02 Whole

Download as pdf or txt
Download as pdf or txt
You are on page 1of 324

 

SOCIAL MEDIA ENGAGEMENT BEHAVIOUR:


A USES AND GRATIFICATIONS PERSPECTIVE

 
 
A thesis submitted in fulfilment

of the requirement for the degree of

DOCTOR OF PHILOSOPHY

By

Rebecca Marie Dolan, B.Com. (Hons)

School of Marketing and Management


Adelaide Business School
University of Adelaide
2015
Table of Contents
 
List of Figures ....................................................................................................................... v
List of Tables ....................................................................................................................... vi 
Abstract ............................................................................................................................... x 
Declaration ........................................................................................................................xii 
Publications ...................................................................................................................... xiii 
Acknowledgements............................................................................................................ xiv 
Key Terms, Definitions and Abbreviations .................................................................... xvi 
CHAPTER 1. Introduction.................................................................................................. 1 
1.1 Background to the Research ........................................................................................1 
1.2 Research Problem and Propositions ............................................................................1 
1.3 Justification for the Research ......................................................................................4 
1.4 Research Context .........................................................................................................8 
1.5 Research Method .......................................................................................................10 
1.6 Delimitation and Scope of the Thesis........................................................................11 
1.7 Outline of the Thesis .................................................................................................12 
1.8 Chapter Summary ......................................................................................................14 
CHAPTER 2. Literature Review ...................................................................................... 15 
2.1 Introduction ...............................................................................................................15 
2.2 Social Media ..............................................................................................................15 
2.2.1 Social Media Definitions ................................................................................16 
2.2.2 Types of Social Media ....................................................................................19 
2.2.3 Social Networking Sites ..................................................................................20 
2.3 Uses and Gratifications Theory .................................................................................24 
2.3.1 Internet Uses and Gratifications ......................................................................26 
2.3.2 Social Media Gratifications ............................................................................27 
2.4 Customer Engagement ..............................................................................................35 
2.4.1 Customer Engagement Theoretical Foundations ............................................36 
2.4.2 Engagement Conceptualisation .......................................................................37 
2.4.3 Customer Engagement Related Concepts, Antecedents and Consequences...44 
2.4.4 Dimensions of Customer Engagement ............................................................51 
2.4.5 Customer Engagement Behaviour ..................................................................51 
2.5 Chapter Summary ......................................................................................................53 
i|Page
CHAPTER 3. Social Media Engagement Behaviour ...................................................... 54 
3.1 Introduction ...............................................................................................................54 
3.2 Social Media Engagement Behaviour .......................................................................55 
3.2.1 Definition of Social Media Engagement Behaviour .......................................55 
3.2.2 Social Media Engagement Behaviour Intensity ..............................................55 
3.2.3 Social Media Engagement Behaviour Valence ...............................................56 
3.2.4 Social Media Engagement Behaviour Construct ............................................57 
3.2.5 Social Media Engagement Behaviour Typologies ..........................................62 
3.3 Chapter Summary ......................................................................................................71 
CHAPTER 4. Conceptual Model Development .............................................................. 73 
4.1 Introduction ...............................................................................................................73 
4.2 The Conceptual Model ..............................................................................................73 
4.3 Hypotheses ................................................................................................................75 
4.3.1 Informational Content .....................................................................................75 
4.3.2 Entertaining Content .......................................................................................77 
4.3.3 Remunerative Content.....................................................................................78 
4.3.4 Relational Content...........................................................................................81 
4.3.5 Simultaneous Presence of Social Media Content Categories .........................82 
4.3.6 Moderating Variables ......................................................................................83 
4.4 Chapter Summary ......................................................................................................88 
CHAPTER 5. Research Design ......................................................................................... 90 
5.1 Introduction ...............................................................................................................90 
5.2 The Research Objective and Questions .....................................................................90 
5.3 Philosophical Stance..................................................................................................91 
5.4 The Research Methods ..............................................................................................92 
5.4.1 Context of the Study .......................................................................................92 
5.4.2 The Research Design ......................................................................................94 
5.4.3 Data Collection Sources ..................................................................................95 
5.4.4 Data Collection ...............................................................................................96 
5.5 Content Analysis .....................................................................................................100 
5.5.1 Defining Content Analysis ............................................................................100 
5.5.2 Purpose of Content Analysis .........................................................................101 
5.5.3 Content Analysis Process ..............................................................................101 
5.6 Descriptive Results ..................................................................................................121 

ii | P a g e
5.6.1 Social Media Content ....................................................................................121 
5.6.2 Social Media Engagement Behaviour ...........................................................123 
5.6.3 Moderating Variables ....................................................................................125 
5.6.4 Control Variables ..........................................................................................126 
5.7 Hypothesis Testing ..................................................................................................129 
5.7.1 Binary Logistic Regression ...........................................................................129 
5.7.2 Process Analysis ...........................................................................................131 
5.8 Chapter Summary ....................................................................................................133 
CHAPTER 6. Results ....................................................................................................... 135 
6.1 Introduction .............................................................................................................135 
6.2 Social Media Content ..............................................................................................135 
6.2.1 Informational Content Presence ....................................................................136 
6.2.2 Informational Content Level .........................................................................145 
6.2.3 Entertaining Content Presence ......................................................................148 
6.2.4 Entertaining Content Level ...........................................................................155 
6.2.5 Remunerative Content Presence ...................................................................157 
6.2.6 Remunerative Content Level ........................................................................166 
6.2.7 Relational Content Presence .........................................................................168 
6.2.8 Relational Content Level ..............................................................................176 
6.2.9 Social Media Content Presence Summary ....................................................178 
6.2.10 Social Media Content Level Summary .......................................................180 
6.3 Interaction Effects ...................................................................................................182 
6.3.1 Informational and Entertaining Content Interaction .....................................184 
6.3.2 Informational and Relational Content Interaction ........................................186 
6.3.3 Informational and Remunerative Content Interaction ..................................187 
6.3.4 Entertaining and Remunerative Content Interaction .....................................187 
6.3.5 Entertaining and Relational Content Interactions .........................................188 
6.3.6 Relational and Remunerative Content Interaction ........................................189 
6.3.7 Interaction Effects Summary ........................................................................190 
6.4 Moderation ..............................................................................................................191 
6.4.1 Hayes PROCESS Moderation Model with Three Category Moderator .......192 
6.4.2 Media Richness .............................................................................................192 
6.4.3 Congruity ......................................................................................................197 
6.4.4 Community Size ............................................................................................206 

iii | P a g e
6.4.5 Moderation Effect Summary .........................................................................218 
6.5 Summary of Results ................................................................................................220 
6.6 Chapter Summary ....................................................................................................221 
CHAPTER 7. Discussion and Conclusion ...................................................................... 222 
7.1 Introduction .............................................................................................................222 
7.2 Contributions of the Research .................................................................................222 
7.2.1 Development of the SMEB Construct ..........................................................222 
7.2.2 Application of the UGT Perspective to Engagement ....................................225 
7.2.3 Establishment of the Relationship between Social Media Content and
SMEB .....................................................................................................................227 
7.2.4 Social Media Data Analytics.........................................................................230 
7.3 Limitations...............................................................................................................232 
7.4 Directions for Future Research................................................................................235 
7.4.1 User Progression through SMEB ..................................................................235 
7.4.2 Identification of Further Antecedents to SMEB ...........................................236 
7.4.3 Investigation of SMEB Consequences ..........................................................238 
7.4.4 Incorporation of the Three Dimensional View of Customer Engagement ...239 
7.5 Managerial Implications ..........................................................................................240 
7.5.1 High Level of Dormancy and Low Engagement Rates among Users ..........240 
7.5.2 Enhancing Engagement through Strategic Content Design ..........................241 
7.6 Concluding Thoughts ..............................................................................................247 
Appendices ...................................................................................................................... 249 
Appendix A: Email to Participating Wine Brands ........................................................249 
Appendix B: NVivo10 Word Frequency Report ...........................................................250 
Appendix C: Word Search Formulas for Post Content Coding ....................................254 
Appendix D: Kappa Coefficient Calculation ................................................................265 
Appendix E: Binary Logistic Regression Results (Interactions)...................................267 
References ...................................................................................................................... 289 

iv | P a g e
List of Figures
Figure 1.1 Conceptual Framework.................................................................................... 2 
Figure 3.1 Social Media Engagement Behaviour Construct ........................................... 59 
Figure 4.1 Conceptual Model of Social Media Content and Engagement
Behaviour ....................................................................................................... 74 
Figure 5.1 Quantitative Content Analysis Process........................................................ 102 
Figure 5.2 PROCESS Model 2 Conceptual Diagram ................................................... 132 
Figure 5.3 PROCESS Model 2 Statistical Diagram ...................................................... 132 
Figure 6.1 A Visual Representation of the Moderation of the Effect of
Informational Content on Contributing Behaviour (Shares) by Richness ... 196 
Figure 6.2 Visual Representation of the Moderation of the Effect of Informational
Content on Contributing Behaviour (Likes) By Congruity. ........................ 201 
Figure 6.3 Visual Representation of the Moderation of the Effect of Entertaining
Content on Creating Behaviour by Congruity. ............................................ 204 
Figure 6.4 Visual Representation of the Moderation of the Effect of Entertaining
Content on Contributing Behaviour (Likes) by Community Size. .............. 210 
Figure 6.5 A Visual Representation of the Moderation of the Effect of
Entertaining Content on Contributing Behaviour (Shares) by
Community Size........................................................................................... 213 
Figure 6.6 A Visual Representation of the Moderation of the Effect of
Entertaining Content on Dormant Behaviour by Community Size. ............ 216 

v|Page
List of Tables
Table 2.1 Recent Studies with a UGT Perspective Applied to Online Media Use ......... 28 
Table 2.2 Customer Engagement Definitions ................................................................. 39 
Table 2.3 Customer Engagement Conceptual Relationships .......................................... 48 
Table 3.1 Positively- and Negatively-Valenced Social Media Engagement
Behaviours ...................................................................................................... 60 
Table 4.1 Hypotheses ...................................................................................................... 89 
Table 5.1 Facebook Insights and NCapture Data Metrics .............................................. 97 
Table 5.2 Brand Profiles ................................................................................................. 98 
Table 5.3 Facebook Insights Post Metrics ...................................................................... 98 
Table 5.4 Number of Comments by Brand ..................................................................... 99 
Table 5.5 Social Media Content Categories.................................................................. 105 
Table 5.6 Informational Content Codes ........................................................................ 106 
Table 5.7 Entertaining Content Codes .......................................................................... 108 
Table 5.8 Remunerative Content Codes ....................................................................... 109 
Table 5.9 Relational Content Codes ............................................................................. 110 
Table 5.10 Social Media Engagement Behaviour Operationalisation .......................... 112 
Table 5.11 Media Richness Operationalisation ............................................................ 113 
Table 5.12 Congruity Operationalisation ...................................................................... 114 
Table 5.13 Community size operationalization ............................................................ 114 
Table 5.14 Image Coding Scheme ................................................................................ 115 
Table 5.15 Kappa Value Interpretation ......................................................................... 120 
Table 5.16 Post Content Categories .............................................................................. 121 
Table 5.17 Descriptive Statistics for Social Media Engagement Behaviour ................ 124 
Table 5.18 Media Richness ........................................................................................... 125 
Table 5.19 Congruity .................................................................................................... 125 
Table 5.20 Community Size.......................................................................................... 126 
Table 5.21 Post Distribution by Week .......................................................................... 127 
Table 5.22 Post Distribution by 12 Months .................................................................. 127 
Table 5.23 Post Distribution by Hour ........................................................................... 128 
Table 5.24 Independent Variable Coding ..................................................................... 130 
Table 5.25 Dependent Variable Coding ........................................................................ 130 

vi | P a g e
Table 5.26 Control Variable Coding ............................................................................. 130 
Table 6.1 Logistic Regression Showing Effect of Informational Content Presence
on Creating Behaviour .................................................................................. 138 
Table 6.2 Logistic Regression Showing Effect of Informational Content Presence
on Contributing Behaviour (Shares) ............................................................. 140 
Table 6.3 Logistic Regression Showing Effect of Informational Content Presence
on Contributing Behaviour (Likes)............................................................... 141 
Table 6.4 Logistic Regression Showing Effect of Informational Content Presence
on Consuming Behaviour ............................................................................. 142 
Table 6.5 Logistic Regression Showing Effect of Informational Content Presence
on Dormant Behaviour ................................................................................. 143 
Table 6.6 Logistic Regression Showing Effect of Informational Content Presence
on Detaching Behaviour. .............................................................................. 144 
Table 6.7 Summarised Logistic Regression Results for H1 ......................................... 145 
Table 6.8 Logistic Regression Showing Effect of Informational Content Level on
SMEB ........................................................................................................... 146 
Table 6.9 Logistic Regression Showing Effect of Entertaining Content Presence
on Creating Behaviour .................................................................................. 149 
Table 6.10 Logistic Regression Showing Effect of Entertaining Content Presence
on Contributing (Likes) Behaviour............................................................... 150 
Table 6.11 Logistic Regression Showing Effect of Entertaining Content Presence
on Contributing (Shares) Behaviour ............................................................. 151 
Table 6.12 Logistic Regression Showing Effect of Entertaining Content Presence
on Consuming Behaviour ............................................................................. 152 
Table 6.13 Logistic Regression Showing Effect of Entertaining Content Presence
on Dormant Behaviour ................................................................................. 153 
Table 6.14 Logistic Regression Showing Effect of Entertaining Content Presence
on Detaching Behaviour ............................................................................... 154 
Table 6.15 Summarised Logistic Regression Results for H2 ....................................... 155 
Table 6.16 Logistic Regression Showing Effect of Entertaining Content Level on
SMEB ........................................................................................................... 156 
Table 6.17 Logistic Regression Showing Effect of Remunerative Content
Presence on Creating Behaviour................................................................... 159 
Table 6.18 Logistic Regression Showing Effect of Remunerative Content
Presence on Contributing Behaviour (Shares).............................................. 160 
Table 6.19 Logistic Regression Showing Effect of Remunerative Content
Presence on Contributing Behaviour (Likes) ............................................... 161 

vii | P a g e
Table 6.20 Logistic Regression Showing Effect of Remunerative Content
Presence on Consuming Behaviour .............................................................. 162 
Table 6.21 Logistic Regression Showing Effect of Remunerative Content on
Dormant Behaviour ...................................................................................... 164 
Table 6.22 Logistic Regression Showing Effect of Remunerative Content on
Detaching Behaviour .................................................................................... 165 
Table 6.23 Summarised Logistic Regression Results for H3 ....................................... 166 
Table 6.24 Logistic Regression Showing Effect of Remunerative Content Level
on SMEB ...................................................................................................... 167 
Table 6.25 Logistic Regression Showing Effect of Relational Content Presence on
Creating Behaviour ....................................................................................... 170 
Table 6.26 Logistic Regression Showing Effect of Relational Content Presence on
Contributing Behaviour (Shares) .................................................................. 171 
Table 6.27 Logistic Regression Showing Effect of Relational Content Presence on
Contributing Behaviour (Likes).................................................................... 172 
Table 6.28 Logistic Regression Showing Effect of Relational Content Presence on
Consuming Behaviour .................................................................................. 173 
Table 6.29 Logistic Regression Showing Effect of Relational Content Presence on
Dormant Behaviour ...................................................................................... 174 
Table 6.30 Logistic Regression Showing Effect of Relational Content Presence on
Detaching Behaviour .................................................................................... 175 
Table 6.31 Summarised Logistic Regression Results for H4 ....................................... 176 
Table 6.32 Logistic Regression Showing Effect of Relational Content Level on
SMEB ........................................................................................................... 177 
Table 6.33 Binary Logistic Regression Results for Social Media Content and Social
Media Engagement Behaviour ..................................................................... 179 
Table 6.34 Interaction Effects Summary ...................................................................... 184 
Table 6.35 Richness Operationalisation........................................................................ 193 
Table 6.36 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Shares) by
Media Richness............................................................................................. 194 
Table 6.37 PROCESS Data for Visualising Conditional Effect of Informational
Content on Contributing Behaviour (Shares) ............................................... 194 
Table 6.38 Conditional Effect of Informational Content on Contributing
Behaviour (Shares) At Values of the Moderator .......................................... 195 
Table 6.39 Congruity Operationalisation ...................................................................... 198 

viii | P a g e
Table 6.40 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Likes) by
Congruity ...................................................................................................... 199 
Table 6.41 PROCESS Data for Visualising Conditional Effect of Informational
Content on Contributing Behaviour (Likes) ................................................. 200 
Table 6.42 Conditional Effect of Informational Content on Contributing
Behaviour (Likes) At Values of the Moderator ............................................ 200 
Table 6.43 Partial Output from PROCESS Model 2 Examining the Moderation of
the Effect of Entertaining Content on Creating Behaviour by Congruity .... 202 
Table 6.44 PROCESS Data for Visualising Conditional Effect of Entertaining
Content on Creating Behaviour .................................................................... 203 
Table 6.45 Conditional Effect of Entertaining Content on Creating Behaviour at
Values of the Moderator ............................................................................... 204 
Table 6.46 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Likes) by
Community Size ........................................................................................... 207 
Table 6.47 PROCESS Data for Visualising Conditional Effect of Entertaining
Content on Contributing Behaviour (Likes) ................................................. 208 
Table 6.48 Conditional Effect of Entertaining Content on Contributing Behaviour
(Likes) at Values of the Moderator............................................................... 209 
Table 6.49 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Shares) by
Community Size. .......................................................................................... 211 
Table 6.50 PROCESS Data for Visualising Conditional Effect of Informational
Content on Contributing Behaviour (Shares) ............................................... 212 
Table 6.51 Conditional Effect of Entertaining Content on Contributing Behaviour
(Shares) At Values of The Moderator .......................................................... 212 
Table 6.52 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Dormant Behaviour by Community
Size ............................................................................................................... 214 
Table 6.53 PROCESS Data for Visualising Conditional Effect of Entertaining
Content on Dormant Behaviour .................................................................... 215 
Table 6.54 Conditional Effect of Entertaining Content on Dormant Behaviour at
Values of The Moderator .............................................................................. 215 
Table 6.55 Summary of Hypotheses and Results.......................................................... 220

ix | P a g e
Abstract
The proliferation of social media platforms in recent years has precipitated a paradigm

shift among consumers, as they become more proactive in their direct interactions with

brands. Practitioners recognise the value of these interactions, and are endeavouring to

build engagement through their social media content. However, despite recent research

in this field, theoretically-based academic guidance on a strategic approach to

developing engagement in new-media social networks remains limited. In addition,

while the Uses and Gratifications theoretical perspective has long claimed that media

users are motivated by a need to engage with content, it is unclear whether this

perspective can explain the engagement of customers in a social media context.

This dissertation aims to advance existing knowledge on social media content types by

examining the effect of informational, entertaining, remunerative and relational content

on the engagement behaviour of social media users. A social media engagement

behaviour (SMEB) construct is developed to provide a richer understanding of the

nature of engagement behaviour in this context. This construct includes six discrete

levels of behavioural intensity that recognise the positively- and negatively-valenced

nature of engagement behaviour.

This study used Facebook Insights and NCapture to extract data from Facebook to

provide insight into the actual behaviours of consumers using social media, rather than

relying on self-reported data to examine the proposed hypotheses. Social media data

was collected from twelve Australian wine brands, yielding a total of 2,236 social media

posts. Quantitative content analysis (QCA), binary logistic regression, and Process

moderation analysis were used to analyse the set of data and establish the significance

of the hypothesised relationships.

x|Page
The results show that the four social media content types have distinct and independent

effects on SMEB, demonstrating the need to consider each individually. Supported by

the notion of information overload, the results demonstrate that for each type of content,

the positive relationship with SMEB only exists at lower levels of each content type.

This demonstrates that the amount of content is an important consideration impacting

on the resultant engagement behaviour. Minimal interaction effects among content types

were found, which suggests that there is little benefit in designing social media content

that attempts to simultaneously appeal to users’ needs for information, entertainment,

remuneration and relational interaction. The results also showed significant moderating

effects of media richness, community size, and congruity of the social media content,

which affect the relationships with SMEB.

This study contributes to our knowledge of engagement by exploring online

engagement behaviour in greater depth and integrating specific levels and valence of

behaviour into a singular construct. It extends the utility of Uses and Gratifications

Theory in engagement research, demonstrating how this theory can be evolved to

explore emerging media such as social networking sites. The study supports the need

for the strategic design of social media content in business by linking specific types of

content to different aspects of SMEB. In doing so, it provides guidance to managers on

delivering social media content to enhance engagement among social media users.

xi | P a g e
Declaration

I certify that this work contains no material which has been accepted for the award of
any other degree or diploma in my name in any university or other tertiary institution
and, to the best of my knowledge and belief, contains no material previously published
or written by another person, except where due reference has been made in the text. In
addition, I certify that no part of this work will, in the future, be used in a submission in
my name, for any other degree or diploma in any university or other tertiary institution
without the prior approval of the University of Adelaide and where applicable, any
partner institution responsible for the joint-award of this degree. I give consent to this
copy of my thesis when deposited in the University Library being made available for
loan and photocopying, subject to the provisions of the Copyright Act 1968.

The author acknowledges that copyright of published works contained within this thesis

resides with the copyright holder(s) of those works.

I also give permission for the digital version of my thesis to be made available on the
Internet, via the University’s digital research repository, the Library Search and also
through Internet search engines, unless permission has been granted by the University to
restrict access for a period of time.

Signed: ____________________________________________

Date: 23rd December, 2015

xii | P a g e
Publications
The following publications are based upon the research presented in this thesis, and may

contain results and materials presented herein.

Dolan, R., Conduit, J., Fahy, J., and Goodman, S. (forthcoming) “Social Media
Engagement Behaviour: A Uses and Gratifications Perspective” Journal of Strategic
Marketing.

Dolan, R., Conduit, J., and Fahy, J. (forthcoming) “Social Media Engagement: A
Construct of Positively and Negatively Valenced Engagement Behaviours” in R.
Brodie, L.Hollebeek and J.Conduit, (Eds.) Customer Engagement: Contemporary Issues
and Challenges. Routledge.

Dolan, R., Conduit, J., Fahy, J., and Goodman, S. (forthcoming) “Big Social Data and
Social Media Analytics: Tools for exploring Social Media Engagement Behaviour”
Australian and New Zealand Marketing Academy Conference, Sydney Australia.

Dolan, R., Conduit, J., Goodman. S., and Fahy, J (forthcoming) “Facebook for Wine
Brands: An Analysis of Strategies for Facebook Posts and User Engagement Actions”
Academy of Wine Business Research Conference, Adelaide Australia.

Dolan, R., Conduit, J., Fahy, J., and Goodman, S. (2014) “Customer Brand Engagement
Behaviour in Online Social Networks: a Conceptual Framework” Australian and New
Zealand Marketing Academy Conference, Brisbane Australia.

xiii | P a g e
Acknowledgements
This PhD has been a challenging, but enjoyable journey which would not have been

possible without the support of many special people.

Firstly, I would like to thank my supervisors Professor. John Fahy, Dr. Jodie Conduit,

and Dr. Steve Goodman.

