13680 ,000 ,5312 1,2100
ASOS ,05972 ,14094 ,914 -,2900 ,4094
AMAZON ZARA -,87063* ,13680 ,000 -1,2100 -,5312
ASOS -,81092* ,13476 ,000 -1,1453 -,4765
ASOS ZARA -,05972 ,14094 ,914 -,4094 ,2900
AMAZON ,81092* ,13476 ,000 ,4765 1,1453
*. The mean difference is significant at the 0.05 level.

Product Information Kruskal-Wallis

Ranks
website N Mean Rank
Productinf ZARA 35 72,20
AMAZON 42 34,65
ASOS 37 69,53
Total 114

Test Statisticsa,b
Productinf
Chi-Square 32,145
df 2
Asymp. Sig. ,000
a. Kruskal Wallis Test
b. Grouping Variable: website

104
 Technical Quality

Test of Homogeneity of Variances

Tecquality
Levene Statistic df1 df2 Sig.
1,067 2 111 ,348

ANOVA
Tecquality
Sum of df Mean Square F Sig.
Squares
Between Groups ,282 2 ,141 ,291 ,748
Within Groups 53,781 111 ,485
Total 54,063 113

Multiple Comparisons
Dependent Variable: Tecquality
Scheffe
(I) (J) Mean Std. Sig. 95% Confidence Interval
website website Difference (I- Error Lower Upper
J) Bound Bound
ZARA AMAZON ,06667 ,15931 ,916 -,3286 ,4619
ASOS -,05238 ,16413 ,950 -,4596 ,3548
AMAZON ZARA -,06667 ,15931 ,916 -,4619 ,3286
ASOS -,11905 ,15694 ,751 -,5084 ,2704
ASOS ZARA ,05238 ,16413 ,950 -,3548 ,4596
AMAZON ,11905 ,15694 ,751 -,2704 ,5084

105
Appendix F: Questionnaire

106
107
108
109
110