John provided me with the clarity, direction and confidence that was needed to take my

research to a higher level. His constant support and belief in me, and my research,

facilitated an extremely productive and enjoyable experience. Jodie has been an

inspiring supervisor, mentor, friend, and teacher. I simply cannot thank her enough for

the many ways in which she has encouraged me to constantly achieve more. I admire

her passion and enthusiasm for research, and the advice and skills that she has shared

with me go well beyond the scope of this PhD. I am sure that they will stay with me

throughout my career. I would not have started my Ph.D. if it was not for the rewarding

experience that I shared during my honours year with Dr. Steve Goodman. Steve has

provided me with so many fantastic teaching and research opportunities, which I am

very grateful for. The assistance of Ray Adams in editing this thesis is also gratefully

acknowledged.

I am very grateful to the many people who have provided me with advice during

presentations and doctoral colloquiums. A special thank you to Dr. Chris Medlin, Dr.

Carolin Plewa, and Professor Rod Brodie for their valuable support and guidance.

I would like to thank the wonderful friends that I have made throughout my PhD

candidature. Thank you to Teagan, who spent many hours at Cibo with me, which

always meant starting my day with a guaranteed laugh. A special thanks to Hande for

xiv | P a g e
her contagious enthusiasm and spirit, and the many hours that she spent helping me with

my thesis. Thank you also to Ervin. We have shared so many wonderful and interesting

experiences since the beginning of our honours year. Ervin has always been there for

me, and I don’t think I would have survived this Ph.D. without his infectious smile and

evil laugh.

I would also like to thank my family for their love, inspiration, understanding and

never-ending support and belief in me. They have always pushed me to achieve

everything that I set out to. The last 8 years of study would definitely not have been

possible without your guidance, love and generosity. I would also like thank my best

friend and sister, Sarah, who has survived being my housemate and tutor for many

years. We share such a special bond, without which I don't think either of us would be

able to survive the pressure of our studies. A special thank you also to my Gran and

Grandpa, for the many years of Monday night dinners which have provided me with a

wealth of encouragement, love, and laughter that I will forever cherish.

To the Marsland family, thank you for being such a significant source of support

throughout my studies, and for providing me with so much happiness and laughter.

A special thanks to Ian for the many hours that he spent proof-reading my chapters.

Finally, I would like to thank Ryan. The gratitude I feel for everything that you have

done for me is beyond words. Your constant encouragement and positivity throughout

this entire process has been so wonderful. You have provided me with endless coffee,

laughter, and love, for which I will be forever grateful.

xv | P a g e
Key Terms, Definitions and Abbreviations
The definitions of selected terms are listed to provide clarity and to set certain

terminologies for the context in which they were utilised in this thesis;

Uses and Gratifications Theory (UGT): An approach to understanding why and how

people actively seek out specific media to satisfy specific needs. UGT is an audience-

centred approach to understanding mass communication. It assumes that audience

members are not passive consumers of media. Rather, the audience participants have

power over their media consumption and assume an active role in interpreting and

integrating media into their own lives (Severin & Tankard, 1997).

Social Networking Sites (SNS): Web-based services that allow individuals to construct

a public or semi-public profile within a bounded system, articulate a list of other users

with whom they share a connection, and view and traverse their list of connections and

those made by others within the system (Boyd and Ellison, 2008, p. 210).

Social Media Content: Social media content in this thesis refers to the content of posts

to users, made by brands via Facebook. This content is categorised into four types:

informational, entertaining, remunerative and relational.

Customer Engagement Behaviour (CEB): defined as “a customer’s behavioural

manifestations that have a brand or firm focus, beyond purchase” (van Doorn et al.,

2010 p.254).

Social Media Engagement Behaviour (SMEB): Social media engagement behaviours

go beyond transactions, and may be specifically defined as a customer’s behavioural

manifestations that have a social media focus, beyond purchase, resulting from

motivational drivers.

xvi | P a g e
Positively-Valenced Social Media Engagement Behaviour: is reflected in favourable

or affirmative user behaviours. This thesis categorises three positively-valenced social

media engagement behaviours: consuming, contributing and creating.

Negatively-Valenced Social Media Engagement Behaviour: Negatively-valenced

engagement behaviour is exhibited through unfavourable behaviours directed towards

the brand (Hollebeek and Chen, 2014). This thesis categorises two negatively-valenced

engagement behaviours in the social media context: detaching and destructing.

Creating SMEB: Users engage with brands and other users by creating positively-

valenced content on social media platforms. Creating epitomises a highly active level of

SMEB. Creating users exhibit specific creating behaviours of knowledge seeking,

sharing experiences, advocating, socialising, co-developing and affirming.

Contributing SMEB: Users contribute to existing content in social media platforms.

Contributing users exhibit a moderate level of positively-valenced SMEB.

Consuming SMEB: Users passively consume content without any form of active

reciprocation or contribution. Consuming users demonstrate a minimum level of

positive, passive SMEB.

Dormant SMEB: A temporary state of inactive, passive engagement by users who may

have previously interacted with the focal brand.

Detaching SMEB: Users take action to remove content of the brand appearing in their

news-feed or equivalent home page. Detaching users exhibit a moderate level of

negatively-valenced SMEB.

xvii | P a g e
Destructing SMEB: Negative, active contributions to existing content on social media

platforms are created by destructive users. Destructive users represent a highest level of

negatively-active SMEB.

Facebook Insights: Facebook Insights is a tool provided to administrators of Facebook

brand pages to enable high-level monitoring of the activities on the Facebook page.

Facebook Insights allows administrators to download data concerning the performance

of a social media post.

NCapture: NCapture is a web browser extension developed by QSR International. It

allows researchers to quickly and easily capture content including web pages, online

PDF’s and social media for analysis within NVivo 10.

Quantitative Content Analysis (QCA): This research follows Neuendorf’s (2002)

approach to quantitative content analysis (QCA), suitable for this study due to its focus

on summarising the quantitative analysis of messages. Content analysis is most

commonly defined as a research technique for the objective, systematic and quantitative

description of the manifest content of communication (Berelson, 1952).

xviii | P a g e
Chapter 1: Introduction

CHAPTER 1. Introduction
1.1 Background to the Research
The focus of this thesis is to investigate and explain the role of social media content in

facilitating engagement behaviours within social media platforms. In doing so, it

contributes a deeper recognition of the nature and dynamics of engagement behaviour

specifically in a social media context. It explores both positively-valenced social media

engagement behaviour (SMEB) and distinguishes this from neutral and negatively-

valenced SMEB. It examines the role of social media content in the form of

informational, entertaining, remunerative and relational posts to achieve these types of

engagement behaviour. The four categories of social media content which facilitate

expressions of SMEB are derived from the underpinning perspective of Uses and

Gratifications Theory (Katz & Foulkes, 1962). This theory suggests that individuals are

motivated to engage with media by their needs for information, entertainment,

economic incentive or reward and a need for social and relational interaction.

1.2 Research Problem and Objectives


The aim of the research is to investigate how uses and gratifications theory (UGT)

contributes to the knowledge and understanding of the influence of social media content

on SMEB. Specifically, the research characterises a new construct of positively- and

negatively-valenced SMEB. This construct is the dependent variable of the study and is

measured through quantitative analysis of SMEB derived from Facebook data. The

independent variables include social media content categories, underpinned by UGT,

defined as informational content, entertaining content, remunerative content, and

relational content.

1|Page
Chapter 1: Introduction
The specific objectives of this research are:

1. To gain a deeper understanding of social media content categories through the

application of the UGT perspective.

2. To offer new insights into the characteristics, levels and valence of social media

engagement behaviour.

3. To determine the impact of social media content (categorised as informational,

entertaining, remunerative and relational) on social media engagement

behaviour.

The conceptual framework presented in Figure 1.1 highlights these research objectives

and presents them together to depict their interrelationships. A detailed explanation and

theoretical justification for these variables is provided in Chapter 4, Conceptual Model

Development.

Figure 1.1 Conceptual Framework

2|Page
Chapter 1: Introduction
Social media content includes four categories; informational, entertaining, remunerative

and relational content. These social media content categories are derived from the

underpinning theoretical foundations of UGT. They are the proposed drivers or

antecedents of SMEB as depicted in Figure 1.1. The underpinning UGT perspective is

discussed in Chapter 2, Section 2.3, leading to the identification of the four categories of

social media content.

The SMEB construct is the outcome variable which identifies and explicates the

different types of engagement behaviour that users exhibit in social media platforms. It

demonstrates that SMEB consists of six distinct types; creating, contributing,

consuming, dormancy, detaching and destructing. While creating, contributing and

destructing represent active engagement behaviours that potentially impact on other

social media users, consuming, dormancy and detachment are more passive and/or

individualised forms of engagement. The SMEB construct also recognises that SMEB

may be positively- or negatively-valenced in nature.

The relationship between social media content and SMEB is supported by the UGT

perspective. UGT provides a framework through which the motivations of individuals

seeking a specific type of media can be further understood. In a social media context,

users are not passive in their media selection or their use of specific media. Social

media is constructed to enable customer interaction and engagement. Therefore, it is

imperative to expand the application of UGT to determine the engagement behaviour

that results from the selection of and interaction with different types of social media

content. It is proposed that social media content which satisfies the needs for

information, entertainment, remuneration and relational interaction will facilitate the

way in which consumers choose to engage with brands and other network users within
3|Page
Chapter 1: Introduction
social media sites.

There are three proposed moderating variables in the conceptual framework: media

richness, congruity of the content and community size. Each is briefly defined and

further explicated within Chapter 4: Conceptual Model Development.

Media Richness relates to the richness of the social media content measured as low,

medium or high, and is hypothesised to partly moderate the relationship between social

media content and SMEB.

Community size refers to the size of the social media community, measured by the total

number of Facebook page likes at the time of the data collection. The size of the

community is hypothesised to partly moderate the relationship between social media

content and SMEB.

Congruity of the social media content relates to the extent to which the content is related

to the focal brand. Measured as low, medium or high congruity, this variable is

hypothesised to partly moderate the relationship between social media content and

SMEB.

1.3 Justification for the Research


The emergence of social media platforms and increasing customer adoption of these

platforms has precipitated a paradigm shift, significantly altering the way customers

communicate and interact with each other and with brands. There are more than one

billion members of Facebook, and Twitter now has more than 280 million monthly

active users (Stieglitz, Dang-Xuan, Bruns, & Neuberger, 2014). The interactive

properties of social media have transformed consumers from passive observers of

4|Page
Chapter 1: Introduction
content to active participants who create vast quantities of user generated content

through their conversations, interactions and behaviours online.

Central to this paradigm shift is the concept of customer engagement, which recognises

that customers can co-create value, co-create strategy and collaborate in the firm’s

innovation process (Bijmolt, Leeflang, Block, Eisenbeiss, Hardie, Lemmens, & Saffert,

2010). Business environments have therefore become more dynamic and interactive,

with customers seeking participation and engagement with unique offerings and

activities of the organisation (Vivek, Beatty, & Morgan, 2012). With the growing

prevalence of social media there has been an emergent focus from both academics and

practitioners on the concept of engagement in social media platforms (Brodie, Ilic,

Juric, & Hollebeek, 2013). Social media platforms provide users with an interactive

avenue to create value and engage with the firm (Brodie et al., 2013; Gummerus,

Liljander, Weman, & Pihlström, 2012).

Social media has become a mainstream media platform that connects one-third of the

world’s population (Nelson-Field & Taylor, 2012). It offers advertisers access to eighty

per cent of global consumer expenditures, a $29 trillion market (Nuttney, 2010). Over

15 million brands globally are registered with the social media site, Facebook (Koetsier,

2013). Customers are inundated with a proliferation of messages and marketers will not

succeed in their efforts without an understanding of how to effectively engineer their

content to facilitate customer engagement (Lee, Hosanagar, & Nair, 2013). Practitioners

have largely been at the forefront of efforts to advise businesses on their social media

strategy. Whilst the list of guidelines and strategies for social media marketing efforts

appears endless, academic research and empirical evidence in the area remains scarce.

5|Page
Chapter 1: Introduction
The notion of engagement has been studied in many fields, including psychology

(Hallberg & Schaufeli, 2006), education (Baron & Corbin, 2012) and management

(Saks, 2006). A recent focus in marketing has centred on customer engagement with a

brand (Hollebeek, 2011b; van Doorn, Lemon, Mittal, Nass, Pick, Pirner, & Verhoef,

2010). Customers engage with a firm or brand through multiple touch-points and

service encounters. Examples of this engagement include interactions with staff, use of

products, physical retail spaces, social media pages and other forms of communication.

Previous authors recognise that there are various focal objects of customer engagement

including product or service offerings (Brodie, Hollebeek, Jurić, & Ilić, 2011), activities

and events (Vivek et al., 2012), and media (Calder, Malthouse, & Schaedel, 2009).

Together these interactions constitute the brand experience of the customer.

Engagement is interactive and therefore context-dependent and can only be properly

understood through an examination of each of these service experiences (Brodie et al.,

2011; Calder et al., 2009; Gummerus et al., 2012).

This thesis explores the behavioural manifestation of the engagement concept,

consistent with previous studies of engagement and social media (e.g. Gummerus et al.,

2012, van Doorn et al., 2010). Customer engagement behaviour is defined in this thesis

as “a customer’s behavioural manifestations that have a brand or firm focus, beyond

purchase” (van Doorn et al., 2010 p.254). Customer engagement behaviour involves

customers’ voluntary resource contributions that have a brand or firm focus, but go

beyond what is fundamental to transactions. These contributions occur in interaction

between the focal engagement object and/or other actors and result from motivational

drivers (Jaakkola & Alexander, 2014).

6|Page
Chapter 1: Introduction
While recent research has explored both the antecedents and consequences of customer

engagement (Gambetti, Graffigna, & Biraghi, 2012; van Doorn et al., 2010), studies that

consider engagement with social media are only beginning to emerge. There is a need to

develop a theoretical understanding of the nature of engagement behaviour in response

to marketing practices within a social network structure (Sashi, 2012). This thesis

explains the role of social media content in facilitating engagement behaviours within

social media platforms. It contributes a deeper recognition of the nature and dynamics

facilitating engagement behaviour within social media platforms. The conceptual model

presented in Chapter 4 explores the processes for stimulating positively-valenced

SMEB and/or dissuading neutral and negatively-valenced SMEB through the use of

social media content. As such, it addresses one of the challenges in the implementation

of organisational tactics and strategies centred on the increasing role of customers, and

focuses on an MSI (2014) key topic of interest. Specifically, the MSI (2014) advocates

research on customer behaviours in multi-media environments; the role of social media

within customer experiences; the conceptualisation, definition and measurement of

engagement; and how social media marketing activities create customer engagement

(MSI, 2014).

This thesis provides an examination of SMEB, which focuses attention on a singular

touch-point in the service experience. It therefore does not reflect customer brand

engagement in its entirety, but rather a singular component of that engagement.

Consistent with calls from previous researchers (Brodie et al., 2011; Vivek et al., 2012),

it is argued that this in-depth examination within a context-specific environment (e.g.

social media) will provide greater insight into the behavioural manifestations of

engagement.

7|Page
Chapter 1: Introduction
The rise of social media sites has also provided a new trail of data detailing customer

interactions and conversations for businesses and academics to explore and understand.

This thesis takes advantage of this emergence through its use of behavioural data

derived from the social media platform, Facebook. Access to social media data is said to

have disrupted traditional approaches to customer relationship management, causing a

need for organisations to consider how to build insights from the large quantities of data

made available by social media (Manovich, 2011). Effectively using this data enables

companies to derive valuable insights about their customers (Malthouse, Haenlein,

Skiera, Wege, & Zhang, 2013). Through understanding this form of data, managers can

measure and hence know radically more about their businesses and directly translate

that knowledge into improved decision making and performance (McAfee,

Brynjolfsson, Davenport, Patil, & Barton, 2012). However, within social media many

companies are unable to identify which activities attract and engage customers

(Malthouse et al., 2013). The major challenge for marketing practitioners and academics

is how to extract insights from these large quantities of data, and how to incorporate

them into models of customer engagement and social customer relationship

management (Bijmolt et al., 2010). In this thesis, this challenge is addressed by

demonstrating how this data can be effectively accessed and analysed in order to

provide an enhanced understanding of engagement.

1.4 Research Context


The context of this thesis is the social media presence of the Australian wine industry.

The Australian wine industry has grown at a phenomenal rate, with 2,573 wine

producers listed in 2014, compared to just 344 in 1983 (Winebiz, 2014). Considering

the high levels of competition within the Australian wine industry, it is not surprising

8|Page
Chapter 1: Introduction
that many brands are seeking new and innovative ways to communicate with

consumers. Social media sites such as Facebook have become an increasingly popular

customer touch point, with the viral and social capabilities of these online networks

creating a new forum for customer interaction with wine brands (Barber, Dodd, &

Ghiselli, 2008; Bulearca & Bulearca, 2010; Keller, 2009). Wine is an experiential

product (Bruwer & Alant, 2009), and as Australian wine brands compete to attract and

retain consumers, many are embracing social media to reach their consumers and

communicate their brand experience, quality and personality (Vinography, 2012).

Marketing practitioners within the wine industry have been quick to recognise the value

of social media platforms, rapidly integrating such platforms into the marketing mix

(Bergen, 2014; Sinclair, 2014; Stelzner, 2014). There are currently more than 2,500

Australian and New Zealand wineries with a presence on Facebook (Mastermind,

2015). However, many practitioners have identified a lack of awareness and knowledge

regarding effective social media strategy, creating a significant challenge as

practitioners navigate through this forum with little guidance and empirical

understanding (Stelzner, 2014). This study empirically explores how practitioners can

strategically design and engineer their social media content in order to facilitate

engagement amongst users.

Research has suggested that 90 percent of wine drinkers use Facebook for at least 6.2

hours per week (Breslin, 2013). Some wine brands have achieved success through

social media, with documented examples demonstrating that small and large wineries

have achieved a positive return on investment through the implementation of successful

social media strategies. For example, Pacific Rim Winery in the United States increased

their website traffic by 7000% and achieved a 15% increase in revenue, with a 73%
9|Page
Chapter 1: Introduction
increase in transactions (Moore, 2012). Several scholarly studies have explored social

media practices within the wine industry. Of wineries studied in Australia, Canada, New

Zealand, Spain, Italy, South Africa and the US, 35% have reported using social media,

with the primary reasons of communicating with customers about events at the winery,

and promoting wines (Alonso, Bressan, O'Shea, & Krajsic, 2013). Scholars have also

suggested that social media assists with wine sales as word of mouth is particularly

effective amongst wine consumers (Leigon, 2011), with the socialisation aspect of

social media acting as an appropriate fit with wine, allowing consumers to exchange

information and encourage others to try different wines (Wilson & Quinton, 2012).

However, the wine industry context has received little attention in the customer

engagement literature (Hollebeek & Brodie, 2009).

1.5 Research Method


This section provides a brief overview of the research method adopted in this thesis. A

detailed description and justification of the procedures is provided in Chapter 5.

A quantitative research approach is adopted in this thesis to investigate the relationships

between the theoretically developed constructs of social media content and SMEB. The

data used for the study was derived from the social media pages of Australian wine

brands. The data was collected by using Facebook Insights and NCapture. Social media

data was collected for a 12 month period. In total, 2,236 Facebook posts were collected

from the 1st of January, 2013 to the 31st of December, 2013. Social media content was

analysed using the process of quantitative content analysis (QCA) defined by Neuendorf

(2002). Following the process of QCA which categorised the 2,236 Facebook posts

according to their level of informational, entertaining, remunerative and relational

content, data analysis was conducted using SPSS v22. Binary logistic regression was
10 | P a g e
Chapter 1: Introduction
used to test the impact of social media content on SMEB. Hayes (2013) PROCESS

computational tool for path-based moderation was used to assess the moderating impact

of media richness, community size and congruity of the content on the relationship

between social media content and SMEB.

1.6 Delimitation and Scope of the Thesis


The scope of this thesis is limited to social media content and SMEB in the Australian

wine industry. The findings and implications of this research are relevant to similar

industry contexts, particularly those in which the product category may also be hedonic.

Further research is suggested in Chapter 7 which would explicate the relationship

between social media content and SMEB in unrelated contexts. Further examination of

the relationships among the key constructs is recommended in various industry settings,

in addition to studies across various social media platforms.

A three dimensional perspective of customer engagement has been widely accepted in

recent literature (Brodie et al., 2011; Hollebeek, 2011b) concerning the

conceptualisation and definition of what constitutes engagement. This perspective

includes recognition of the cognitive, emotional and behavioural dimensions of

engagement. Within this thesis, the focus is on one dimension of engagement:

behaviour. An inclusion of the cognitive and emotional aspects of engagement

experienced by social media users would provide a more comprehensive insight into

social media engagement.

A broader study, encapsulating individual user characteristics by way of a survey would

allow a greater investigation of the various factors which may also influence SMEB. By

extending the scope of the study, further antecedents to engagement, beyond social

11 | P a g e
Chapter 1: Introduction
media content may be identified. These antecedents are discussed in Chapter 2, and

include factors such as identification (Nambisan & Baron, 2007; Tsai, Huang, & Chiu,

2012), identity (Eisenbeiss, Blechschmidt, Backhaus, & Freund, 2012; van Doorn et al.,

2010), hedonism (Gambetti et al., 2012; Nambisan & Baron, 2007), interaction, rapport

(Hollebeek, 2011b), satisfaction (van Doorn et al., 2010) and trust (Bowden, 2009).

This thesis does not capture the extent to which these antecedents impact on SMEB.

SMEB is predicted in this thesis as a result of social media content. SMEB is shown to

vary in intensity as well as the extent to which it is positively- or negatively-valenced.

The examination provides a cross-sectional analysis of the resultant effects of social

media content on SMEB at one point in time. This thesis does not address the dynamic

nature of the levels of engagement behaviour in the SMEB construct. Analysis of user

progression through or within the six typologies is not considered. Engagement that

may occur in a cyclical form with reciprocal effects between antecedents and

consequences present (Fehrer, Woratschek, & Germelmann, 2013; Hollebeek, Glynn, &

Brodie, 2014) was not analysed in this thesis.

In the next section, the outline of the thesis is presented, including the key topics

addressed within each chapter.

1.7 Outline of the Thesis


The structure of this thesis is comprised of seven chapters.

CHAPTER 2 provides a review of the literature regarding social media, UGT and

customer engagement. The emergence of social media and specific social networking

sites is discussed. UGT is reviewed as an appropriate theoretical foundation of this

thesis. The theory supports the notion of users’ active selection and engagement with

12 | P a g e
Chapter 1: Introduction
specific focal objects, including social media content. The theoretical foundations of

customer engagement are discussed, including the importance of S-D logic and

relationship marketing. Recent conceptualisations of engagement are discussed, in

addition to a provision of the important distinction of engagement from its related

concepts, antecedents and consequences. This leads to a discussion of the dimensions of

customer engagement, with a specific focus on the behavioural dimension of customer

engagement.

CHAPTER 3 describes the SMEB construct developed by following the literature on

customer engagement behaviour and social media. This chapter provides an overview of

the importance of considering both the intensity and valence of engagement behaviour

in a social media setting. This leads to the development of six distinct SMEBs; creating,

contributing, consuming, dormancy, detaching and destructing.

CHAPTER 4 presents the conceptual model of the thesis, outlining the specific

hypothesised relationships between social media content and SMEB. These

relationships are underpinned by UGT.

CHAPTER 5 describes the research design used to identify and examine the

relationships between the key constructs. This chapter provides the philosophical stance

of the researcher, leading to the adoption of a quantitative approach to research.

Quantitative content analysis (QCA) is defined as an appropriate method within this

thesis, including a detailed overview of the QCA process and the descriptive results

resulting from the QCA. A discussion of the moderating variables is provided. The

processes for hypothesis testing including binary logistic regression and Hayes (2013)

moderation analysis are presented.

13 | P a g e
Chapter 1: Introduction
CHAPTER 6 addresses the main hypotheses of this thesis and presents the results.

Social media content is examined with reference to its impact on SMEB. Binary logistic

regression investigates the influence of the presence and level of social media content

type, informational, entertaining, remunerative and relational, on SMEB. The chapter

also demonstrates the moderating effects of community size, media richness and

congruity on the relationship between social media content and SMEB.

CHAPTER 7 integrates the key findings of the study with the relevant literature

addressed in Chapter 2. The discussion of the key findings leads an analysis of the

important contributions of the research. The managerial implications are discussed. The

chapter concludes with identifications of the limitations of the thesis, directions for

future research and concluding thoughts.

1.8 Chapter Summary


This chapter set out the foundations for this thesis. The background of the research was

introduced, leading to the identification of the research problem and objectives. The

justification of the research was provided; the research context and method were briefly

discussed. The delimitation and scope of the thesis were provided. The structure of the

thesis was provided. In the next chapter, the relevant literature concerning social media,

UGT and customer engagement is reviewed. This leads to the development of a new

construct of social media engagement behaviour which is presented in Chapter 3.

14 | P a g e
Chapter 2: Literature Review

CHAPTER 2. Literature Review


2.1 Introduction
This chapter begins with a review of the literature on social media and social

networking sites. This is followed by a discussion on the theoretical perspective of UGT

applied to user motivations for social media engagement. Relevant literature concerning

customer engagement theoretical foundations, conceptualisation, related concepts,

antecedents, consequences, and dimensions is then outlined. This leads to a focused

discussion on customer engagement behaviour, and its examination in a social media

context.

2.2 Social Media


The advent of social media has facilitated a fusion between sociology and technology,

shifting communication between individuals and firms from a monologue of one to

many, into a dialogue of many to many. As a result, social media have radically altered

the way individuals communicate, interact and manage relationships (Shirky, 2009).

Correspondingly, the lines of division between content providers and consumers have

begun to diminish (Giurgiu & Barsan, 2008).

The rise of social media channels in the past decade has enabled new forms of

customer/firm interaction. The role of social media within marketing has rapidly

developed in recent years, attracting interest in both academic and non-academic

literature. Social media has given consumers a rise in power, flexibility and visibility

regarding marketing content, changing the way individuals and organisations interact.

As a result, customers have transformed from passive receivers of marketing content to

active participants in the brand message (Mangold & Faulds, 2009). Interactive

15 | P a g e
Chapter 2: Literature Review
customer experiences through social media act as a significant influencing factor of

many consumer behaviour aspects, including information acquisition, purchase

behaviour and post-purchase communication (Mangold & Faulds, 2009).

Organisations are increasingly recognising and utilising this opportunity, with more

than 15 million brands registered with the social media site Facebook (Koetsier, 2013).

As businesses seek to communicate with customers through the social medium more

effectively, it offers a significant research area for scholars to better anticipate and

understand customer engagement in online social groups and subsequent brand-related

behaviours (Pagani, Hofacker, & Goldsmith, 2011; Pelling & White, 2009).

Despite significant academic and practitioner interest in the field of social media in

recent years, a lack of clarity remains evident regarding the precise definition of social

media, as discussed in the following section.

2.2.1 Social Media Definitions

The emergence of social media has powered many attempts to develop a definition of

the social media domain within the marketing literature. The term social media is a

construct derived from two underlying areas of research: communication science and

sociology (Peters, Chen, Kaplan, Ognibeni, & Pauwels, 2013). From the

communication science perspective, social media are a means for storing and delivering

information and data. Comparatively, from the perspective of sociology, social media

are viewed as social structures made up of a set of social actors linked by a complex set

of dynamic ties (Peters et al., 2013). Combining both perspectives, social media can be

described as “communication systems that allow their social actors to communicate

through multiple dyadic ties” (Peters et al., 2013, p.282). Hence, in contrast to

16 | P a g e
Chapter 2: Literature Review
traditional and other online media, social media are more egalitarian in nature. Unlike

traditional media platforms, social media resemble dynamic, interconnected, egalitarian,

and interactive organisms beyond the control of any organisation (Peters et al., 2013).

Rapid emerging technologies and communication forms alter the processes and

capabilities of social media, causing difficulty in its precise definition (Tuten, 2008).

Despite the lack of clarity in defining social media, most scholars agree that social

media is founded on participation and engagement (Mayfield, 2008). Participation

within social media occurs through the provision, sharing and discussion of user

generated content, through highly interactive mobile and web-based technology. This

concept lies at the centre of most attempts to define social media, and often definitions

incorporate a range of activities undertaken by the users of the social media page. To

illustrate, social media can be broadly defined as any “internet based applications that

help consumers share opinions, insights, experiences and perspectives” (Kaplan

and Haenlein, 2010, p. 565). Similarly, social media can be characterised as platforms

in which users have the ability to create, initiate, circulate and use online information

(Blackshaw & Nazzaro, 2006). Authors have made distinctions as to what social media

is not, sometimes specifying that the term should exclude data creation, data storage and

the interpersonal connections established in any application (Kaplan & Haenlein, 2010).

Ryan (2014) recognised the participation and contribution of user generated content in

his definition. However he expanded the concept to describe the roles and actions of

users, describing social media as

“The umbrella term for web-based software and services that allow individuals to come
together online and exchange, discuss, communicate and participate in a form of social
interaction. That interaction can encompass text, audio, images, video and other media,

17 | P a g e
Chapter 2: Literature Review
individually or in combination. It can involve the generation of new content; the
recommendation of and sharing of existing content; reviewing and rating products,
service and brands; discussing the hot topics of the day; pursuing hobbies, interests and
passions; sharing experience and expertise” (p.151).

Social media can also be conceptualised in terms of functionality, including networking,

socialisation and navigation (Thelwall & Stuart, 2009). Networking involves social

media serving as a function for people-finding by supporting non-social interpersonal

communication, for example through the platform ‘LinkedIn’. The socialisation

function of social media supports the social interaction of members, while the

navigation function supports the finding of resources such as blogs, videos and web

pages (Thelwall & Stuart, 2009). Brand communities established within social media

allow for socialisation and navigation by means of facilitating member to member

communication in addition to exchange with the brand and company.

While there are many definitions of social media provided in recent literature, it is

important to note that these definitions do not contradict one another, but rather build

upon preceding definitions in order to provide a more comprehensive definition and

description. For example, Ryan’s (2014) definition extends upon the description of

Kaplan and Haenlein (2010) to explain exactly how consumers share their opinions,

insights, experiences and perspectives.

Social media sites are unique from previous online communities in a range of ways.

Social media sites are based largely on one’s existing ‘real-world’ social network

structure, despite their ability to form new relationships or further online relationships

(Liu, Rau, & Gao, 2010). Previous forms of computer-mediated brand and online

communities tended to revolve around communication with strangers in the network.

18 | P a g e
Chapter 2: Literature Review
Additionally, social networking sites mimic offline network structures, through

connected individual nodes rather than groups. Social media also offer member profiles

allowing visual person-to-person exploration; whereas previous online communities

focussed on a given topic, with individuals interacting based on that topic, for example

in a chat room forum or blog. A number of specific types of social media have emerged

in recent years, outlined in the following section.

2.2.2 Types of Social Media

Social media take on many forms including blogs, business networks, enterprise social

networks, microblogs, photo sharing, product/services reviews, social bookmarking,

social gaming, social networks, video sharing and virtual worlds (Aichner & Jacob,

2015). Social media can be distinguished by six overarching categories (Kaplan &

Haenlein, 2010). Social media can include collaborative projects such as Wikipedia.

These enable joint and simultaneous creation content, in which the joint effort of many

actors leads to a better outcome than any actor could achieve individually (Kaplan &

Haenlein, 2010). Social media in the form of blogs represent the earliest form of social

media, generally defined as websites that display date-stamped entries in reverse

chronological order (Kaplan & Haenlein, 2010). The third form of social media is

content communities, in which the main objective is the sharing of media content

between users. Content communities enable sharing of content such as photos (e.g.

Flickr), videos (e.g. YouTube) and PowerPoint presentations (e.g. Slideshare). The

fourth type of social media defined by Kaplan and Haenlein (2010) is virtual game

worlds, which replicate a three-dimensional environment, where users develop avatars

and interact with each other. Similarly, virtual social worlds involve users interacting

with avatars in a three dimensional virtual environment. However there are no rules

19 | P a g e
Chapter 2: Literature Review
governing the restriction of possible interactions as observed within virtual game

worlds. Finally, social networking sites are a form of social media that allow users to

create profiles, establish friendship with other users and exchange information.

This research focuses on one type of social media; social networking sites, discussed in

the following section.

2.2.3 Social Networking Sites

Social networking sites (SNSs hereafter) are commonly defined as

Web-based services that allow individuals to construct a public or semi-public profile


within a bounded system, articulate a list of other users with whom they share a
connection, and view and traverse their list of connections and those made by others
within the system (Boyd and Ellison, 2008, p. 210).

Personal profiles established on SNSs include photos, video, audio files and blogs

(Kaplan and Haenlein, 2010). Consumers are increasingly becoming more active

participants within SNS, through interactive processes comprising multiple feedback

loops and highly immediate communication (Brodie et al., 2013). The interactive

properties of SNSs have transformed consumers from passive observers to active

participants, with SNSs serving as an ideal forums for product and brand-related

advocacy (Chu & Kim, 2011; Riegner, 2007), customer-led content generation (Vivek

et al., 2012) and customer-created product innovations (Hoyer, Chandy, Dorotic, Krafft,

& Singh, 2010; Sawhney, Verona, & Prandelli, 2005). Therefore, there is a significant

amount of social and network value provided to both users and organisations through

SNSs, as users comment, review and share information online.

SNSs have become a popular topic of academic enquiry, with scholars exploring the

20 | P a g e
Chapter 2: Literature Review
concept from varying perspectives, including usage motivations of participants

(Joinson, 2008; Raacke & Bonds-Raacke, 2008), social interactions, usage patterns

(Golder, Wilkinson, & Huberman, 2007; Hsu & Lin, 2008; Lampe, Ellison, &

Steinfield, 2006) and characteristics of users (Gjoka, Sirivianos, Markopoulou, & Yang,

2008; Hargittai, 2007). Less academic attention has been paid regarding the role of

SNSs from a marketing perspective. Practitioners have largely been at the forefront of

efforts to advise businesses on the design of their social media content, with an

inundation of industry blogs, websites and guides on the best practice for marketing

within the social network sphere emerging in recent years (Steeves, 2013). Whilst the

list of guidelines and strategies for marketing efforts within SNSs appears endless,

academic research and empirical evidence in the area remains scarce. Further, while

millions of brands have adopted sites such as Facebook, as discussed in the next section,

theoretically grounded academic enquiry guiding marketing and communications

strategies in this forum remains limited. This research focuses on one specific social

networking site, Facebook, as discussed in the following section.

Facebook

The world’s largest SNS, Facebook (Hughes, Rowe, Batey, & Lee, 2012), was

originally developed to help students at Harvard University communicate with each

other (Krivak, 2008). A decade on, Facebook is the world’s most successful social

networking company (Hansson, Wrangmo, & Solberg Søilen, 2013). As of April 2015,

Facebook had over 1.3 billion monthly active users (Social Bakers, 2015). Registered

Facebook users interact with other users through the creation of a user profile, by which

the exchange of messages, status updates, photos and videos occurs.

In 2006, organisations were allowed to become active members on Facebook and create
21 | P a g e
Chapter 2: Literature Review
public profiles, resulting in more than 4000 organisations joining within the first two

weeks (Waters, Burnett, Lamm, & Lucas, 2009). The practice of marketing

communications via Facebook is now a well-functioning concept, with many companies

considering Facebook as the most attractive SMS for B2C marketing purposes

(Lillevalja, 2010). Over 40 million brands, globally have company pages registered with

the social media site (Facebook, 2015).

Millions of companies have created Facebook fan pages, by which consumers receive

information from the company. Based on the current definitions from Facebook

(Facebook, 2015), the ‘like page’ is the official name for all Facebook pages which are

not user profile pages. ‘Like pages’ are for businesses, brands and organisations to share

their stories and connect with people. These pages are free public spaces companies can

utilise to continually update their consumers about company news, products and events

(Facebook, 2015). Within this thesis, these pages are referred to as ‘brand profile

pages’. Content shared on brand profile pages is referred to as posts and appears on the

central part of the page, known as the wall or timeline. Brand profile pages can have

one or multiple administrators who are responsible for the creation of content. The

brand profile page can have any number of members, referred to in this study as ‘users’.

Within a brand profile page, users can engage with a company in the following four

ways (Cvijikj & Michahelles, 2013); Posting content on the wall (dependent on the

communication policy set by the company), commenting on an existing post shared by

the administrator, indicating interest in an existing post by pressing the ‘like’ button,

referred to as ‘liking’ and sharing the post on their personal profile wall. Each of these

actions results in the generation of a story which appears on the wall and ‘news feed’ of

the user’s personal network of friends. In addition to these actions, companies can
22 | P a g e
Chapter 2: Literature Review
utilise functions on their brand profile pages including discussion boards, events,

photos, reviews, videos and notes (Smith & Treadaway, 2010). Further, companies can

adopt third party applications such as Facebook badges, contests, games, gifts, quizzes

and survey polls (Smith & Treadaway, 2010). Introduced in July 2011, Facebook Page

Insights allows Facebook page administrators to view metrics associated with their

posted content. Administrators have access to the Facebook Page Insights dashboard

where they can examine their page’s success based on user engagement. Within

Facebook, user engagement with a page is measured by clicks, shares and likes of page

posts (Facebook, 2015).

The increasing popularity of Facebook as a marketing and communication platform has

stimulated the interest of scholars, with research investigating user personalities (Ryan

& Xenos, 2011), online identity (Hum, Chamberlin, Hambright, Portwood, Schat, &

Bevan, 2011), self-disclosure (Park, Jin, & Jin, 2011), uses and motivations (Cheung,

Chiu, & Lee, 2011). While SNSs such as Facebook are said to enable interactive

consumer experiences which contribute to the development of customer engagement

with specific brands (Brodie et al., 2013), the extent to how much and how often this

process occurs remains largely unknown. Further, it is unclear whether efforts to

stimulate interaction and engagement amongst existing and potential customers on

brand pages has a measurable, beneficial influence on the brands they promote (Jahn &

Kunz, 2012; Nuttney, 2010).

Despite millions of brands adopting SNSs as a marketing tool, it appears that

stimulating customer interaction within these forums is a significant challenge for

marketers. Recent studies have indicated that less than five percent of customers engage

(defined as commenting, sharing, liking, answering a question, checking-in,


23 | P a g e
Chapter 2: Literature Review
or RSVP’ing to an event) within social media, regardless of the product category

(Nelson-Field & Taylor, 2012). As of March, 2015, within Australia, the average

Facebook post engagement rate was estimated to be just 0.41 percent (Social Bakers,

2015). Post engagement rate in this context is defined as the average number of

interactions per post on a given day, divided by the total number of fans for the page

(Social Bakers, 2015). Within the alcohol industry, large brands such as Corona, with

over 6 million fans are reaching maximum engagement rates of only 2.05% (Social

Bakers, 2015). Such statistics indicate a significant challenge for marketers attempting

to increase levels of engagement with their social media pages.

As customers are inundated with a proliferation of messages, it is evident marketers will

not succeed in their social media strategy efforts without an understanding of how to

effectively engineer their content to facilitate engagement (Lee et al., 2013). The

academic community recognises the need and research is advocated on customer

behaviours in multi-media environments; the role of social media within customer

experiences; the conceptualisation, definition and measurement of engagement; and

most pertinently for this research, how social media marketing activities

create customer engagement (MSI, 2014). This thesis addresses this need and UGT is

discussed in the next section of the literature review as a theoretical foundation through

which to consider consumer active choice and use of specific media, such as social

media content.

2.3 Uses and Gratifications Theory


UGT is an approach to understanding why and how individuals actively seek out and

use specific media to satisfy specific needs (Katz & Foulkes, 1962). UGT emerged in

the 1940’s when psychologist Herzog (1944) used the term gratifications to describe
24 | P a g e
Chapter 2: Literature Review
specific types or dimensions of satisfaction reported by audience members for daytime

radio programs. Subsequently, researchers became interested in why audiences engaged

in various forms of media behaviour, such as listening to the radio and reading the

newspaper (Wimmer & Dominick, 1994). UGT addresses how individuals deliberately

choose media that will satisfy their needs, allowing one to realise gratifications such as

knowledge enhancement, entertainment and relaxation, social interaction, reward or

remuneration, and personal identity (Calder et al., 2009; Ko, Cho, & Roberts, 2005).

UGT was one of the first approaches to consider the active role of the audience in media

choice, suggesting that individuals actively search for, identify with, and employ media

to fulfil specific gratification needs (Ku, Chu, & Tseng, 2013). UGT therefore posits

that individuals have free will in determining their interaction and engagement with

media. This perspective constitutes a shift from the traditional mechanistic approach,

which suggests that individual media consumers are passive.

Communication theorists and advertisers applied the UGT perspective in the context of

various mass media including television and electronic bulletins (Leung & Wei, 2000;

O'Keefe & Sulanowski, 1995). The concepts and perspectives of uses and

gratifications research are particularly useful in explaining continuing use of a particular

medium, such as continued reading, listening or viewing (McGuire, 1974). Swanson

(1987) advocated the need to understand the role of message content within uses and

gratifications research. This indicated that audience members seek and find different

gratifications within media content, affecting consumption of the content. Audience

members seek main gratifications from their media consumption, including

informational benefits, entertainment, economic or remunerative rewards and social

interaction. Entertainment and information gratifications, derived through the content of

25 | P a g e
Chapter 2: Literature Review
television programs contribute to substantial increases in television

viewing levels (Rubin, 1983). Within electronic bulletin boards, Rafaeli (1984) found

that factual or informative content is skipped least often, with increasing diversity of

content significantly and positively related to user contribution levels.

The well-established theoretical perspective of UGT can shed interesting insights on

new, interactive mediums including online media. This medium requires a higher level

of interactivity from its users, when compared to more traditional forms of media (Ko et

al., 2005; Ruggiero, 2000). As the underlying assumption of UGT is that users are

actively involved in media usage and highly motivated by their needs in their selection

of the communication media, the theory has become increasingly relevant in studies of

media channels that allow for consumer choice. Many theorists posit that UGT is a

research tradition eminently suited for internet and social media study (Kaye &

Johnson, 2003; Ruggiero, 2000), as discussed in the following section.

2.3.1 Internet Uses and Gratifications

Extending from studies of UGT in mass and traditional media, the UGT perspective has

been successfully applied to a range of new media studies. Given the inherent

interactivity and user-directed nature of internet media, this user-level approach of UGT

is well suited for examining consumer internet use (Stafford, Stafford, & Schkade,

2004). In the online context, consumers have significant control over the information

they search for and receive, unlike their experience in more traditional mass-marketing

communication. Consumers are generally responsible for initiating the flow of

communication through their decisions regarding what websites to search for or what

communities to join (Stafford & Stafford, 2001). Based on this reasoning, authors posit

that the internet is ‘intentionally consumed’ (Rayburn, 1996) and hence UGT provides
26 | P a g e
Chapter 2: Literature Review
the necessary theoretical framework for understanding the specific reasons that motivate

consumers to approach, and engage with online content.

Researchers have examined the psychological and behavioural aspects of internet users

in order to identify the appropriate underlying dimensions of internet use motivations

(LaRose, Mastro, & Eastin, 2001; Lin, 1999). Items such as social escapism,

transactional security and privacy, information seeking, interactive control,

socialisation, entertainment and economic motivation have been suggested as key

motivations for internet use (Korgaonkar & Wolin, 1999). Further, motivations such as

interpersonal utility, pastime, information seeking, convenience and entertainment have

been suggested (Papacharissi & Rubin, 2000). The effect of these motivations have

been applied to consequences such as interaction on websites, attitude toward websites,

attitude towards brands, purchase intention and satisfaction (Ko et al., 2005; Luo, 2002),

with consumer motivations significantly altering psychological and behavioural actions

of internet users. In addition to internet use, the advent of social media triggered further

enquiry of the UGT perspective in order to understand user motives and behaviours

within this forum, discussed in the following section.

2.3.2 Social Media Gratifications

Based on UGT, previous studies have used the motivations for using social media to

predict users’ specific behaviours concerning social media sites (Baek, Holton, Harp, &

Yaschur, 2011; Quan-Haase & Young, 2010). Constructs based on the

theoretical underpinnings of UGT, such as the need for social interaction, need for

entertainment, information seeking and sharing needs, and desire for reward or

remuneration have been explored in recent literature investigating consumer choices of

online and social media. Table 2.1 provides a summary of this literature.
27 | P a g e
Chapter 2: Literature Review
Table 2.1 Recent Studies with a UGT Perspective Applied to Online Media Use

Motivations Conceptual Relationships and Outcomes

Entertainment Individuals share links on Facebook because it is easy and entertaining (Baek et al., 2011).
Entertainment needs are a significant predictor of the use of comments among Facebook users (Smock, Ellison, Lampe, & Wohn, 2011).
Users with entertainment needs participate in Facebook groups for leisure and amusement (Park, Kee, & Valenzuela, 2009).
Persuasive content including emotional and philanthropic content increases engagement on Facebook in the form of likes and shares (Lee et al., 2013).
Entertaining content increases customer engagement on Facebook through increasing levels of liking, commenting and sharing, in addition to having a
positive effect on interaction duration (Cvijikj & Michahelles, 2011, 2013).
Entertainment needs are not significantly linked to attitudes towards social media content because using social media is no longer entertaining to users as it
is a common practice in everyday life (Chung & Austria, 2010).
Entertainment needs are linked to consuming, contributing to and creating brand-related content in social media (Muntinga, Moorman, & Smit, 2011).
Individuals use social media as a source of entertainment through playing games, listening to music and watching videos, in addition to looking for humour
and comic relief and to listen to jokes (Whiting & Williams, 2013)

Information The motivation to share information significantly predicts an individual’s frequency of sharing links on Facebook (Baek et al., 2011).
Seeking and
Expressive information sharing is a significant predictor of the use of Facebook groups and the use of status updates on Facebook (Smock et al., 2011).
Sharing
Individuals use Facebook groups to satisfy information seeking needs through the acquisition of knowledge regarding products, events and services (Park et
al., 2009).
Informational content on Facebook such as mentions of prices, availability and product features reduces engagement in the form of likes and comments
(Lee et al., 2013).
Informational needs are linked to individuals consuming brand-related content in social media (Muntinga et al., 2011)
Posts offering brand-related information increase the level of engagement within Facebook through liking and commenting, but do not cause an effect on

28 | P a g e
Table 2.1 Recent Studies with a UGT Perspective Applied to Online Media Use (Cont.) Chapter 2: Literature Review
the number of shares, in addition to causing the greatest increase in interaction duration (Cvijikj & Michahelles, 2011, 2013).
Social media information gratification has a positive relationship with attitudes towards social media content (Chung & Austria, 2010)
Individuals use social media to seek out information about sales, deals, products, events, birthdays, parties and information about businesses (Whiting &
Williams, 2013).

Social & Interaction gratifications through social media have a positive relationship with attitude towards social media content (Chung & Austria, 2010)
Relational
Social interaction needs are linked to consumers creating and contributing to social media content (Muntinga et al., 2011)
Interaction
Individuals use social media for social interaction (Whiting & Williams, 2013)
Social connection gratifications lead to an increased frequency of use of Facebook (Joinson, 2008)
Individuals post links on Facebook as a tool for interacting and socialising with others (Baek et al., 2011)
Social interaction needs are a significant predictor of the use of comments, individuals writing on a friends walls, private message use, the use of ‘Facebook
chat’, and the use of Facebook groups (Smock et al., 2011)
Individuals participate in Facebook groups to satisfy socialising needs through meeting and talking with others, getting peer support and a sense of
community (Park et al., 2009)

Monetary Remuneration needs are linked to individuals consuming brand-related content in social media (Muntinga et al., 2011).
incentives,
Remunerative content has a positive effect on the number of comments within Facebook; however no effect exists over the number of shares in addition to a
remuneration or
negative effect over the number of likes. Further, this content type has no effect over the interaction duration of consumers (Cvijikj & Michahelles, 2013)
reward

29 | P a g e
Chapter 2: Literature Review

In social media, a brand’s overt goal is to attract an audience by providing value, or

gratification, through its content (Malthouse et al., 2013). Content must therefore be

designed in a way which creates value for individual consumers to build a stronger level

of engagement and facilitate value outcomes (Malthouse et al., 2013). Based on

the UGT perspective, this thesis posits that social media content can be categorised into

four main groups, based on its level of information (Cvijikj & Michahelles, 2013; De

Vries, Gensler, & Leeflang, 2012; Ducoffe, 1996), entertainment (De Vries et al., 2012;

Ducoffe, 1996; Taylor, Lewin, & Strutton, 2011), remunerative (Cvijikj & Michahelles,

2013; Lee et al., 2013) and relational (Muntinga et al., 2011) content. Each of these

categories is examined in detail in the following sections.

Information

The information construct identified by UGT can be defined as the extent to which the

web provides users with resourceful and helpful information (Chen, Clifford, & Wells,

2002; Ducoffe, 1995). Attaining various forms of information has been suggested as the

most important reason for consumers to use the internet (Maddox, 1998), and levels of

information and attitude to the website have been found to be positively related (Chen et

al., 2002). Further, the relationship between the ability of an advertisement to provide

information to viewers and advertising acceptance has been well documented (Bauer &

Greyser, 1968). Advertising value and attitude to advertising have also been found as

positive consequences of informative advertising (Ducoffe, 1995, 1996).

Whilst the importance of delivering information through advertisements has been

recognised for more traditional media (Rubin, 2002), the role of informational

advertising and content in the online social domain has also received attention.

Searching for and receiving information about a brand is one of the main gratifications
30 | P a g e
Chapter 2: Literature Review
of consumer participation in online brand communities (Muntinga et al., 2011; Raacke

& Bonds-Raacke, 2008; Ulusu, 2010). The desire to seek information directly from

brands is a motivating factor for consumers to continue to use social media sites

(Dholakia, Bagozzi, & Pearo, 2004; Lin & Lu, 2011). Content gratifications such as

information seeking, knowledge and learning can predict consumers internet use

patterns (Stafford et al., 2004), attitudes towards websites and brands, purchase

intentions, and interaction behaviours (Ko et al., 2005).

Studies relating to consumer interaction with brands as a result of motives such as

information seeking are transferable to the field of customer engagement. Within this

perspective, scholars have demonstrated that consumers are motivated by informational

needs to engage with a brand on social media. This engagement is most likely to

manifest through consumer actions such as clicking on links, staying on websites

longer, reading details and threads and using multimedia features (Ko et al., 2005). This

form of interaction is referred to as human-message interaction, and denotes passive

engagement with the brand, rather than active engagement in the form of commenting

and contributing to online brand communities (Ko et al., 2005).

Entertainment

The entertainment construct refers to the extent to which web media content is fun and

entertaining to media users (Eighmey & McCord, 1998). UGT research has

demonstrated that the value of entertaining media is embedded in its ability to fulfil

users’ needs for escapism, hedonistic pleasure, aesthetic enjoyment and emotional

release (McQuail, 1983). Previous research has suggested that providing a higher

entertainment value to users is likely to lead to an advantage for media users, motivating

them to use the media more often (Chung & Austria, 2010). Early research which
31 | P a g e
Chapter 2: Literature Review
considered the role of entertaining content on the web discovered that web users who

perceive banner advertisements on the web as entertaining tended to have greater brand

loyalty to the advertised products and a higher likelihood to purchase (Stern &

Zaichowsky, 1991). The concept of entertaining advertising has been discussed

extensively in the literature, with empirical evidence demonstrating that entertaining

advertisements lead to positive attitudes toward the advertisements (Taylor et al., 2011),

positive attitudes toward the brand and a desire to return to the websites (Raney, Arpan,

Pashupati, & Brill, 2003).

Through the application of UGT (Katz & Foulkes, 1962) in the context of brand

communities and social media, authors have shown that consuming entertaining content

is an important factor for participation in brand communities (Dholakia et al., 2004;

Raacke & Bonds-Raacke, 2008). The entertainment value of a social networking site

can be an important reason for consumers to adopt it (Cheung et al., 2011; Dholakia et

al., 2004; Lin & Lu, 2011; Park et al., 2009). Extant studies within the social media

context highlighted the importance of entertaining or persuasive content as one of the

antecedents to customer engagement behaviour. Entertaining ads are said to lead users

to consume, create or contribute to brand-related content online (Muntinga et al., 2011).

Remuneration

In addition to considering whether brand content offers information and entertainment,

the level of remuneration offered to the consumers has been studied as a driver of

consumer decisions to contribute to online communities (Muntinga et al., 2011).

Consumers may engage in social media use as they expect to gain some kind of reward,

such as an economic incentive, job-related benefit or personal wants (Muntinga et al.,

2011).
32 | P a g e
Chapter 2: Literature Review
Social Exchange Theory (SET) (Emerson, 1976) posits that virtual interactions need to

be rewarding for both the product (brand, company) and the participant (Anderson,

Challagalla, & McFarland, 1999). SET has been a valuable approach to analyse user

behaviour within the online community context (Hemetsberger, 2002; Smith & Kollock,

1999). The theory demonstrates however that monetary benefits or incentives are not

required for community members to make contributions. Füller (2006) points out that

whilst managers often believe the offering of monetary incentives such as bonus points,

drawing prizes or sharing product success is beneficial to stimulating user engagement,

it is often mistaken. Rather, factors such as the ability to learn something new, the

possibility to get exclusive content and the ability to gain acknowledgement and support

from the community have a far greater impact on community members’ motivation to

contribute to virtual communities (Füller, 2006).

Social interaction

Consumer needs including the need for integration and social interaction and desire for

social benefits (Hennig-Thurau, Gwinner, Walsh, & Gremler, 2004) have been defined

as key motivations for users to access the internet. Related specifically to social media

use, sub-motivations include gaining a sense of belonging, connecting with friends,

family and society, seeking support and substituting for real-life partnership (Muntinga

et al., 2011). Social identification is an important factor in user’s contributions to social

media sites (Boyd & Ellison, 2008). Social interaction involves consumers gaining

insight into the circumstances of others, social empathy, identifying with others, gaining

a sense of belonging, finding a basis for conversation, helping carry out social roles and

enabling individuals to connect with family, friends and society (McQuail, 1983).

Social interaction needs have also been linked to consumer motivations to provide user
33 | P a g e
Chapter 2: Literature Review
generated content online, with research showing that users find the internet a

comfortable place to reveal their feelings, share views and experiences, and to let their

family and friends know about their latest information (Leung, 2009). Internet users

expressed the view that through the online content generation process, they would have

the opportunity to be recognised, publicise their expertise, learn more of the world,

socialise with friends and be entertained (Leung, 2009). Park et al. (2009) found that

socialising is a significant reason for users to participate in Facebook groups.

Socialising involves motivations such as getting peer support from others, meeting

interesting people, belonging to a community, talking about something with others and

staying in touch with friends (Park et al., 2009).

Brodie et al. (2013) define ‘socialising’ as one of the five sub-processes of customer

engagement which may occur within a virtual brand community. Socialising, in this

context, refers to two-way, non-functional interactions through which consumers

develop attitudes, norms and/or community language. Chen (2011) studied the

relationship between the social needs of Twitter users and the degree to which they are

engaged in Twitter use. The study demonstrated that usage increased the more the

person gratified a need for an informal sense of camaraderie (or connection) with other

users. Similarly, Ko et al. (2005) demonstrated that consumers with high social

interaction motivations are more likely to engage in human-to-human interaction.

Human-to-human interaction refers to behaviours such as providing comments,

feedback, and personal information to an advertiser and participating in on-line

discussion or forums. These studies suggest that the social gratification is a significant

predictor in the use of SNSs. In the following section, customer engagement is

introduced, including an examination of recent literature regarding its theoretical

34 | P a g e
Chapter 2: Literature Review
foundations, conceptualisation, dimensionality and related concepts.

2.4 Customer Engagement


The concept of engagement has been investigated across various disciplines, including

psychology, organisational behaviour, sociology and political science. Further,

engagement has been applied in a range of contexts, including community engagement

(Algesheimer, Dholakia, & Herrmann, 2005; Keener, 1999), student engagement (Kahu,

2013; Skinner & Belmont, 1993), employee engagement (Kahn, 1990; Saks, 2006),

civic engagement (Jennings & Stoker, 2004; Mondak, Hibbing, Canache, Seligson, &

Anderson, 2010), social engagement (Achterberg, Pot, Kerkstra, Ooms, Muller, &

Ribbe, 2003; Binning, Unzueta, Huo, & Molina, 2009; Huo, Binning, & Molina, 2010)

and stakeholder engagement (Andriof & Waddock, 2002; Greenwood, 2007). Across

these disciplines and contexts, the engagement concept has some conceptual

consistencies, including recognition of emotional, cognitive and behavioural activation

states (Kahn, 1990; Schaufeli, Salanova, González-Romá, & Bakker, 2002).

Customer engagement has emerged as an important construct in marketing research,

literature and practice (Brodie et al., 2011). In recent literature exploring customer

engagement, authors have focussed their attention on defining the concept, in addition

to conceptualising the stages, levels, or processes embodied within the customer

engagement concept. This section of the literature review explores customer

engagement in detail, exploring recent literature which establishes the theoretical

foundations of the concept, and outlines the definitions and dimensionality of customer

engagement. A number of related yet distinct concepts are addressed, in addition to

consideration of the antecedents and consequences of customer engagement.

35 | P a g e
Chapter 2: Literature Review
2.4.1 Customer Engagement Theoretical Foundations

Customer engagement stems from the theoretical foundations of relationship marketing

and the service-dominant (S-D) logic perspective (Ashley, Noble, Donthu, & Lemon,

2011). While the classical view of marketing is characterised by its consideration of

customers as passive recipients of value created by companies, the focus of marketing

has shifted from a product-centric to a customer-centric view of marketing (Day &

Montgomery, 1999; Deshpandé, Farley, & Webster, 1993) in line with the relationship

marketing approach. Relationship marketing is characterised as marketing activities

which establish, develop and maintain successful relational interaction (Morgan &

Hunt, 1994). Within this broadened relationship marketing notion, the firm focuses on

existing and prospective customers, in addition to consumer communities and co-

creative networks (Vivek et al., 2012).

Compared to the traditional goods-dominant (G-D) perspective where consumers are

provided with the value created by firms and act as receivers (Lusch, 2007), the S-D

logic perspective advocates an interactive view of the customer-brand relationship,

whereby value co-creation through customer collaboration occurs. The S-D logic

perspective addresses the service as the main purpose within business exchange, and

emphasises the co-creation of value resulting from interactions among firms, customers

and other stakeholders (Karpen, Bove, & Lukas, 2011). Interactive consumer

experiences co-created with other actors can be interpreted as the act of ‘engaging’

(Lusch, Vargo, & Tanniru, 2010). Co-creation therefore occurs when the customer

participates through behaviours that uniquely customise the customer-to-brand

experience, beyond the selection of predetermined options as in co-production (van

Doorn et al., 2010).

36 | P a g e
Chapter 2: Literature Review
2.4.2 Engagement Conceptualisation

The conceptualisation and definition of engagement varies across multiple disciplines

and contexts. For example, organisational behaviour literature suggests that engagement

is physically, emotionally or cognitively expressed through task behaviours (Bowden,

2009). By comparison, in the discipline of social psychology, engagement is described

as an initiative and adequate response to social stimuli (Jennings & Stoker, 2004).

Student engagement includes academic investment, motivation and commitment to an

institution, in addition to perceived psychological connection, comfort and sense of

belonging (Bryson & Hand, 2007; Hu & Wolniak, 2010).

Within marketing literature, engagement has been characterised by a range of forms

including customer engagement (Patterson, Yu, & De Ruyter, 2006), customer

engagement behaviours (van Doorn et al., 2010), customer brand engagement

(Hollebeek, 2011b), consumer engagement (Vivek et al., 2012) and simply engagement

(Higgins, 2006). Customer engagement can be defined as a psychological process by

which customers move towards being loyal toward a brand (Bowden, 2009), or an

ongoing emotional, cognitive and behavioural activation state (Brodie et al., 2011).

Customer engagement has also received significant attention by authors regarding its

specific behavioural dimension, who define engagement as behavioural manifestations

toward an object (e.g. a brand) other than purchase, which result from motivational

drivers (van Doorn et al., 2010; Verleye, Gemmel, & Rangarajan, 2013). Other

engagement scholars go beyond behaviour to incorporate motivational drivers and/or

psychological aspects as part of the engagement construct. To illustrate, Hollebeek

(2011b) states that customer brand engagement is ‘the level of an individual customer’s

motivational, brand-related, context-dependent state of mind characterised by specific

37 | P a g e
Chapter 2: Literature Review
levels of cognitive, emotional and behavioural activity in direct brand interactions’

(p.790). This three dimensional perspective of customer engagement, incorporating

cognitive, behavioural and emotional aspects is widely accepted in the customer

engagement literature (Brodie et al., 2011). Despite the widely accepted three

dimensional conceptualisation of engagement, there is little specific detail known

regarding the specific sub-dimensions, operationalisation and measures of the cognitive,

behavioural and emotional aspects. Further, there remains a lack of consistency and

clarify regarding what the specific dimensions are, as demonstrated in Table 2.2. Table

2.2 provides a summary of the definitions proposed by recent scholars including

engagement, consumer engagement, customer engagement, customer brand

engagement, customer engagement behaviour, advertising engagement, media

engagement, brand community engagement and online engagement.

38 | P a g e
Chapter 2: Literature Review
Table 2.3 Customer Engagement Definitions

Author Concept Definition Engagement


Dimensionality

McEwen (2004) Engagement A measure of the strength of a company’s customer relationships based on the -
extent to which customers have formed both emotional and rational bonds with
a brand

Peppers and Rogers (2005) Engagement Engagement is a series of customised informational and financial transactions Informational and financial
that occur over time and increase both the consumer value to the company and transactions
the value of the company to the consumer

Foley (2006) Engagement Engagement is a multidimensional concept, even a multidimensional process, Multidimensional:
with the end result defined as consumer connection in terms of cognitive,
behavioural, emotional, and aspirational facet Cognitive, behavioural,
emotional and aspirational

Higgins (2006) Engagement Being engaged is to be involved, occupied and interested in something Multidimensional:

Cognitive, hedonic, social

Scholer and Higgins (2009) Engagement Engagement is defined as a state of being involved, occupied, fully absorbed, -
or engrossed in something

Gambetti and Graffigna Engagement Rather than a specific definition, the following marketing-based sub-forms are Focal engagement sub-forms
(2010) identified: consumer-, customer-, brand- advertising- and media engagement may comprise the following
dimensions: soft (relational),
pragmatic (managerial)

39 | P a g e
Table 2.2 Customer Engagement Definitions (Cont.) Chapter 2: Literature Review

Author Concept Definition Engagement


Dimensionality

Abdul-Ghani, Hyde, and Engagement Requires consumer connection (e.g. with specific media) Multidimensional:
Marshall (2011)
Utilitarian, hedonic, social

Appelbaum (2001) Consumer Consumer engagement consists of both rational loyalty (includes overall Multidimensional: rational
engagement satisfaction, intent to repurchase, and intent to recommend) and emotional loyalty, emotional
attachment (including confidence in a brand, belief in its integrity, pride in the attachment
brand, and passion for it)

Ghuneim (2006) Consumer Consumer engagement is a consumer-based measurement that relates to -


engagement interaction with an aspect of a brand or media property

Harris (2006) Consumer Consumer engagement is a multidimensional concept: a brand’s ability to -


engagement connect meaningfully with the consumer

Campanelli (2007) Consumer Consumer engagement is the emotional connection and empowerment of Emotional
engagement consumers

Heath (2007) Consumer Consumer engagement is a subconscious emotional construct. Level of One-dimensional: emotional
engagement engagement is the amount of subconscious ‘‘feeling’’ going on when an
advertisement is being processed

Shevlin (2007) Consumer Consumer engagement is repeated and satisfying interactions that strengthen One-dimensional: emotional
engagement the emotional connection a consumer has with a brand (or product or company)

Vivek et al. (2012) Consumer The intensity of a consumer’s participation and connection with an Multidimensional:
Engagement organisation’s offerings and/or its organised activities
40 | P a g e
Table 2.2 Customer Engagement Definitions (Cont.) Chapter 2: Literature Review

Author Concept Definition Engagement


Dimensionality

Awareness, enthusiasm,
interaction, activity,
extraordinary experience

Patterson et al. (2006) Customer The level of a customer’s physical, cognitive and emotional presence in their Multidimensional:
engagement relationship with a service organisation
Vigor, dedication,
absorption, interaction

Bowden (2009) Customer A psychological process comprising cognitive and emotional aspects. Multidimensional:
engagement
Cognitive and emotional

MSI (2010) Customer Customers’ behavioural manifestation toward a brand or firm beyond purchase, One-dimensional:
engagement which results from motivational drivers including: word-of-mouth activity, behavioural
recommendations, customer-to-customer interactions, blogging, writing
reviews, and other similar activities.

Smith and Wallace (2010) Customer Customer engagement (CE) refers to the types of connections consumers make -
engagement with other consumers, companies, and specific brands; CE is viewed as being
conducive to enhancement of brand loyalty

So, King, and Sparks (2014) Customer A customers’ personal connection to a brand as manifested in cognitive, Multidimensional:
engagement affective, and behavioural actions outside of the purchase situation identification, enthusiasm,
attention, absorption,
interaction

41 | P a g e
Table 2.2 Customer Engagement Definitions (Cont.) Chapter 2: Literature Review

Author Concept Definition Engagement


Dimensionality

Hollebeek (2011b) Customer brand The level of an individual consumer’s motivational, brand-related and context- Multidimensional:
engagement dependent state of mind characterised by specific levels of cognitive, emotional
and behavioural activity in brand interactions. Cognitive, emotional,
behavioural

Bijmolt et al. (2010) Customer The behavioural manifestation from a customer toward a brand or a firm which One-dimensional:
engagement goes beyond purchase behaviour behavioural
behaviour

van Doorn et al. (2010) Customer The customer’s behavioural manifestation toward the brand or firm, beyond Multidimensional:
engagement purchase, resulting from motivational drivers
behaviours Valence, form, scope,
nature, customer goals.

Algesheimer et al. (2005) Brand Positive influences of identifying with the brand community through the Multidimensional:
community consumers’ intrinsic motivation to interact/cooperate with the community
engagement members Utilitarian, hedonic, social

Davis Mersey, Malthouse, and Media A motivational experience; being connected to a specific media Multidimensional:
Calder (2010) Engagement
Transportation, irritation,
promotion, rejection

Phillips and McQuarrie (2010) Advertising Modes of engagement are routes to persuasion Multidimensional:
Engagement
Consumers engage in ads to:
act, identify, feel, immerse.

42 | P a g e
Table 2.2 Customer Engagement Definitions (Cont.) Chapter 2: Literature Review

Author Concept Definition Engagement


Dimensionality

Owyang (2007) Online Online engagement indicates the level of authentic involvement, intensity, Multidimensional:
engagement contribution and ownership, summarised by ‘apparent interest’ involvement, intensity,
contribution and ownership

Peterson (2007) Consumer Consumer online engagement is an estimate of the degree and depth of visitor One-dimensional:
online interaction on the site, measured against a clearly defined set of goals. Each behavioural.
engagement organisation’s version of engagement will be unique. It will be derived from a
number of root metrics, probably under a dozen. Common root metrics include
frequency, recency, length of visit, purchases, and lifetime value

Mollen and Wilson (2010) Brand The cognitive and affective commitment to an active relationship with the Multidimensional:
engagement brand as personified by the website or other computer-mediated entities
(online) designed to communicate brand value Cognitive and affective

43 | P a g e
Chapter 2: Literature Review

2.4.3 Customer Engagement Related Concepts, Antecedents and Consequences

The imprecision regarding a universal definition of customer engagement has led to

varied interpretations of what customer engagement is, and is not. Specifically, scholars

have focussed attention on providing a clear distinction between the relatively new

domain of customer engagement, and closely related but conceptually differing

concepts such as brand experience, loyalty, satisfaction, involvement, participation,

commitment, interactivity and flow. This section of the literature review addresses these

related but distinct concepts.

The conceptual distinction between involvement and engagement is discussed frequently

in customer engagement literature (Bowden, 2009; Brodie et al., 2011; Brodie et al.,

2013; Mollen & Wilson, 2010). Involvement reflects focused attention or engrossment

with an engagement object (Mollen & Wilson, 2010). Defined as a “state of mental

readiness that typically influences the allocation of cognitive resources to a

consumption object, decision or action” (Thomson, MacInnis & Park, 2005, p. 41),

involvement considers the cognitive processing, relevance and interest that a consumer

experiences with regards to engagement objects. Involvement can also be defined as the

perceived relevance of the object based on inherent needs, values and interests

(Zaichkowsky, 1985). Comparatively, engagement goes beyond involvement to

encompass an active dimension, depicted through interaction between engagement

objects and subjects (Mollen & Wilson, 2010). Involvement has been suggested as an

antecedent to behavioural customer engagement (Vivek et al., 2012) and psychological

customer engagement (Cheung et al., 2011).

Customer engagement can be viewed as a psychological state which drives customer

loyalty (Bowden, 2009). Loyalty can manifest behaviourally, through repeated


44 | P a g e
Chapter 2: Literature Review
purchases prompted by attitudinal loyalty, a strong internal disposition (Jacoby &

Kyner, 1973). Loyalty does not reflect part of the engagement concept itself. Rather it is

seen as a potential customer engagement consequence (Bowden, 2009; Patterson et al.,

2006). Similarly, satisfaction is conceptually distinct from customer engagement.

Satisfaction is considered as an evaluative outcome of customer engagement for new

customers (Brodie et al., 2011). However, some discrepancy occurs regarding its

conceptual relationship with engagement. Authors have also proposed that satisfaction

could be an antecedent to customer engagement behaviour for existing and experienced

customers (van Doorn et al., 2010). In this case, it is expected that satisfied, existing

customers of a brand are more likely to engage with the focal brand. From Brodie et

al.’s (2011) perspective, a customer’s satisfaction with the brand occurs following the

engagement experience.

Participation refers to the degree to which customers produce and deliver service

(Bolton & Saxena-Iyer, 2009). While related constructs such as involvement capture the

psychological elements of customer engagement, participation reflects the behavioural

facet of engagement. Multidimensional views of customer engagement (Brodie et al.,

2011) advocate inclusion of the cognitive, emotional and behavioural dimensions in

order to capture the full notion of customer engagement. Participation fails to

encapsulate this notion, only reflecting the behavioural dimension. Additionally, Vivek

(2009) suggests that participation refers to customers’ connections with the firm in

exchange situations. However customer engagement is a broader concept which goes

beyond such exchange-centric perspectives.

Commitment is closely related to the emotional/affective dimension of customer

engagement, encompassing a psychological attachment (Bowden, 2009). Commitment


45 | P a g e
Chapter 2: Literature Review
reflects an emotional attachment in which a customer is committed when “his or her

values, self-image, and attitudes are strongly linked to a specific choice alternative”

(Bowden 2009, pg. 70). This state causes the customers to view a specific object as the

only acceptable choice. Compared to involvement, customer commitment goes beyond

interest and relevance. However it does not incorporate behavioural and cognitive

dimensions as required by customer engagement. Commitment has been suggested as a

consequence of customer engagement (Brodie et al., 2013; Chan & Li, 2010).

The concept of interactivity is closely related to the behavioural dimension of customer

engagement. While there is little consensus about the definition of interactivity in the

literature, it has been proposed that interactivity is an experiential phenomenon, in

which customers perceive communication as “two-way, controllable and responsive to

their actions” (Mollen and Wilson, 2010. p.5). This concept appears to capture the

behavioural dimension of customer engagement. However it excludes the application of

psychological and motivational elements as captured within the emotional and cognitive

aspects of customer engagement. Through its depiction of the behavioural element of

customer engagement, interactivity has been discussed as one dimension of engagement

(So et al., 2014), an antecedent to engagement (Hollebeek, 2011b) and a consequence of

customer engagement.

Flow can be defined as a state of optimal experience that is characterised by focused

attention, a clear mind, mind and body unison, effortless concentration, complete

control, loss of self-consciousness, distortion of time and intrinsic enjoyment

(Csikszentmihalyi, 1990). Flow is a distinct construct to engagement characterised as a

psychological state (Mollen & Wilson, 2010). Flow may act as an antecedent to

customer engagement in certain contexts, including the online environment (Brodie et


46 | P a g e
Chapter 2: Literature Review
al., 2011).

Closely related concepts to engagement including involvement, loyalty, participation,

commitment, interactivity, and flow as discussed have often been positioned as

antecedents and/or consequences to customer engagement depending on the context.

The following section of the literature review discusses customer engagement

antecedents and consequences in order to bring further clarity regarding customer

engagement.

Discussion surrounding the closely related concepts to customer engagement has given

rise to academic debate regarding the potential antecedents and consequences of

engagement. Factors such as identification (Nambisan & Baron, 2007; Tsai et al., 2012),

identity (Eisenbeiss et al., 2012; van Doorn et al., 2010) and hedonism (Gambetti et al.,

2012; Nambisan & Baron, 2007) have been described as antecedents to customer

engagement. Consequences of customer engagement include loyalty (Bowden, 2009;

Brodie et al., 2011; Hollebeek, 2011a), customer value (Vivek et al., 2012), word-of-

mouth (WOM) (Vivek et al., 2012) and product innovation (Hoyer et al., 2010;

Sawhney et al., 2005). Whilst these concepts are distinct in their positioning as either

antecedents or consequences of engagement, a lack of conceptual clarity emerges

regarding constructs such as satisfaction, trust, rapport, commitment and interaction

(Brodie et al., 2011; Gambetti et al., 2012; Hollebeek, 2011b; Tsai et al., 2012; van

Doorn et al., 2010; Vivek et al., 2012). Table 2.3 provides a further summary of

customer engagement and its conceptual relationships. The constructs are presented in

three groups: antecedents of customer engagement, consequences of customer

engagement, and constructs that have been categorised as both antecedents and

consequences depending on the context.


47 | P a g e
Chapter 2: Literature Review
Table 2.4 Customer Engagement Conceptual Relationships
Construct Definition Conceptual Relevant Engagement Literature
Relationship to CE

Antecedents

Involvement An individual’s level of interest and personal relevance in relation to a focus Antecedent Hollebeek (2011a),
object/decision in terms of his or her basic values, goals and self-concept Brodie et al. (2011),
(Lassar, Mittal, & Sharma, 1995) Vivek et al. (2012),
Nambisan and Baron (2007)

Participation The degree to which customers produce and deliver service (Bolton & Saxena- Antecedent Brodie et al. (2011)
Iyer, 2009) Vivek et al. (2012)

Flow A state of optimal experience characterised by focused attention, clear mind, Antecedent Hollebeek (2011a)
mind and body unison, effortless concentration, complete control, loss of self-
consciousness, distortion of time, and intrinsic enjoyment (Csikszentmihalyi,
1990)

Identification Identification refers to a person's self-conception, according to the defining Antecedent Nambisan and Baron (2007)
features of a self-inclusive social category (e.g., brand community) that renders Tsai et al. (2012)
the self stereotypically “interchangeable” with other group members and distinct
from outsiders (Bergami & Bagozzi, 2000; Bhattacharya & Sen, 2003)

Identity a psychological state in which people acquire a social identity (as part of their Antecedent Eisenbeiss et al. (2012),
self-concept) through a group when they perceive group membership (Reed, van Doorn et al. (2010)
2002; Terry, Hogg, & White, 1999)

48 | P a g e
Table 2.3 Customer Engagement Conceptual Relationships (Cont.) Chapter 2: Literature Review
Construct Definition Conceptual Relevant Engagement Literature
Relationship to CE

Consequences

Loyalty Repeated purchases (behavioural loyalty) prompted by a strong internal Consequence Bowden (2009)
disposition (attitudinal loyalty) (Day, 1976) over a period of time (Guest, 1944).

Customer value A customer’s overall assessment of the utility of a product/service based on Consequence Hollebeek (2011)
perceptions of what is received and what is given (Zeithaml, 1988)

Antecedents and Consequences

Interaction A variable characterised by some form of customer-firm interaction (Bolton & Antecedent Hollebeek (2011a)
Saxena-Iyer, 2009)
Consequence De Vries et al. (2012)
Tsai et al. (2012)

Rapport Perceived level of harmonious, empathetic or sympathetic connection to another, Antecedent (existing Hollebeek (2011a)
which is viewed in some way as congruent to the self (Brooks, 1989); A sense of customers)
genuine interpersonal sensitivity and concern (Ashforth & Humphrey, 1993)
Consequence (new Brodie et al. (2011)
customers)

Customer A customer’s overall evaluation of the performance of an offering to date Antecedent (existing van Doorn et al. (2010)
Satisfaction (Gustafsson, Johnson, & Roos, 2005; Johnson & Fornell, 1991). customers)

Consequence (new Tsai et al. (2012)


customers)

49 | P a g e
Chapter 2: Literature Review
Table 2.3 Customer Engagement Conceptual Relationships (Cont.)
Construct Definition Conceptual Relevant Engagement Literature
Relationship to CE

Trust Consumer-perceived security/reliability in brand interactions and the belief that Antecedent of CE Bowden (2009)
the brand acts in consumers’ best interests (Delgado-Ballester, Munuera- behaviour van Doorn et al. (2010)
Aleman, & Yague-Guillen, 2003; Rotter, 1967)
Antecedent (existing Hollebeek (2011a)
customers)

Consequence (new Tsai et al. (2012)


customers) Gambetti et al. (2012)

Commitment Valuing an ongoing relationship with a specific other party so as to warrant Consequence Bowden (2009)
maximum efforts at maintaining it, i.e. a desire to maintain the relationship
(Morgan & Hunt, 1994) Antecedent (existing van Doorn et al. (2010)
customers)

Antecedent Hollebeek (2011a)


(behaviour) Gambetti et al. (2012)

50 | P a g e
Chapter 2: Literature Review

2.4.4 Dimensions of Customer Engagement

As discussed throughout Sections 2.4.2, ‘Engagement Conceptualisation’ and 2.4.3

‘Engagement Related Concepts, Antecedents and Consequences’, a lack of clarity

remains regarding the definition and dimensionality of customer engagement. However,

the concept has been generally conceptualised into three overarching dimensions;

cognitive, emotional and behavioural engagement (Brodie et al., 2011; Cheung et al.,

2011; Hollebeek, 2011b; Mollen & Wilson, 2010; Patterson et al., 2006; Taheri, Jafari,

& O'Gorman, 2014; Vivek et al., 2012). The multi-dimensional view of customer

engagement defines engagement as a ‘psychological state’ through its incorporation of

cognitive and emotional aspects (Brodie et al., 2011; Hollebeek, 2011b; Patterson et al.,

2006). Comparatively, one-dimensional engagement perspectives have a dominant

focus on the behavioural manifestations of engagement (Brodie et al., 2011)

The cognitive dimension of customer engagement refers to individual’s levels of

concentration and/or engrossment in the brand. Emotional activity can be represented by

a customer’s level of brand-related inspiration and pride. Thirdly, behavioural

engagement activity can be expressed through a customer’s level of energy exerted in

interacting with a focal brand (Hollebeek, 2011a, 2011b).

This thesis explores the behavioural manifestation of the concept, consistent with

previous studies of engagement and social media (e.g. Gummerus et al., 2012, van

Doorn et al., 2010). Section 2.4.5 introduces customer engagement behaviour (CEB)

and its positioning within this thesis.

2.4.5 Customer Engagement Behaviour

CEB is defined as “a customer’s behavioural manifestations that have a brand or firm

51 | P a g e
Chapter 2: Literature Review
focus, beyond purchase” (van Doorn et al., 2010 p.254). CEB involves customers’

voluntary resource contributions that have a brand or firm focus, but go beyond what is

fundamental to transactions. These contributions occur in interaction between the focal

engagement object and/or other actors and result from motivational drivers (Jaakkola &

Alexander, 2014). In line with the rise of online social networks as discussed earlier in

Section 2.2, non-transactional customer behaviour is an increasingly important

consideration, as customers and firms rapidly and easily interact online (Verhoef,

Reinartz, & Krafft, 2010). The achievement of customer engagement is said to incur

beneficial consequences for brands, through interactions between customers and

employees, in which customers can give suggestions for service improvement, resulting

in cost advantages for firms (Hoyer et al., 2010; Verleye et al., 2013). Further, CEB

among customers through the creation of word of mouth, referrals and online reviews

can affect other customer attitudes and behaviours towards brands (Gupta & Harris,

2010). CEB directed toward the firm and its employees may involve elements of

cooperation, feedback and compliance.

The proliferation of social media platforms and corresponding consumer adoption in

recent years has precipitated a paradigm shift, significantly altering the way customers

behave and engage with brands. While traditional marketing communications

approaches were characterised by one way, controlled communication from the

marketer to consumer, the social media paradigm shift has allowed for interactive and

dynamic communications between customers and brands. The social media environment

offers users a touch-point through which they can actively and behaviourally engage

with brands through reading, commenting, reviewing and sharing information online

(Calder et al., 2009). In the next chapter, the concept of CEB specifically within the

52 | P a g e
Chapter 2: Literature Review
social media forum is introduced and defined.

2.5 Chapter Summary


This chapter reviewed the relevant literature regarding social media, UGT and customer

engagement. The chapter began by discussing the emergence of social media, its

definitions and types, specifically focussing on social networking sites. It is evident

from the foregoing literature presented in this chapter that a significant challenge for

marketing academics and practitioners has emerged, with recent academic enquiry

showing a significant lack of knowledge regarding the strategic development of

successful customer engagement within social media. Additionally, while recent

research has explored the conceptualisation, related concepts, and antecedents and

consequences of customer engagement, studies that consider customer engagement

with social media are only beginning to emerge. In particular, there is a need to develop

a theoretical understanding of the nature of engagement behaviour in response to

marketing practices within a social network structure (Sashi, 2012).

This chapter examined the relevant literature regarding social media, UGT and customer

engagement. Four overarching social media gratifications were identified based on

recent UGT literature; information seeking, entertainment seeking, desire for economic

reward or remuneration, and desire for social and relational interaction. Through

the application of the UGT perspective, a greater understanding is developed regarding

specific user motivations for customer engagement through social media.

53 | P a g e
Chapter 3: Social Media Engagement Behaviour

CHAPTER 3. Social Media Engagement Behaviour


3.1 Introduction
Customer engagement is interactive and context-dependent in nature (Brodie et al.,

2011; Calder et al., 2009; Gummerus et al., 2012). To comprehensively understand

customer engagement, examination of specific focal objects of engagement is required.

Focal objects of customer engagement include product or service offerings (Brodie et

al., 2011), media (Calder et al., 2009), and activities and events (Vivek et al., 2012).

While customers engage with a firm or brand through the multiple touch-points and

service encounters, constituting the entire brand experience, there is little research that

examines engagement with a specific focal object.

In this chapter, the focus is on one touch-point; social media. Extending from the

literature reviewed in the previous chapter, the examination of engagement behaviour

focuses on a singular focal object of engagement (social media), and therefore does not

reflect customer brand engagement in its entirety. The examination within this context-

specific environment provides greater insight into the behavioural manifestations of

engagement within social media platforms in order to further develop an understanding

of the nature of engagement at different intensities and with different valence (Brodie et

al., 2011; Vivek et al., 2012).

Specifically, this chapter outlines the development of a new construct termed ‘social

media engagement behaviour’ (SMEB). The construct explicates six specific types of

SMEB. These types are characterised by varying levels of intensity in addition to their

positively and negatively-valenced nature.

54 | P a g e
Chapter 3: Social Media Engagement Behaviour

3.2 Social Media Engagement Behaviour


3.2.1 Definition of Social Media Engagement Behaviour
This thesis adapts the definition of customer engagement behaviour from van Doorn et

al., (2010, p. 254) to reflect social media engagement:

Social media engagement behaviours go beyond transactions, and may be


specifically defined as a customer’s behavioural manifestations that have a social
media focus [adapted], beyond purchase, resulting from motivational drivers.

In order to further understand SMEB, this chapter draws from previous literature in

order to investigate the intensity and valence of engagement behaviours that may exist

in the social media context. SMEB intensity and SMEB valence are discussed in the

following sections, leading to a discussion of the SMEB construct developed for this

study.

3.2.2 Social Media Engagement Behaviour Intensity

Customers engage with focal brands and brand related content within social media

platforms (Chu, 2011; Chung & Austria, 2010; Hollebeek et al., 2014). Traditional

categorisations of ‘users’ of social media: distinguish between users who create content

such as ‘posters’, compared to those who are members of a community but do not post,

referred to as ‘lurkers’ (Nonnecke & Preece, 1999; Preece, Nonnecke, & Andrews,

2004). This basic categorisation of online users is limited in its general nature and fails

to take into account the diverse number of possible roles available to users in dynamic

platforms. Previous research categorised social media users who ‘like’ brands on

Facebook into groups based on their brand loyalty, brand love, use of self-expressive

brands, and word of mouth (Wallace, Buil, De Chernatony, & Hogan, 2014). In an

attempt to define customer engagement intensity within social media, scholars have also

55 | P a g e
Chapter 3: Social Media Engagement Behaviour
characterised engagement behaviours on a continuum of low to high activity (Muntinga

et al., 2011).

Muntinga et al. (2011) propose three social usage types: consuming (low level of brand

related activity), contributing (medium level) and creating (highest level). Muntinga et

al’s (2011) development of the “COBRA’s” (consumers online brand related activities)

as described here has not been empirically tested. Based on its description, consuming

brand related content may reflect a level of dormancy whereby consumers do not

actively contribute to the brand relationship. Malthouse et al. (2013) distinguish

between two levels of engagement, lower and higher. Lower engagement describes

situations in which customers exhibit passive engagement, consuming content or using

very basic forms of feedback, such as ‘liking’ a page on Facebook. Comparatively,

higher engagement occurs in cases when customers more actively process the role of the

brand in their lives, participating in forms of co-creation through writing reviews,

comments or creating content (Malthouse et al., 2013). Other online engagement

scholars (Cvijikj & Michahelles, 2013; De Vries et al., 2012; Lee et al., 2013) have

proposed more numeric measures of online engagement behaviours, such as the number

of likes, comments, shares and interaction duration as indicators of how much

‘engagement’ was demonstrated by consumers.

3.2.3 Social Media Engagement Behaviour Valence


Online engagement behaviours such as blogging, writing online reviews and word-of-

mouth activity have the potential to be positive or negative for the firm, based on the

valence of the content (van Doorn et al., 2010). Positive customer engagement includes

those actions that in both the short and long run have positive consequences (financial

and nonfinancial) for the firm. Actions such as recommending the brand to friends and

56 | P a g e
Chapter 3: Social Media Engagement Behaviour
family may be predominantly positive, however also have the potential to be negative

(van Doorn et al., 2010). The valence of customer engagement behaviour is one of five

dimensions of engagement behaviour as proposed by van Doorn et al. (2010). The other

dimensions are scope, form/modality, nature of impact and customer goals (van Doorn

et al., 2010). It has been proposed that customer-based, firm-based, and context-based

factors act to facilitate the five proposed dimensions of customer engagement

behaviour.

Positively-valenced engagement behaviour is reflected in favourable or affirmative

behaviours, whereas negatively-valenced engagement behaviour is exhibited through

unfavourable behaviours (Hollebeek & Chen, 2014). Positively-valenced behaviours

often reflect heightened levels of customer engagement and include activities such as

‘sharing’ a brand post to a friend with a recommendation to experience the offer (van

Doorn et al., 2010).

Extant literature pays little attention to negatively-valenced behaviours that involve such

activities as customers frequenting anti-brand communities, or visiting social media

platforms to vent negative feelings and views about brands.

3.2.4 Social Media Engagement Behaviour Construct

The SMEB construct outlined in this section proposes six distinct types of behaviour

that epitomise a hierarchy of SMEB and reflect both positively- and negatively-

valenced nature of the behaviour. The construct incorporates lower intensity and more

passive engagement behaviours, such as dormancy and consumption. Additionally, it

recognises more active engagement behaviours with a moderate intensity: detachment

and contribution. Finally, the construct demonstrates the occurrence of highly active

engagement behaviour such as creation and destruction. Whilst recognising the different
57 | P a g e
Chapter 3: Social Media Engagement Behaviour
forms of SMEB, the construct also reflects the valence of behaviours. Each of these

types of SMEB will be discussed in Section 3.2.5.

Studies addressing the processes and levels of social behaviour in the online context

provide foundational insights. However, theoretical coherence of the concept and

corresponding measurement techniques remain sparse. The construct developed

depicting SMEB offers value to engagement researchers through the context-specific

detailed investigation of engagement behaviour. Further, the description of various

engagement behaviours offers managerial clarity regarding exactly how users engage,

both positively and negatively, within social media platforms. The construct includes a

neutral inactive level termed dormancy, three positively-valenced behaviours:

consuming, contributing and creating, and two negatively-valenced behaviours:

detachment and destruction. Each of these typologies of SMEB is discussed in the

following section. Table 3.1 outlines each of the types of SMEB, with their respective

definitions and examples. Additionally, the six SMEB’s are presented visually in Figure

3.1.

58 | P a g e
Chapter 3: Social Media Engagement Behaviour
Figure 3.1 Social Media Engagement Behaviour Construct

59 | P a g e
Chapter 3: Social Media Engagement Behaviour
Table 3.1 Positively- and Negatively-Valenced Social Media Engagement Behaviours

Behaviour Definition Examples


Creating Users engage with brands and other users by creating positively- Knowledge Seeking: Content is created by users with the objective of learning, through the
valenced content on social media platforms. Creating epitomises acquisition of competencies that consumers apply to purchase consumption decisions.
a highly active level of SMEB. Creating users exhibit specific Sharing Experiences: Users provide content that is designed to disseminate personal
creating behaviours of knowledge seeking, sharing experiences, relevant information, knowledge and experiences.
advocating, socialising, co-developing and affirming.
Advocating: Users recommend specific brands, products/services and organisations, or
ways of using products and brands.
Socialising: Users’ content reflects two-way, non-functional interactions
Co-developing: Content which assists in the development of new products and services
Affirming: The creation of content with the aim of disseminating support, encouragement
and acknowledgement of the focal firm, brand or organisation’s success
Contributing Users contribute to existing content in social media platforms. Facebook: ‘like’ content and ‘share’ content to personal profile or friends profile
Contributing users exhibit a moderate level of positively- Twitter: ‘favourite’ brand related tweets,‘re-tweet’ brand-related content to personal
valenced SMEB. profile.
Instagram: ‘Like’ brand related images, ‘re-gram’ brand-related images to personal profile
YouTube: ‘like’ content, share video to personal social networks.
Consuming Users passively consume content without any form of active Viewing brand-related video
reciprocation or contribution. Consuming users demonstrate a Listening to brand-related audio
minimum level of positive, passive SMEB.
Viewing pictures and photos posted by the brand
Reading brand posts
Reading post comment threads and conversations
Reading product/brand reviews within the social media page
Dormancy A temporary state of inactive, passive engagement by users who Brand-related content is delivered to the user via the social media news feed or home page
may have previously interacted with the focal brand. but the user takes no action.

60 | P a g e
Chapter 3: Social Media Engagement Behaviour
Table 3.1 Positively- and Negatively-Valenced Social Media Engagement Behaviours (Cont.)
Behaviour Definition Examples
Detaching Users take action to remove content of the brand appearing in ‘Unliking’ or ‘unsubscribing’ to a social media brand page
their news-feed or equivalent home page. Detaching users ‘Unfollowing’ a brand on social media
exhibit a moderate level of negatively-valenced SMEB
Terminating a subscription for further updates and content from the brand
Selecting to hide future posts

Negative, active contributions to existing content on social media Conversing negatively on brand-related content
Destructing
platforms are created by destructive users. Destructive users Making negative contributions to brand forums
represent a highest level of negatively-active SMEB.
Publicly rating products and brands negatively
Commenting negatively on posts, blogs, videos and pictures posted by the brand
Writing a public complaint on the brand page
Writing negative product reviews and testimonials on social media content
Reporting brand or brand-related social media content for misconduct of use on social
media

61 | P a g e
Chapter 3: Social Media Engagement Behaviour

3.2.5 Social Media Engagement Behaviour Typologies

Creating

Users who create original content within social media platforms exhibit the highest

level of positively-valenced SMEB. Users make unique, positive, active contributions to

social media content by disseminating their knowledge, resources and experiences

(Brodie et al., 2013). This behaviour goes beyond relaying (e.g. sharing, liking) content

created by the brand and reflects a user’s contribution to the brand’s social media site.

Although not all user comments and content creation is favourable towards a brand, the

categorisation recognises the creation of negative content as destructing engagement

behaviours and these will be discussed in detail later in this section. Therefore, within

the typology of SMEB, it is posited that positively-valenced, active and highly engaged

users fall under the creating type of engagement behaviour. They exhibit interactive

creation behaviours specific to social media platforms including knowledge seeking,

sharing experiences, advocating, socialising and co-developing the brand experience or

offering (Brodie et al., 2013).

Six creating engagement behaviours, specific to social media platforms are derived

from Brodie et al., (2013). These different categories reflect the objective and intent of

the content that is created and posted to the social media site by users. These categories

may occur independently or in conjunction (e.g. sharing experiences and advocating).

The categories include knowledge seeking, sharing experiences, advocating, socialising,

affirming, and co-developing. Each of the creation behaviours are discussed in the

following sections.

Knowledge seeking: Users create content within social platforms with the objective of

seeking knowledge from other users that can be applied to purchase and consumption
62 | P a g e
Chapter 3: Social Media Engagement Behaviour
decision-making (Brodie et al., 2013). The concept of knowledge acquisition through

media content has strong links to the uses and gratifications theoretical perspective

(Calder et al., 2009). Content created may include specific questions about the product

or service, directed towards other users of the brands who may share their experience

and knowledge. Social media platforms enable and increase the collaboration and

learning from customers in various ways, such as providing and receiving feedback

regarding new products and services (Kärkkäinen, Jussila, & Leino, 2012). Customer-

related learning can benefit both the supplier and receivers of information, (i.e. the

brand and the customer), as both parties learn by receiving and adopting novel

information and knowledge (Kärkkäinen et al., 2012). This exchange has been referred

to as ‘interactive learning’, described as the informal exchange and sharing of

knowledge resources with suppliers and/or customers that is conducive to the firm

(Meeus, Oerlemans, & Hage, 2001). Social media platforms have been credited with

facilitating processes of interactive learning, including facilitating the mobilisation of

tacit knowledge (Ribiere & Tuggle, 2010), enhancing information and knowledge

sharing (Levy, 2009), and facilitating knowledge acquisition (Schneckenberg, 2009).

Sharing Experiences: Users disseminate personally relevant information, knowledge

and experiences (Brodie et al., 2013; Jaakkola & Alexander, 2014). Users share their

personal experiences and personally relevant information in this way through

storytelling. Traditionally it has been difficult for managers to acquire this information

of customer conversations, opinions and desires (Gorry & Westbrook, 2011). Few

managers could hear customers speak in their own words about their experiences (Gorry

& Westbrook, 2011). However, within social media platforms, the barrier is reduced as

customers freely comment and create stories regarding their brand experiences. In a

63 | P a g e
Chapter 3: Social Media Engagement Behaviour
social media platform, user sharing of their personal experiences and knowledge allows

them to pass along their information about services or products purchased (Black &

Kelley, 2009). Through such forms of content creation, users co-create experiences

within social media platforms. This concept goes beyond the practice of forwarding

brand posts, as users are actively creating the content to share with the social media

community.

Advocating: Advocating is an expression of engagement, which occurs when users

recommend their preference for specific brands, products/services and organisations or

ways of using products and brands (Sashi, 2012). This form of content creation is aimed

at influencing other user’s perceptions, preferences or knowledge regarding the brand

(Jaakkola & Alexander, 2014). Advocating may occur through positive electronic word-

of-mouth (e-WOM) (van Doorn et al., 2010), the significance of which is well

recognised in the marketing and advertising literature (Engel, Kegerreis, & Blackwell,

1969; Gilly, Graham, Wolfinbarger, & Yale, 1998). Social media platforms represent an

ideal tool for e-WOM, as consumers freely create and disseminate brand-related

information in their established social networks composed of friends, classmates and

other acquaintances (Vollmer & Precourt, 2008). Further, advocating focal brands and

brand-related experiences through e-WOM behaviours on social media will likely

impact purchase behaviour and increase customer value (van Doorn et al., 2010).

Socialising: Socialising behaviour denotes the creation of content through two way,

non-functional interactions (Brodie et al., 2013). The social value derived from

membership in social media platforms has been argued to drive the adoption and usage

levels of the platforms (Hennig-Thurau, Malthouse, Friege, Gensler, Lobschat,

Rangaswamy, & Skiera, 2010). Social media platforms, incorporating brand pages,
64 | P a g e
Chapter 3: Social Media Engagement Behaviour
provide greater opportunities for interactions where consumers can derive social value

from computer-mediated interactions with one another (De Vries & Carlson, 2014).

Within social media platforms members perceive other members as similar to

themselves and have the opportunity to interact, meet and communicate with them (Jahn

& Kunz, 2012). It has been suggested that a higher perception of social-interaction

value of social media brand pages may lead to the customer using the page more

frequently, and subsequently becoming more engaged with the brand (De Vries &

Carlson, 2014). Social-interactive engagement occurs in online communities, whereby

users experience intrinsic enjoyment and value the input from the larger community of

users (Calder et al., 2009). Social engagement has been identified as a fundamental

dimension of engagement in online (Calder et al., 2009) and off-line environments

(Altschwager, Conduit, & Goodman, 2013).

Users who exhibit a high level of socialising behaviour interact on a brand’s social

media page and may develop a sense of belonging and feeling of knowing each other

(Park et al., 2009). This form of content is created with the sole purpose of interacting

and communicating with the brand and other members in a social manner. This is

distinct to the behaviour of sharing experiences, in which users share personal relevant

information, knowledge and experiences within the social media platform (Brodie et al.,

2013).

Co-developing: Users contribute to the brand by assisting in the development of new

products, services, brands or brand meanings (Brodie et al., 2013). Users engaged in co-

developing behaviours assist in the collaborative innovation of new products and

services, allowing firms to draw upon customer knowledge, experience and capabilities

(Greer & Lei, 2012). When customers are involved in design and innovation processes,
65 | P a g e
Chapter 3: Social Media Engagement Behaviour
there is a positive impact on new product performance (Menguc, Auh, & Yannopoulos,

2014). Within the social media context, collaborative innovation occurs when users

contribute their knowledge, resources and skills to facilitate the focal firm’s developing

of its offering, through sharing ideas for improved products and services (Jaakkola &

Alexander, 2014). Users also answer questions or quizzes related to the brand within

social media platforms, which provides the organisation with customer insight for future

development of its offerings. The use of questions and quizzes as a form of gamification

in order to engage users in solving problems has been found to increase user

contributions and engagement (Huotari & Hamari, 2012)

Affirming: Affirming denotes the specific creation of content by users with the aim of

disseminating their support, encouragement and acknowledgement of the brand’s

success. Drawing from the self-concept theory (Mehta, 1999; Sirgy, 1982), affirming

behaviour refers to the words and deeds of others that act to reinforce an individual’s

perceptions of competencies, traits and values (Schmidt & Scholl, 2004). Types of

affirming behaviour include positive feedback directed toward the brand, recognition of

skills and worth and recognising significant achievements (Schmidt & Scholl, 2004). In

social media platforms, affirming behaviour occurs in a customer-to-brand manner

whereby users and customers recognise and communicate the brand’s skills, worth and

significant achievements. Affirming behaviour differs from advocating or word-of-

mouth behaviour in that the content is directed toward the brand. Comparatively,

advocating behaviour occurs when users engage in user-to-user interactions with the

aim of recommending brands, products and services (Jaakkola & Alexander, 2014).

Affirming allows users who are highly engaged a way of demonstrating their

appreciation and support for the focal brand.

66 | P a g e
Chapter 3: Social Media Engagement Behaviour
It is posited that the ‘creating’ form of SMEB consists of the six creation behaviours as

described above. The occurrence of creating engagement behaviours can be triggered

when a consumer recognises a need to solve a problem or satisfy a need, performs a

search, identifies relevant social media platforms, and posts a comment. The nature of

the posted comment and resultant conversations characterise the creation of knowledge

seeking, sharing experiences, advocating, socialising, co-developing, and affirming

behaviours. The construct of SMEB recognises that whilst customers can be actively

and positively engaged with the brand through the six creating behaviours, additional

behaviours may occur within a social media environment. These are less active in

nature; however still represent a positively-valenced expression of engagement through

contributing to and consuming brand related content.

Contributing

The second overarching SMEB proposed, ‘contributing’ sees users forward or

contribute to existing content. However, they do not create any additional or new

content in the form of writing a comment or post. Contributing users represent a

moderate level of positively-valenced SMEB. Through functions such as ‘sharing’

content on Facebook, and ‘re-tweeting’ messages on Twitter, users contribute by

forwarding brand content. Users are therefore distributors of pre-existing content,

passing along information to members of their own social networks. Additionally, users

contribute to content by indicating their preferences for specific social media content

through selecting the ‘Like’ function on Facebook and Instagram, tagging friends and

other users in comments, and functions such as the ‘favourite’ option on Twitter.

Through these actions, users contribute to the popularity of social media content and

become message senders for the focal brand, passing on content to actors within their

67 | P a g e
Chapter 3: Social Media Engagement Behaviour
own networks. When users undertake these actions, they not only increase the original

reach and exposure of the social media content, they also become advocates of the focal

brand. This increases the likelihood of friends and other social media users engaging

with the brand (Chu, 2011).

Consuming

Consuming is defined as the passive consumption of brand related content through

reading reviews, discussion and comments, viewing photos, watching videos and

clicking on content and links. Consuming reflects the minimum level of positively-

valenced SMEB. Consumption behaviour is passive, whereby consumers exhibit a level

of engagement however do not actively contribute to or create content. Through the

consumption of content within social media platforms, individuals may extract

individual value. Users’ behaviour is individualistic and independent of other users, and

thus will not impact on other users of the social media platform. Reading discussions

(e.g. to find information) is a form of passive engagement, whereas posting comments is

active engagement (Gummerus et al., 2012; Shang, Chen, & Liao, 2006). Only a small

number of customers actively interact with content and other members, with most

customers using brand communities to consume content as a source of information,

reading messages rather than contributing through likes, shares and comments

(Gummerus et al., 2012). Consuming users may consciously choose not to contribute to

social media content for several reasons including a desire for privacy, time pressures,

and an overload of messages (Nonnecke & Preece, 1999). Hence, within this thesis

there is a distinction in non-contributing behaviour, between consuming behaviours and

dormancy.

Dormancy
68 | P a g e
Chapter 3: Social Media Engagement Behaviour
A dormant user is one who has made zero active or passive contributions to the

community. Users do not behaviourally engage with the brand passively or actively,

through consuming, contributing to or creating content. Rather, dormant users exhibit a

temporary state of inactive engagement (Brodie et al., 2013). This state does not

necessarily reflect inactive cognitive or emotional engagement, but there is no visible

interaction between the user and the brand from the perspective of an independent

observer (e.g. another user). Dormant users are considered as neutral in their

engagement behaviour valence, and do not exhibit negatively-valenced engagement

behaviours such as detachment or destruction. Studies of Facebook user behaviour have

indicated a significant rate of dormancy, with less than five percent of Facebook users

engaging with the brand they are a fan of, regardless of product category (Nelson-Field

& Taylor, 2012). Such a high rate of inactive users presents a challenge for marketers,

particularly those wishing to increase expressions of SMEB. To date, little research has

been undertaken in the social media arena to determine effective communication efforts

and strategies that may act to facilitate superior levels of engagement amongst existing

users who remain dormant.

Detaching

Detaching represents a negatively-valenced SMEB, which involves users actively and

yet privately removing themselves from social media brand pages through selecting to

hide brand related content, or ‘unlike’ and ‘unsubscribe’ from the page. Detachment

represents a moderate level of negatively-valenced SMEB. Detaching users have made

a decision to terminate their interaction with the brand, meaning there is a temporary or

permanent conclusion to the consumers’ behavioural engagement with the brand

community. As users privately and quietly remove themselves from the platform, it is

69 | P a g e
Chapter 3: Social Media Engagement Behaviour
unlikely to impact on other users of the page.

Customer detachment from a brand relationship is observed in interpersonal relationship

scholarly research through the concept of relationship ending (Duck & Perlman, 1985).

Similarly, relationship termination, withdrawal, dissolution, discontinuation, uncoupling

and break-up reflect similar processes of detachment (Stewart, 1998). Referring to

detachment as a process of disengagement, Bowden, Gabbott, and Naumann (2014)

define disengagement as a process, stimulated by a trauma or disturbance leading to

relationship termination, dependent on prior levels of engagement:

“A process by which a customer-brand relationship experiences a trauma or disturbance


which may lead to relationship termination; which involves a range of trigger based events;
which varies in intensity and trajectory; which occurs within a specific set of category
conditions and which is dependent on prior level of customer engagement”. (p.6)

This view of customer disengagement suggests a permanent state of detachment.

However, disengagement or detachment with a focal brand or brand community may be

more temporary in nature. Consumers may choose to re-join the community, following

the trauma or disturbance.

Destructing

Destructive social media users make negatively-valenced active contributions to social

media brand pages that would be visible to other users. Negatively-valenced, destructive

content is created by users within social media platforms with the aim to disseminate

negative word-of-mouth, or e-WOM, and vent negative brand related feelings, causing a

destruction of brand value (Bowden et al., 2014; Hollebeek & Chen, 2014; Plé &

Cáceres, 2010). While co-creation refers to the process in which providers and

customers collaboratively create value, co-destruction refers to the collaborative

70 | P a g e
Chapter 3: Social Media Engagement Behaviour
destruction, or diminishment of value by providers and customers (Plé & Cáceres,

2010). Destructive behaviour can be driven by users’ perception of the brand’s

reputation, product involvement, self-confidence, perceived worthiness of complaining

and the proximity of others, and attitudes to the business in general (Lau & Ng, 2001).

3.3 Chapter Summary


This chapter introduced and developed a new construct termed ‘social media

engagement behaviour’, referred to as SMEB. This construct includes six specific

typologies of SMEB, and considers their valence (positive, negative, and neutral) as

well as the relevant intensity of each type of SMEB. The behaviours are modelled in

Figure 3.1 and defined in Table 3.1.

The proposed construct of SMEB contributes to the literature through the development

of a deeper understanding of the nature of engagement behaviour. It encompasses a new

typology of SMEB: specifically describing creating, contributing, consuming, dormant,

detaching and destructing behaviours. Whilst previous engagement scholars have

explored the customer engagement cycle, a construct for understanding both positive

and negative SMEB has not yet been developed. The typology captures both positively-

valenced engagement behaviours and negatively-valenced engagement behaviours and

considers the intensity of this engagement, reflected at both ends of the spectrum.

Further, the development of the SMEB construct provides clarity to managers who wish

to understand not only why, but also how customers engage with a focal brand.

Building on the previous discussion, an integrative model of social media content and

SMEB is proposed in Chapter 4. The model explores the processes for stimulating

positively-valenced social media engagement behaviour and/or dissuading neutral and

71 | P a g e
Chapter 3: Social Media Engagement Behaviour
negatively-valenced social media engagement behaviour through the use of social

media content. The model hypothesises that social media content can be categorised

into four main groups, based on its level of information (Cvijikj & Michahelles, 2013;

De Vries et al., 2012), entertainment (De Vries et al., 2012; Taylor et al., 2011),

remunerative (Cvijikj & Michahelles, 2013; Lee et al., 2013) and relational (Muntinga

et al., 2011) content. Delivery of these varying forms of content will gratify customer

motives for social media use, therefore resulting in expressions of social media

engagement behaviours, as discussed further in the following chapter.

72 | P a g e
Chapter 4: Conceptual Model Development

CHAPTER 4. Conceptual Model Development


4.1 Introduction
Theoretically grounded academic guidance concerning marketing practice and customer

engagement in new-media social networks is limited. The conceptual model and

hypotheses developed in this chapter explicate the role of social media content in

facilitating engagement behaviour within a social media context. Based on UGT, the

conceptual model shows how social media content can stimulate positively- and

negatively-valenced engagement behaviour in this forum.

This chapter begins by introducing the conceptual model of social media content and

SMEB. The four social media content categories; informational, entertaining,

remunerative and relational are discussed. The relationship between social media

content and SMEB is introduced, which leads to a discussion of the main hypotheses of

the study. Interaction effects are addressed, followed by a discussion of media richness

theory and congruity. The moderating roles of media richness, congruity and

community size are then justified and hypothesised. The hypotheses of the study are

summarised at the end of the chapter in Table 4.1.

4.2 The Conceptual Model


Figure 4.1 illustrates the conceptual model, which positions the four categories of social

media content derived from UGT as antecedents to SMEB. Social media content

includes informational content, entertaining content, remunerative content and relational

content. The impact of each of these content types on each of the types of SMEB

(destructing, detaching, dormancy, consuming, contributing and creating) are tested by

this model. The concepts of media richness, congruity and community size as

73 | P a g e
Chapter 4: Conceptual Model Development
moderators are defined and explored.

Figure 4.1 Conceptual Model of Social Media Content and Engagement Behaviour

In order to understand consumer motives for participation in social media, the classical

components of UGT have been employed. This provides an avenue through which to

understand consumer gratifications sought through engagement with social media

content. The conceptual model depicts the relationships between the four categories of

social media content determined from UGT; informational, entertaining, remunerative

and relational content and the resultant SMEB.

As discussed in Chapter 2, Section 2.3, UGT is an approach to understanding why and

how individuals actively seek out and use specific media to satisfy specific needs (Katz

& Foulkes, 1962). UGT provides a framework through which the motivations of

individuals seeking a specific type of media can be further understood. In a social media

context, users are not passive in their media selection or their use of specific media.

Social media is constructed to enable customer interaction and engagement. It is

proposed that social media content which satisfies the need for information,

74 | P a g e
Chapter 4: Conceptual Model Development
entertainment, remuneration and social interaction will facilitate the way in which

consumers choose to engage with brands and other network users within social media

sites. The relationships between social media content categories and SMEB are

hypothesised in the following section.

4.3 Hypotheses
4.3.1 Informational Content

The relationship between the ability of an advertisement to provide informational

content to viewers and advertising acceptance has been well documented (Bauer &

Greyser, 1968). Whilst the importance of delivering information through advertisements

has been recognised for more traditional media (Rubin, 2002), the role of informational

advertising and content in the online, social domain has only recently received attention.

Attaining various forms of information has been suggested as the most important reason

for consumers to use the internet (Maddox, 1998) . Levels of informativeness and

attitude towards websites have been found to be positively related (Chen et al., 2002).

De Vries et al. (2012), Cvijikj and Michahelles (2013) and Lee et al. (2013) have

empirically demonstrated the relationship between informative content and engagement

behaviour. Informative content negatively impacts levels of user engagement in the

form of likes and comments, when compared to emotional content (Lee et al., 2013).

Similarly, Cvijikj and Michahelles (2013) found that posts which contain information

about the brand cause a lower level of engagement compared to entertaining content.

This could be explained by the fact that highly informational content may be considered

specific to the page in which it is posted, and hence lose its significance when shared by

the fans on their own walls, to friends outside of the brand community who might not be

interested in that particular information. Further, informational content is not generally


75 | P a g e
Chapter 4: Conceptual Model Development
designed to appeal to consumer engagement actions such as commenting and

conversing with other users in the same way as entertaining or relational content.

Informational content, such as a product release date, is less likely to stimulate

conversation amongst fans, compared to content which is entertaining, contains humour,

or poses an interesting question to the audience. Drawing from empirical studies

exploring the link between informational content and consumer response, it is

hypothesised that informational social media content will facilitate positively-valenced

SMEB (H1).

The positive consequences of informational content could be challenged, as

informational content has been previously found to decrease likes and shares (Lee et al.,

2013). Further, customers’ who seek informational value from social media content are

more likely to consume content rather than interact through comments and discussion

(Ko et al., 2005). As such, informational content is further surmised to have a

significant relationship with passive, positively-valenced SMEB:

H1: The presence of informational content facilitates passive, positively-valenced

social media engagement behaviour

Further, it is hypothesised that an optimal level of informational content exists, beyond

which increasing levels of informational content will have a detrimental effect on

positively-valenced SMEB. This expected relationship is derived from research

regarding information overload in advertising messages (Eppler & Mengis, 2004). As

receivers of the content have a limited cognitive processing capacity, when information

overload occurs it is likely that a reduction in decision quality and attention will occur

(Speier, Valacich, & Vessey, 1999). Within computer mediated communication,

76 | P a g e
Chapter 4: Conceptual Model Development
information overload refers to the delivery of too many communications, causing an

increase in density that gives individuals exposure to more communication elements

that they can easily respond to (Hiltz & Turoff, 1985). In both traditional and electronic

media, information overload causes individuals to fail to respond, inaccurately and

incorrectly respond, systematically ignore or filter out the message, or quit (Hiltz &

Turoff, 1985; Sheridan & Ferrell, 1974). Individuals commonly filter and ignore

information as the primary effective way of coping with high levels of information

overload (Eppler & Mengis, 2004). It is thus hypothesised that high levels of

informational content will weaken the relationship with positively-valenced SMEB as

predicted by H1.

H1a: High levels of informational content weaken the relationship with positively-

valenced social media engagement behaviours.

4.3.2 Entertaining Content

The concept of entertaining advertising has been discussed extensively in literature,

with empirical evidence demonstrating that entertaining advertisements lead to positive

attitudes toward the advertisements (Taylor et al., 2011), attitude toward the brand, and

desire to return to the websites (Raney, Janicke, & Tamborini, 2013). Entertaining

advertisements are said to motivate users to consume, contribute to and create brand

related content online (Muntinga et al., 2011).

Social media content is considered entertaining when it includes small talk, banter, or

attempts to appeal to a person’s emotions. This has been empirically found to increase

engagement behaviour in the form of likes and comments (Lee et al., 2013). Further,

entertaining social media content may not focus on the brand or product, but may be

77 | P a g e
Chapter 4: Conceptual Model Development
written in the form of a teaser, slogan or word play, which increases the number of

likes, comments and shares made on Facebook posts (Cvijikj & Michahelles, 2013).

However, entertaining social media content has also been found to have the opposite

effect, failing to facilitate active SMEBs as it may not be related to the brand and does

not offer the consumer any relevant informational value (De Vries et al., 2012).

In this study, entertaining content is proposed to be a driver of positively-valenced

SMEB. It is argued that if a brand post is entertaining, brand fans’ motivations to

engage with the content are met. Hence, brand fans may exhibit a more positive

response toward entertaining brand posts compared to non-entertaining brand posts.

Based on this foundation, it can be hypothesised that entertaining brand posts will

facilitate active, positively-valenced SMEB:

H2: The presence of entertaining content facilitates active, positively-valenced social

media engagement behaviour.

It is further hypothesised that with greater levels of entertaining content, the relationship

with positively-valenced engagement behaviour will be weakened. An optimal level of

entertaining content is expected to exist, beyond which information overload will occur

and the user will be presented with too many cues to correctly process (Hiltz & Turoff,

1985). Hence;

H2a: High levels of entertaining content weaken the relationship with positively-

valenced social media engagement behaviours.

4.3.3 Remunerative Content

The level of remuneration offered to the consumers has been studied as an antecedent of

78 | P a g e
Chapter 4: Conceptual Model Development
consumer decisions to contribute to online communities (Muntinga et al., 2011). Brands

often use monetary incentives including loyalty points, lucky draws, and price

promotions to encourage engagement in online brand communities (Aksoy, van Riel,

Kandampully, Wirtz, den Ambtman, Bloemer, Horváth, Ramaseshan, van de Klundert,

& Gurhan Canli, 2013). Customers may engage in social media as they expect to gain

some kind of reward, such as an economic incentive, job-related benefit or personal

wants (Muntinga et al., 2011). Rewarding or remunerating content may include

monetary incentives, giveaways, prize drawings or monetary compensations (Füller,

Bartl, Ernst, & Mühlbacher, 2006).

Monetary benefits or incentives are not required for community members to make

contributions, and thus it is not hypothesised that a positive relationship exists between

remunerative offers and SMEB. Limited support has been found for the use of monetary

incentives in stimulating engagement (Dumas, Begle, French, & Pearl, 2010). While

monetary incentives have been shown to increase short-term engagement of online

community members, a stronger effect has been observed for passive compared to

active members (Aksoy et al., 2013). Monetary rewards decrease active online

community members intentions to participate, suggesting a reduction of active SMEB

as a result of this content type (Aksoy et al., 2013). While some authors (van Doorn et

al., 2010) suggest that firms can successfully affect customer engagement behaviour by

providing rewards and other incentives to customers, this relationship has not been

empirically tested. Füller et al. (2006) discussed that whilst managers often believe the

offering of monetary incentives such as bonus points, drawing prizes, or sharing product

success results in positive consequences for engagement, they are often mistaken.

Rather, factors such as the ability to learn something new, the possibility to get

79 | P a g e
Chapter 4: Conceptual Model Development
exclusive content and the ability to gain acknowledgement and support from the

community have a far greater impact on community members’ motivation to contribute

to virtual communities (Füller et al., 2006).

Social media content that offers remuneration to fans includes contests and sweepstakes

organised within the Facebook brand community (Cvijikj & Michahelles, 2013). This

content is negatively related to the number of likes on a post, but has been found to be a

significant, positive factor in predicting the number of comments. Remunerative content

had no effect on the number of shares (Cvijikj & Michahelles, 2013). Hence, it is

expected that a low level of engagement would occur as a result of a post containing a

reward or offer, for example ‘consuming’ rather than contributing or creating behaviour

(Muntinga et al., 2011). Content which includes economic or remunerative details such

as price mention or deal/promotion has a negative impact on the number of comments

(Lee et al., 2013). Further, this form of content also has a negative impact on the

number of likes obtained on the post. Hence, it may be argued that content which

provides remunerative or economic benefit to the user leads to less active expressions of

SMEB;

H3: The presence of remunerative content facilitates passive, positively-valenced social

media engagement behaviour.

In line with H1a and H2a as previously discussed, with increased levels of content it can

be hypothesised that users will experience information overload;

H3a: High levels of remunerative content weaken the relationship with positively-

valenced social media engagement behaviours.

80 | P a g e
Chapter 4: Conceptual Model Development
4.3.4 Relational Content

Customers are motivated by social needs when creating user generated content online,

as they find it a comfortable space in which to reveal feelings and share views and

opinions (Leung, 2009). Within Facebook, users are strongly motivated by socialising

needs when participating in specific interest groups (Park et al, 2009). The socialisation

benefits gained include gaining support from other members, meeting new and

interesting people, and a developing sense of belonging to the community (Park et al.,

2009). In virtual communities, socialising demonstrates a level of customer

engagement, depicted by two-way, non-functional interactions through which customers

develop attitudes, norms and community language (Brodie et al., 2013). Customers who

are engaged in social media platforms benefit by gaining an informal sense of

camaraderie with other users (Chen, 2011). Customers highly motivated by socialisation

motivations frequently participate in human-to-human interactions as defined by Ko et

al. (2005), rather than human-message (content) interactions.

It is hypothesised that relational social media content which stimulates interaction

amongst customers will be successful in facilitating positively-valenced SMEB;

H4: The presence of relational content facilitates active, positively-valenced social

media engagement behaviour.

It is also hypothesised that expressions of SMEB may be dependent on the level of

relational content embedded within a post, again supported by the notion of information

overload;

H4a: High levels of relational content weaken the relationship with positively-valenced

social media engagement behaviours.

81 | P a g e
Chapter 4: Conceptual Model Development
4.3.5 Simultaneous Presence of Social Media Content Categories

Within social media, there is potential for content to possess multiple cues and

simultaneously contain entertaining, informative, remunerative and relational content.

The four social media content categories outlined in H1 to H4 are therefore not

mutually exclusive.

In traditional media research authors have found that high levels of informational value,

combined with high levels of entertainment value can increase the likelihood that

consumers will stop viewing a TV commercial (Elpers, Wedel, & Pieters, 2003).

Research supporting this concept suggests that consumers access different processing

styles when exposed to utilitarian and hedonic content, and that these processing styles

may be incompatible when required at a specific point in time (Bless, 2000; Fiedler,

2001; Forgas, 2001). Highly entertaining advertisements are said to require consumers

to focus on their general knowledge structures and free associations, compared to

informational content which requires consumers to focus on the details and data within

the message (Elpers et al., 2003). These processing styles are not mutually exclusive

(Fiedler, 2001). However they do pose conflicting demands on consumers. Based on

this rationale, it has been suggested that when faced with conflicting processing

demands simultaneously (for example, exposure to highly informational and highly

relational content), consumers are more likely to avoid exposure to the content and

therefore disengage.

Previous studies of online brand content and resultant user engagement have also

explored the interaction between content types. Lee et al. (2013) identify a positive

interaction effect between emotional and informational content. This finding conflicts

with Elpers et al. (2003) who found that when these forms of content were combined,
82 | P a g e
Chapter 4: Conceptual Model Development
consumers avoided the content. Adopting the stream of thought proposed by Elpers et

al. (2003) and the conflict of information processing styles (Bless, 2000; Fiedler, 2001;

Forgas, 2001) it can be predicted that as utilitarian and hedonic content place conflicting

demands on consumers, they are less likely to facilitate positive active engagement

behaviours. Hence, it would be expected that utilitarian social media content

(informative and remunerative) presented simultaneously with hedonic social media

content (entertaining and relational) will cause a conflict of processing styles, resulting

in a lack of attention or content avoidance by the user. However, content types that have

required a similar processing style will have a positive effect on engagement,

facilitating positively-valenced SMEB. Thus:

H5a: The simultaneous presence of informational and entertaining content facilitates


negatively-valenced SMEBs.

H5b: The simultaneous presence of informational and relational content facilitates


negatively-valenced SMEBs.

H5c. The simultaneous presence of informational and remunerative content facilitates


positively-valenced SMEBs.

H5d. The simultaneous presence of entertaining and remunerative content facilitates


negatively-valenced SMEBs.

H5e: The simultaneous presence of entertaining and relational content facilitates


positively-valenced SMEBs.

H5f: The simultaneous presence of remunerative and relational content facilitates


negatively-valenced SMEBs.

4.3.6 Moderating Variables

There are three variables depicted in Figure 4.1 which are expected to moderate the

relationship between social media content and SMEB. The hypothesised relationships
83 | P a g e
Chapter 4: Conceptual Model Development
between social media content and SMEB are either directly or inversely related to the

three moderating variables; media richness, community size and congruity.

Media Richness

Media Richness Theory (Daft & Lengel, 1986) is based on the assumption that the goal

of any communication is the resolution of ambiguity and the reduction of uncertainty. It

states that media differ in the degree of richness they possess, measured by the amount

of information they allow to be transmitted in a given time interval. Media Richness

Theory is a widely known theory of media use, and posits that communication

efficiency will be improved by matching media to users’ information needs (Daft &

Lengel, 1986). Media richness is a function of characteristics: the ability to handle

multiple information cues simultaneously, the ability to facilitate rapid feedback, the

ability to establish a personal focus, and the ability to utilise natural language. Whilst a

majority of studies compare richness between forms of media, e.g. telephone vs. direct

mail marketing, the new communication landscape provides marketers an opportunity to

provide both ‘rich’ and ‘lean’ advertising and marketing content, within a single media

type such as a website.

Recently, media richness has been applied in the field of online and digital marketing

(Shaw, Chen, Harris, & Huang, 2009). Online rich media include a range of interactive

methods that display motion and exploit sensory traits such as video, audio and

animation (Rosenkrans, 2009). The term ‘rich media’ provides an umbrella expression

to describe online content that has multimedia elements such as sounds, video, or

content that moves when a user clicks on the page that features the content (Shaw et al.,

2009). Previous research has found that communication media such as face-to-face

meetings have more richness than communication media and written documents as the
84 | P a g e
Chapter 4: Conceptual Model Development
latter lack nonverbal feedback cues such as facial expression, direction of gaze, posture

and dress (King & Xia, 1997). However, in a virtual, social environment, individuals

can perform communication in a style that is similar to face-to-face communication.

This results in an increased richness of content within social media platforms (Cheung

et al., 2011).

Stemming from the foundations of Media Richness Theory, the term ‘vividness’ has

been applied to studies of brand communication in the online sphere. Vividness reflects

the richness of a brand post’s formal features or the degree to which a brand post

stimulates the different senses (De Vries et al., 2012). Vividness can be achieved by the

inclusion of dynamic animations, colours and pictures (Cho, 1999; Fortin & Dholakia,

2005). The degree of vividness can differ in the way that it stimulates multiple senses.

For example a video is more vivid than a picture because a video stimulates sight and

hearing as opposed to just sight (De Vries et al., 2012). Highly vivid banner

advertisements are more effective in generating intentions to click (Cho, 1999) and click

through rates (Lohtia, Donthu, & Hershberger, 2003). Within social media, vividness is

most commonly operationalised on a low to high scale (Cvijikj & Michahelles, 2013;

De Vries et al., 2012). Social media content which contains only text is categorised as

low vividness. Content that is categorised as ‘moderately vivid’ includes text and a

picture or photo. Finally, highly vivid social media content is presented in the form of a

video. Higher levels of vividness have been significantly and positively related to the

number of ‘likes’ on the content (Cvijikj & Michahelles, 2013; De Vries et al., 2012).

It is proposed that the richness of social media content (low, medium and high)

moderates the relationships between social media content and SMEB as outlined in

Hypothesis 6. As scholars have demonstrated significant and positive relationships


85 | P a g e
Chapter 4: Conceptual Model Development
between highly rich content and engagement behaviours such as liking (Cvijikj &

Michahelles, 2013; De Vries et al., 2012) and clicking (Cho, 1999; Lohtia et al., 2003) it

is hypothesised that the moderation will be positive:

H6: The strength of the relationship between social media content and social media

engagement behaviour is directly related to media richness.

Congruity

The extent to which social media content is congruent or incongruent to the brand may

enhance or mitigate different forms of SMEB. It is proposed that entertaining,

informational, remunerative and relational posts vary in their degree of congruity with

the brand.

Congruity concerns the extent that structural correspondence is achieved between the

entire configuration of attribute relations associated with an object, such as a product,

and the configuration specified by the scheme (Meyers-Levy & Tybout, 1989). Authors

have proposed that the concept of congruity is closely related to the theoretical

underpinnings of Hastie’s (1980) Associative Storage and Retrieval Model. This model

proposes that information which is incongruent with one’s expectation would be

recalled better than information which is congruent with one’s expectation. Hastie

(1980) proposed that when consumers receive incongruent information, they spend

more time processing and comprehending the information. The time taken for the

encoding of incongruent information causes the formation of a larger number of

associative paths, between incongruent messages and the present knowledge stored in

the consumer’s memory. Following Hastie’s (1980) propositions, the theory was tested

and supported in a marketing context (Heckler & Childers, 1992; Houston, Childers, &

86 | P a g e
Chapter 4: Conceptual Model Development
Heckler, 1987; Lee & Mason, 1999).

Scholars have looked at the effects of congruity and incongruity of advertising in a

range of contexts. It has been proposed that incongruent information leads to greater

brand recall (Hastie, 1980), enhanced arousal (Gardner, Mitchell, & Russo, 1985),

curiosity and interest (Muehling & Laczniak, 1988), and increased message

involvement (Lee, 2000). Further, when a mismatch occurs between one’s expectations

and the product information presented, consumers are said to engage in more elaborate

analytical processing and evaluation (Goodstein, 1993; Sujan, 1985). Similarly,

information that is moderately incongruent has been found to increase consumer

attention (Halkias & Kokkinaki, 2013). Hence, it is hypothesised;

H7: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to congruity.

Community Size

Research in sociology has shown that increased community size has a negative effect

over the interactions between individuals (Simmel, 1950). Participation in smaller

communities has been found to result in stronger interpersonal relationships and a

greater tendency for social engagement (Dholakia et al., 2004). Members of a smaller

community are likely to be more connected to the brand community, resulting in a

higher level of brand engagement. In larger communities, scholars have suggested that

there is likely to be a loss of essential intimacy required for a successful level of

interaction and participation (McWilliam, 2012).

In smaller online communities, users may join more often for friendship and

socialisation motives, therefore having a higher propensity to engage within the

87 | P a g e
Chapter 4: Conceptual Model Development
community (Algesheimer et al., 2005). It has also been suggested that in smaller

communities, there is a sense of ‘knowing one another’ which leads to stronger and

multifaceted interpersonal relationships between consumers, and a greater interest in

engaging in social activities (Algesheimer et al., 2005; Dholakia et al., 2004).

Moreover, members of smaller online communities have been found to develop higher

community loyalty and engage in word of mouth for the community and for the brand

(Scarpi, 2010). Community trust and community and perceived social value are higher

for users of smaller community sizes, compared to larger community sizes (Hsiao &

Chiou, 2012).

The size of the community is expected to moderate the relationship between social

media content and SMEB. This moderation is expected to be negative, as it is expected

that the strength of the relationship between social media content and SMEB will

weaken as the community size is increased:

H8: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to community size.

4.4 Chapter Summary


Building upon the theoretical background of the study, and literature review presented

in Chapter 2 and Chapter 3, this chapter proposed the conceptual model of social media

content and SMEB. The model posits that social media content categories, derived from

UGT can be positioned as antecedents of SMEB. The conceptual model presented in

this chapter also considers the moderating role of media richness, content congruity, and

community size. The hypotheses of the study as developed within this chapter are

summarised in Table 4.1. The research design and methodology for testing these

88 | P a g e
Chapter 4: Conceptual Model Development
hypotheses is presented in Chapter 5.

Table 4.1 Hypotheses

H# Hypothesis
H1 The presence of informational content facilitates passive, positively-valenced
social media engagement behaviour.
H1a High levels of informational content weaken the relationship with positively-
valenced social media engagement behaviours.
H2 The presence of entertaining content facilitates active, positively-valenced social
media engagement behaviour.
H2a High levels of entertaining content weaken the relationship with positively-
valenced social media engagement behaviours.
H3 The presence of remunerative content facilitates passive, positively-valenced
social media engagement behaviour.
H3a High levels of remunerative content weaken the relationship with positively-
valenced social media engagement behaviours.
H4 The presence of relational content facilitates active, positively-valenced social
media engagement behaviour.
H4a High levels of relational content weaken the relationship with positively-valenced
social media engagement behaviours.
H5a The simultaneous presence of informational and entertaining content facilitates
negatively-valenced social media engagement behaviours.
H5b The simultaneous presence of informational and relational content facilitates
negatively-valenced social media engagement behaviours.
H5c The simultaneous presence of informational and remunerative content facilitates
positively-valenced social media engagement behaviours.
H5d The simultaneous presence of entertaining and remunerative content facilitates
negatively-valenced social media engagement behaviours.
H5e The simultaneous presence of entertaining and relational content facilitates
positively-valenced social media engagement behaviours.
H5f The simultaneous presence of remunerative and relational content facilitates
negatively-valenced social media engagement behaviours.
H6 The strength of the relationship between social media content and social media
engagement behaviour is directly related to media richness.
H7 The strength of the relationship between social media content and social media
engagement behaviour is inversely related to congruity.
H8 The strength of the relationship between social media content and social media
engagement behaviour is inversely related to community size.

89 | P a g e
Chapter 5: Research Design

CHAPTER 5. Research Design


5.1 Introduction
Following the discussion of the hypothesis development and conceptual model in the

previous chapter, chapter five outlines the research design adopted for testing the

conceptual model and hypotheses. The chapter outlines the research objectives and

questions of the study, the philosophical stance and research method decisions. The

process of content analysis including identification of the appropriate theory and

rationale for the study, conceptualisation decisions, operationalisation of measures,

coding, sampling, training, and reliability testing and reporting decisions are presented.

The chapter then presents the methods adopted for hypothesis testing: binary logistic

regression and process analysis, followed by the chapter summary.

5.2 The Research Objective and Questions


The specific objectives of this research were:

1. To gain a deeper understanding of social media content categories through the

application of the UGT perspective.

2. To offer new insights into the characteristics, levels and valence of social media

engagement behaviour.

3. To determine the impact of social media content (categorised as informational,

entertaining, remunerative and relational) on social media engagement

behaviour.

In order to achieve these aims, two phases of research design are required. Phase 1

involves QCA to determine the presence of information, entertaining, remunerative and

relational content within social media posts. In addition, QCA is used to code and

90 | P a g e
Chapter 5: Research Design
classify elements of the dependent variable, SMEB.

Upon completion of QCA and corresponding coding of the independent and dependent

variables of the study, hypothesis testing was conducted. Phase two, hypothesis testing,

was completed using binary logistic regression to test the direct relationships between

social media content presence, levels, and SMEB. Hayes (2013) PROCESS Modelling

was used to test the moderating effects of media richness, congruity and community

size. This approach to quantitative analysis allowed testing of Hypotheses 1 to 7 as

developed in Chapter 4.

5.3 Philosophical Stance


The philosophical stance adopted is an important influencing factor when designing a

research methodology (Weaver & Olson, 2006). A positivist approach to research is

based on knowledge gained from positive verification of observable experience through

scientific methods (Cole, 2006). The positivist philosophical stance supports the notions

of prediction and control, in that there are general patterns of cause-and-effect that can

be used as a basis for predicting and controlling natural phenomena, with the goal of

research being to discover these patterns. The stance also supports empirical verification

through observations and measurement as accurate sources of data. Finally, following a

positivist stance, research is understood to be ‘value-free’ providing a strict

methodological protocol is followed, allowing objectivity to be achieved and subjective

bias to be removed. In this study, strict methodological protocols including the

processes of QCA, binary logistic regression and process analysis were implemented.

Within this study, the author views reality; social media content and SMEB, as

independent of the researcher. This reality is based on quantitative data, derived from

91 | P a g e
Chapter 5: Research Design
social media content and customer behaviours collated over a twelve month period.

Based on this positivist ontology, the research objectively explains this reality (social

media content and its relationship with SMEB). Under the positivist model, this

objective knowledge can be provided through science and data, and the purpose of the

researcher is to independently expose the objective truths (Weaver & Olson, 2006).

A key tenet underlying the positivist view is that only science may produce objective

knowledge. Therefore, the purpose of the research is to capture and accurately reflect

objective truth and reality, independent of the researcher. The role of the researcher is

therefore as an outsider or objective observer, who gathers data and reports objectively

on the data. As the positivist position is situated within the epistemological tradition of

objectivism, where objects in the world have meaning that exists independently from

any subjective consciousness of them (King & Horrocks, 2010), the research design

aims to provide objective knowledge, unbiased by the researcher and research process.

This stance is adopted due to its appropriateness and applicability to this research study,

with the quantitative methodology discussed in the following sections based on the

positivist paradigm.

5.4 The Research Methods


The following sections outline the nature of the research problem and context, and the

corresponding research approach to empirical investigation.

5.4.1 Context of the Study

This study is conducted with data derived from the Australian wine industry. Wine is an

experiential product (Bruwer & Alant, 2009), and as Australian wine brands compete to

attract and retain consumers, many are embracing social media to reach their consumers

92 | P a g e
Chapter 5: Research Design
and communicate their brand experience, quality and personality (Vinography, 2012).

Further, research has suggested that 90 percent of wine drinkers use Facebook for at

least 6.2 hours per week (Breslin, 2013). Some wine brands are achieving success

through social media, with documented examples demonstrating that small and large

wineries have achieved a positive return on investment through the implementation of

successful social media strategies. Several scholarly studies have explored social media

practices within the wine industry. Of wineries studied in Australia, Canada, New

Zealand, Spain, Italy, South Africa and the US, 35% have reported using social media,

with the primary reasons of communicating with customers about events at the winery,

and promoting wines (Alonso et al., 2013). Scholars have also suggested that social

media assists with wine sales as word of mouth is particularly effective among wine

consumers (Leigon, 2011), with the socialisation aspect of social media acting as an

appropriate fit with wine, allowing consumers to exchange information and encourage

others to try different wines (Wilson & Quinton, 2012).

Marketing practitioners have been quick to recognise the value of social media

platforms, rapidly integrating such platforms into the marketing mix (Sinclair, 2014;

Stelzner, 2014). There are there are currently more than 2,500 Australian and New

Zealand wineries with a presence on Facebook (Mastermind, 2015). However, many

practitioners have identified a lack of awareness and knowledge regarding effective

social media strategy, creating a significant challenge as practitioners navigate through

this forum with little guidance and empirical understanding (Stelzner, 2014).

Regardless, over 80 percent of brands are said to use the social media platform

Facebook, instigating a significant crowding effect as the competition for customer

attention within social media platforms intensifies (Koetsier, 2013). Customers are
93 | P a g e
Chapter 5: Research Design
inundated with the proliferation of social media content, causing a challenge for wine

brands as they attempt to succeed in this environment. Without empirical understanding

of how to effectively engineer content, managers will fail in their endeavours to attract

and engage social media users (Lee et al., 2013).

Whilst studies have demonstrated that social media has been accepted and is widely

used as a communication tool in the wine industry, research concerning customer

engagement with wine brand communication via social media remains scarce. In order

to explore SMEB within social media, this study extracts and analyses engagement

behaviour through the use of Facebook Insights data and NCapture, discussed in the

following section.

5.4.2 The Research Design

The research design consists of two phases. The first phase comprises a quantitative

content analysis of social media content and SMEB. The data for this phase is collected

from Facebook Insights and NCapture (see Section 5.4.3). The quantitative content

analysis was conducted following Neuendorf’s (2002) process. This involved

determination of the appropriate theory and rationale to guide the categories,

conceptualisation decisions, operationalisation of measures, development of human and

computer coding schemes, sampling, training and initial reliability, coding, tabulation

and reporting. Each of these stages is discussed throughout Section 5.5.3 of this chapter.

The second phase of the research design involved hypothesis testing using binary

logistic regression, and moderation testing using (Hayes, 2013) conditional process

analysis. This will be discussed later in Section 5.7.

94 | P a g e
Chapter 5: Research Design
5.4.3 Data Collection Sources

Data required to complete the two phases of the research were collected from two

sources: Facebook Insights and NCapture.

Facebook Insights

Facebook Insights is a tool provided to administrators of Facebook brand pages to

enable high-level monitoring of the activities on the Facebook page. Facebook Insights

allows administrators to download data concerning the performance of a post, such as

the number of people the post reached, the number of people who clicked the post, and

the number of people who liked, commented on or shared the post. If the post is a video,

Insights data also shows the total number of video views, and length of video views.

Insights data provides page administrators with an ‘Engagement Rate’. Facebook

defines the ‘Engagement Rate’ as a post level metric, calculated as the percentage of

people who saw a post (post reach) and liked, shared, clicked or commented on it

(Facebook, 2015).

Despite these capabilities, there are some limitations to the use of Facebook Insights

data. Firstly, the tool is provided exclusively to administrators of the Facebook page,

meaning analysis of multiple brands data, or competitor data is allowed only with

permissions and access granted by the page owner. Secondly, whilst Facebook Insights

data collects the ‘number of comments’ as a metric of engagement, the content of these

comments is not captured. A post may have received a significant number of comments

but Insights data will not display the content or author details of the comment. Finally,

the features of Facebook Insights are controlled by Facebook. This may cause a

limitation in that a change in policy could mean that a metric considered valuable to a

study may become no longer available, or a change in measurement structure may thus

95 | P a g e
Chapter 5: Research Design
affect the analysis. For example, in March 2014 Facebook removed the ‘People Talking

About This’ (PTAT) metric from Page Insights. The PTAT metric split into separate

elements including page likes and engagement rate (Facebook, 2014). Many of these

shortcomings of the Facebook Insights tool can be overcome through the simultaneous

use of the program ‘NCapture’.

NCapture

NCapture is a web browser extension, developed by QSR International. It allows

researchers to quickly and easily capture content including web pages, online PDF’s and

social media for analysis within NVivo 10. NCapture allows the downloading of

Facebook wall posts and comments from any URL, meaning that the access and

permissions required when relying on Facebook Insights data are no longer necessary.

NCapture collects similar data to Facebook Insights, including the content, type and

timing of brand page posts, number of likes and number of comments. NCapture has an

advantage over Facebook Insights in its ability to collect the content and authorship of

fan comments, relevant to each moderator post. Depending on the individual fan’s

privacy settings, NCapture also attempts to collect demographic data such as gender,

age and location for the users who engage with a post. Data missing from the NCapture

extension includes the number of shares made on a post, post reach, the number of

clicks, video views and amount of negative feedback received by the post, all functions

which are available through Facebook Insights.

5.4.4 Data Collection

By combining the available data sources of Facebook Insights and NCapture, a

comprehensive collection of behavioural data related to social media content and SMEB

was created. Table 5.1 provides an overview of the available metrics from Facebook

96 | P a g e
Chapter 5: Research Design
Insights and NCapture data.

Table 5.1 Facebook Insights and NCapture Data Metrics

Metric Data Description


source
Post Type Both Status, link, photo, video
Comments Both Total number of comments made in response to the page post
Likes Both Total number of likes received by the page post
Shares Insights Total number of shares received by the page post
Created Time Both Date, day and time that the post was created
Clicks to play Insights Relevant to video post type only. Total number of times the
video was clicked to play
Link click Insights Total number of clicks on a link within the page post
Other click Insights Number of clicks on the page post (excluding link clicks,
clicks to play video and photo view clicks)
Photo view Insights Total number of times the page post photo was viewed
Post reach Insights Post reach is the number of people who have seen the post.
The post counts as reaching someone when it's shown in the
“News Feed”. Figures are for the first 28 days after a post was
created and include people viewing the post on desktops and
mobiles.
Negative Feedback Insights Total number of negative feedback clicks, separated into four
metrics; ‘hide post’, ‘hide all posts’, ‘report as spam’ and
‘unlike page’
Individual or brand NCapture The name, brand or other page tagged within the post
tagged
Picture NCapture A link to the picture/photo file provided with the brand’s post
Video NCapture A link (URL) to the video provided with the brand’s post
Commenter username NCapture Username of comment author
Comment text NCapture Content of the comment made by author
Comment likes NCapture Number of likes received on the comment
Comment time NCapture Time the comment was made
Commenter gender and NCapture Gender and age of the comment author if listed in personal
age* profile
Commenter location* NCapture Location of comment author when comment was posted
Commenter NCapture Relationship status of comment author if listed in personal
relationship status* profile
Commenter NCapture Hometown of comment author if listed in personal profile
hometown*
Commenter religion* NCapture Religion of comment author if listed in personal profile
*dependent on user privacy settings.

As access to Facebook page insights is restricted to page administrators, an introductory

email was sent to wine brands outlining the study and requesting access to the required

97 | P a g e
Chapter 5: Research Design
data (see Appendix A). Facebook data for 12 Australian wine brands that responded and

agreed to participate in the study was collected over a twelve month period, beginning

on the 1st of January 2013 and concluding on the 31st of December 2013. The brand

profiles and number of posts collected is presented in Table 5.2. The names of the

brands included in the study are not provided due to confidentiality agreements.

The total number of posts in the data set was 2,236. The total number of fans across the

12 brand pages was 54,069. The number of fans was recorded as the current, exiting

number of fans at the time the dataset was downloaded from the Facebook Insights

platform.

Table 5.2 Brand Profiles

Brand Region No. Page fans at time of No. posts (Jan 01


data collection 2013 – Dec 31 2013)
A Barossa Valley, SA. 916 25
B Barossa Valley, SA. 1,348 355
C Yarra Valley, VIC 1,330 145
D Margaret River, WA 4,500 295
E McLaren Vale, SA 7,496 383
F McLaren Vale, SA 13,958 226
G McLaren Vale, SA 12,551 179
H McLaren Vale, SA 1,434 47
I McLaren Vale, SA 3,749 177
J Eden Valley, SA 3,684 191
K Barossa Valley, SA 1,495 106
L Margaret River, WA 1,608 107
TOTAL 54,069 2,236

Facebook Page Insights also allowed access to specific post metrics for each of the

2,236 posts made in 2013 as displayed in Table 5.3.

Table 5.3 Facebook Insights Post Metrics

Insights Post Description


Metric
Type Status, link, photo, video

98 | P a g e
Chapter 5: Research Design
Comments Total number of comments made in response to the page post
Likes Total number of likes received by the page post
Shares Total number of shares received by the page post
Clicks to play Relevant to video post type only. Total number of times the video way clicked to
play
Link click Total number of clicks on a link within the page post
Other click Number of clicks on the page post (excluding link clicks, clicks to play video and
photo view clicks)
Photo view Total number of times the page post photo was viewed
Post reach Post reach is the number of people who have seen the post. The post counts as
reaching someone when it's shown in the “News Feed”. Figures are for the first 28
days after a post was created and include people viewing the post on desktop and
mobile.
Negative Total number of negative feedback clicks, separated into four metrics; ‘hide post’,
Feedback ‘hide all posts’, ‘report as spam’ and ‘unlike page’
Created Time Date, day and time that the post was created

Analysis of the content and nature of each comment allowed for a more rigorous

understanding of SMEB. As Facebook Insights data does not include access to fan

comments, NCapture was employed. A total of 5,699 comments were made in response

to the 2,236 posts as shown in Table 5.4.

Table 5.4 Number of Comments by Brand

Brand Region No. Page fans at time of Total number of comments


data collection
A Barossa Valley, SA. 916 40
B Barossa Valley, SA. 1,348 283
C Yarra Valley, VIC 1,330 331
D Margaret River, WA 4,500 349
E McLaren Vale, SA 7,496 701
F McLaren Vale, SA 13,958 2216
G McLaren Vale, SA 12,551 382
H McLaren Vale, SA 1,434 165
I McLaren Vale, SA 3,749 173
J Eden Valley, SA 3,684 652
K Barossa Valley, SA 1,495 171
L Margaret River, WA 1,608 206
Total 54,069 5,569

In order to analyse the content of the 2,236 posts collected, and the comments made

with respect to those posts, a quantitative content analysis (QCA) was conducted, as

described in the following section.


99 | P a g e
Chapter 5: Research Design

5.5 Content Analysis


This section defines and describes the quantitative content analysis process adopted

within this study.

5.5.1 Defining Content Analysis

Content analysis is most commonly defined as a research technique for the objective,

systematic and quantitative description of the manifest content of communication

(Berelson, 1952). The objective, systematic and quantitative nature of content analysis

remains central to other definitions of the concept, including Carney (1972) who

describe content analysis as a research technique for making inferences by objectively

and systematically identifying specified characteristics of messages, and Stone,

Dunphy, Smith, and Ogilvie (1968) who describe content analysis as a research

technique for making inferences by systematically and objectively identifying specified

characteristics within texts.

This research follows Neuendorf’s (2002) approach to quantitative content analysis

(QCA), suitable for this study due to its focus on summarising the quantitative analysis

of messages. Under this approach, QCA relies on aspects of scientific method, including

attention to objectivity-intersubjectivity, a priori design, reliability, validity,

generalisability, replicability and hypothesis testing. Noteworthy about Neuendorf’s

definition of content analysis is the argument that media content analysis is quantitative

research, rather that qualitative. QCA is widely accepted as the systematic and

replicable examination of symbols of communication, which have been assigned

numeric value according to valid measurement rules, and the analysis of those

relationships involving those values using statistical methods, in order to describe the

communication (Riffe, Lacy, & Fico, 2014).


100 | P a g e
Chapter 5: Research Design
QCA was deemed suitable in this study as it allows for a non-intrusive research method

incorporating examination of a wide range of data over an extensive time period.

Another benefit of QCA is that it can be conducted frequently (Neuendorf, 2002).

5.5.2 Purpose of Content Analysis

There are five main purposes of conducting a QCA of Facebook brand posts and SMEB

relevant to this study. Firstly, QCA allows the researcher to describe substantive

characteristics of message content. The second purpose of QCA is to describe formal

characteristics of message content. Thirdly, QCA allows researchers to make inferences

about the producers of content, in this study, the wine brands. Fourth, QCA allows

researchers to make inferences about the audiences of content, or the social media users

demonstrating SMEB. Finally, through QCA, researchers can predict the effects of

content on audiences (Berelson, 1952).

5.5.3 Content Analysis Process

The first necessary step in the research design is to analyse social media content

according to the four categories as predetermined through the application of UGT. This

involves a process of QCA, adopting both human and computer coding techniques.

Designing and conducting a content analysis involves an 8-step process as suggested by

Neuendorf (2002); determination of the appropriate theory and rationale for the

research, conceptualisation decisions and operationalisation of measures, coding

decisions, sampling, training and initial reliability, coding, and tabulation and reporting

as displayed.

101 | P a g e
Chapter 5: Research Design
Figure 5.1 Quantitative Content Analysis Process

Step 1. Theory and Rationale

The preliminary step in designing a QCA involves determining what content is to be

examined, and why. This involves consideration of the theories and perspectives that

indicate that the particular message content is important to study. This step also

involves determining the hypotheses of the study. The content analysis builds upon the

theory and rationale of the study as developed throughout Chapter 2 regarding UGT.

Step 2. Conceptualisation decisions

The second step of designing QCA involves determining what variables will be used in

the study and how they are defined conceptually. There are 12 conceptual variables in

the study. The four independent variables of the study are derived from the

underpinning UGT; informational, entertaining, remunerative and relational content

(refer to Chapter 2, Section 2.3.2 for conceptualisation). The SMEB construct as

developed and conceptualised within Chapter 3 provides the dependent variable of the

102 | P a g e
Chapter 5: Research Design
study, comprised of six typologies of behaviour: creating, contributing, consuming,

dormancy, detaching and destructing. The three moderating variables: media richness,

congruity and community size are conceptualised and discussed in Chapter 4, Section

4.3.6.

Step 3. Operationalisation of measures

Following the conceptualisation decisions made regarding each variable of the study as

presented in the preceding section, the next stage of conducting a content analysis

involves determining the operationalisation of measures. In order for the research to be

successful, codes and corresponding measures must exhibit three traits. Firstly, they

must be exhaustive. Every aspect of the sample (social media content and SMEB) that is

of relevance to the research must have an identifying code. The second rule of coding is

that each code must be exclusive. Therefore, codes cannot overlap in definition. Finally,

the coding categories must be enlightening. It is crucial that codes deconstruct the focal

content in a way that would be analytically relevant and interesting. In order to ensure

this, codes should be based on previously established norms in the literature (Riffe et

al., 2014) in addition to being relevant to the research questions.

The measures selected for each variable matched the conceptualisations as specified in

Step 2. The researcher then selected what unit of data collection was to be used. Further,

the researcher tested if the variables were measured well (categories that are exhaustive

and mutually exclusive). This study adopted custom dictionaries for text analysis in

order to measure the presence of social media content.

Dictionaries for text analysis

A dictionary is a set of words, phrases, parts of speech, or other word based indicators

103 | P a g e
Chapter 5: Research Design
that is used as the basis for a search of texts (Neuendorf, 2002).

For the purpose of this study, development of a ‘custom dictionary’ was deemed

appropriate. The dictionaries constructed by the researcher are referred to as custom

dictionaries and within this study included variables selected from theory, past research

and researcher immersion in the message pool. By using a large number of narrowly

defined definitions (e.g. ‘discount’ and its synonyms such as reduction, price cut, mark

down, sale) the researcher had the option of creating a variety of flexible index

combinations of the dictionaries (Neuendorf, 2002). Dictionaries are often emergent

from the data, as is the case in this study. This approach to dictionary construction is to

base the lists on actual word frequencies from the message sample (Neuendorf, 2002).

This was achieved through a basic quantitative output from the message sample, a word

count, which reported the frequency of each word occurring in a text of a set of texts.

The qualitative data analysis computer software program, ‘NVivo 10’ was used to

design and develop the custom dictionaries applied for this study. A word frequency

report of the text-based data within the sample was created using the Word Frequency

Query function of NVivo10. The test demonstrated the top 200 words used across the

dataset of social media posts (n = 2,236). The frequency report is presented in Appendix

B and was used to develop and refine the coding schemes and operationalisation of

social media content. The coding schemes for all variables in the study are provided in

the following sections and corresponding tables.

Social media content operationalisation

In order to understand SMEB with social media content, the classical components of

UGT have been employed. This provides the possibility to understand the responses to

104 | P a g e
Chapter 5: Research Design
different dimensions of gratification, which lead towards customer engagement. The

proposed theoretical model suggest that if the content posted by a page moderator on a

Facebook brand pages satisfies particular user needs and is designed well, this would

lead to a higher level of positively-valenced SMEB.

As discussed in Chapter 4, Section 4.3, it is hypothesised that marketing communication

content can be categorised into four main groups. These groups are informational

content, entertaining content, remunerative content and relational content. Each of the

four categories of social media content as shown in Table 5.5 and their relative coding

schemes are described in the following sections.

Table 5.5 Social Media Content Categories

Social Media Content Categories


Informational Entertaining
Remunerative Relational

Informational content: This form of content relates to content aimed at specifically

delivering category, brand and product related information to community members.

Informational content may contain details on price, availability, location and product

names (Lee et al., 2013). Further, information content may contain explanatory images

referring to the brand’s location, facilities and products. Information may also relate to

brand contact details such as the provision of contact phone numbers, email addresses,

links to a website and opening hours where applicable. In order to capture these

elements, the coding scheme displayed in Table 5.6 has been developed based on

previous research testing informational content presence (Cvijikj & Michahelles, 2013;

De Vries et al., 2012; Lee et al., 2013).

105 | P a g e
Chapter 5: Research Design
The informational content category contains 24 codes, shown in Table 5.6.

Additionally, this table shows the custom dictionary applied to mechanically generate

indications of the occurrence of this content. For items such as product variety and

product region, Australian wine industry statistics were consulted. According to Wine

Australia (2015), there are 39 wine varieties grown in Australia. These 39 varieties are

therefore included as the custom dictionary items for product variety. Further, there are

88 wine growing regions within Australia, included as the custom dictionary for product

region (Wine Australia, 2015).

Table 5.6 Informational Content Codes

Informational Dictionary for Text Analysis


Content Codes
1 Brand name [insert brand name]
2 General Information Newspaper and magazine press coverage, new website announcements,
media mentions, hiring and job availability advertisements
3 Product image Image contains a picture of the product: wine bottle, wine label, glass of
wine
4 Vineyard image Image contains a picture of the vineyard
5 Winery image Image contains a picture of the winery: winery facilities, production
6 Price [$], [price], [dollar]
7 Website Post contains a link or reference to the company website [http] [www]
[.com]
8 Venue image Image contains a picture of a review or award: medal, wine review
screenshot or newspaper/magazine clipping, trophy
9 Product review Image contains a picture of a review or award: medal, wine review
image screenshot or newspaper/magazine clipping, trophy
10 Product award Image of a trophy, medal or certificate awarded to the brand.
image
11 Tasting and [tasting], [taste], [tried], [samples], [try], [trying]
sampling
12 Product variety [Chardonnay], [Pinot Grigio], [Riesling], [Sauvignon Blanc], [Viogner],
[Chenin Blanc], [Gewürztraminer], [Semillon], [Verdelho], [Cabernet
Sauvignon], [Pinot Noir], [Tempranillo], [Carmenere], [Durif],
[Grenache], [Sangiovese], [Zinfandel], [Mouvedre], [Mataro], [Syrah],
[Savignin], [Traminer] [colombard] [Muscat Gordo Blanco] [Muscat a
Petits Grains Blanc] [Malbec]. [Nebbiolo] [Ruby Cabernet] [Petit Verdot]
[Dolcetto] [Durif] [Barbera] [Cabernet Franc] [Muscat a Petits Grains
Rouge] [Merlot] [Pinot Gris] [Roussane] [Sultana] [Trebbiano] [Arneis]
[Crouchen] [Marsanne] [Tarrango] [Touriga]
13 Product [Barossa Valley], [Eden Valley], [High Eden], [Currency Creek],
region/origin [Kangaroo Island], [Langhorne Creek], [McLaren Vale], [Southern
Fleurieu], [Coonawarra], [Mount Benson], [Padthaway], [Wrattonbully],
[Robe], [Bordertown], [Riverland], [Adelaide Hills], [Lenswood],

106 | P a g e
Chapter 5: Research Design
[Piccadilly Valley], [Adelaide Plains], [Clare Valley], [North West],
[Tamar Valley], [Pipers River], [East Coast], [Coal River], [Derwent
Valley], [Southern Bendigo], [Goulburn Valley], [Nagambie Lakes],
[Heathcote], [Strathbogie Ranges], [Upper Goulburn], [Gippsland],
[Alpine Valleys], [Beechworth], [Glenrowan], [Rutherglen], [Murray
Darling], [Swan Hill], [Geelong], [Macedon Ranges], [Mornington
Peninsula], [Sunbury], [Yarra Valley], [Grampians], [Henty], [Pyrenees]
[Peel], [Perth Hills], [Swan Valley], [Blackwood Valley], [Geographe],
[Great Southern], [Albany], [Denmark], [Frankland River], [Mount
Barker], [Porongurup], [Manjimup], [Margaret River], [Pemberton]
[Murray Darling], [Perricoota], [Riverina], [Swan Hill], [Central Ranges],
[Cowra], [Mudgee], [Orange], [Hunter Valley], [Broke Fordwich],
[Northern Rivers], [Hastings River], [Northern Slopes], [South Coast],
[Shoalhaven Coast], [Southern Highlands], [South Australia], [Victoria],
[New South Wales], [Western Australia], [Tasmania], [Australian Capital
Territory], [SA], [WA], [NSW], [VIC], [TAS], [ACT]
14 Product [range] [wine] [product]
15 Product making and [winemaking], [ferment], [crop], [pick], [harvest], [crush], [bottle],
processing [bottling], [press], [rack], [barrel], [blend], [vintage], [veraison], [bud
burst], [fertilise], [spray], [plant], [prune], [decant]
16 vineyard [vineyard], [vines], [winery]
17 Opening hours [open], [closed], [hours], [opening], [times], [shut], [am], [pm]
18 Year made [19XX], [20XX]
19 Contact details [phone], [email], [contact], [address], [location], [website], [get in touch],
[reach], [connect]
20 Brand Fact/News [did you know], [fact], [news], [update], [blog]
21 Service [service], [facility], [facilities], [venue], [event], [function], [occasion],
[wedding], [party], [celebration], [set up], [setting up]
22 Wine show, awards [wine show], [win], [won], [award], [awarded], [received], [achieved],
and reviews [successful], [medal], [trophy], [result], [points], [score], [review], [silver],
[gold], [bronze], [presented], [presenting], [star], [judge], [named],
[listed], [finalist], achievement], [success], [rating], [wine of the year],
23 Event [event], [tickets], [festival], [fork in the road], [sea and vines]
24 Product description [red], [white], [fruit], [tannin], [oak], [fresh], [clean], [crisp], [elegant],
[soft], [smooth], [bold], [chocolate], [rich], [full bodied], [yum], [tasty],
[delicious,] [spice], [zest], [acid], [aroma], [dense], [palate], [flavour],
[fragrance], [balanced], [caramel], [complementing], [citrus], [chalky],
[characters], [notes], [raisin], [mocha], [tannic], [toffee], [tannin],
[vibrant], [colour], [sweet], [sugar]

Entertaining content: Social media content can be entertaining when it includes small

talk, banter, or attempts to appeal to a person’s emotions. Further, entertaining content

may not focus on the brand or product, but may be written in the form of a teaser,

slogan or word play, which has been found to increase the number of likes, comments

and shares made on Facebook posts (Cvijikj & Michahelles, 2013).

In order to capture these elements, the coding scheme displayed in Table 5.7 has been

107 | P a g e
Chapter 5: Research Design
developed. Codes of humour and slang are derived from Lee et al. (2013) and Cvijikj

and Michahelles (2013). Entertaining content may include humour, images of animals

and memes. The entertaining content category contains 13 codes, shown in Table 5.7.

The table shows the custom dictionary applied to generate indications of the occurrence

of this content.

Table 5.7 Entertaining Content Codes

Entertaining Content Dictionary for Text Analysis


Codes
1 Food/Recipe [recipe], [food], [cooking], [baking], [breakfast], [lunch], [dinner],
[oven], [stove], [boil], [grill], [cooked], [eat], [chef], [chicken],
[duck], [peach], [chocolate], [dessert], [morning tea], [porchetta],
[pork belly], [chorizo], [scallops].
2 Emoticon [] [;-)]
3 Weather [weather], [forecast], [sun], [shine], [rain]. [cold], [wind], [chilly],
[frosty], [sunshine], [humid], [mild], [freezing], [icy], [foggy], [hot],
[heat], [cloudy], [stormy], [winter], [summer], [spring], [autumn],
[hail], [snow], [storm], [fire], [rainbow], [sleet], [cloudy], [thunder],
[lightening], [fog], [sunrise], [sunset], [degrees], [temperature].
4 Humour [Fun], [funny], [banter], [joke], [gag], [happy], [joking], [kidding],
[April fools], [hilarious], [cool], [whimsical], [exciting], [haha],
[hehe], [entertain], [laugh], [giggle], [humour], [priceless],
[amusing], [laughable], [laughing]
5 Interesting/Fun Image contains and interesting artefact, relates to the history of the
fact/Historic image brand or provides a fun fact
6 Scenic Image Image is a scenic photo of the vineyard
7 Occasion image Image includes customers or staff at event, special occasion or party
hosted by the brand
8 Food and produce image Image includes pictures of food, produce and recipes used by the
brand
9 Celebrity Image includes a celebrity of popular figure
10 Meme Image Image or picture, typically humorous in nature, often in cartoon or
pictorial form
11 Animal Image Image contains a picture of an animal or pet
12 Animal [Cat], [dog], [kitten], [puppy], [pet], [animal [bird], [kitty] [budgie]
13 Slang [Lol], [omg], [jk], [wtf], [l8r], [plz], [ttyl], [cheers], [guys], [wow],
[arvo], [aussie], [gr8], [mate], [m8]

Remunerative content: Rewarding or remunerative content may include monetary

incentives, special offers, giveaways, prize drawings, monetary compensations (Füller,

2006), contests and sweepstakes (Cvijikj & Michahelles, 2013).

The remunerative content overarching code contains 5 indicator codes, shown in Table
108 | P a g e
Chapter 5: Research Design
5.8. Additionally, this table shows the custom dictionary applied in order to generate

indications of the occurrence of remunerative content.

Table 5.8 Remunerative Content Codes

Remunerative Dictionary for Text Analysis


Content Codes
1 Deal/Offer [Special], [discount], [exclusive], [deal], [sale], [promotion], [clearance],
[bargain], [on sale], [marked down], [low price], [free], [gift]
2 Purchase [Buy], [order], [purchase], [order form], [shop], [store]
instructions
3 Competition image Image contains details and instructions about a competition/contest and/or
prize
4 Sale/Promotion Image contains details about a sale, discount, promotion or special price.
image
5 Competition [Win], [reward], [free], [prize],

Relational Content: According to McQuail (1983), gratifications of integration and

social interaction involve members gaining insight into the circumstances of others,

social empathy, identifying with others, gaining a sense of belonging, finding a basis for

conversation and social interaction, helping carry out social roles, and enabling a user to

connect with family, friends and society.

Relational content includes the use of emotion. In order to develop a custom dictionary

of emotion, the vocabulary of emotions developed by Drummond (2004) was

incorporated into the coding scheme. This dictionary is comprised of ten emotions;

happy, caring, depression, inadequateness, fear, confusion, hurt, anger, loneliness and

remorse. It is posited that social media content containing emotion will reflect a level of

relational integration as suggested by McQuail (1983), providing an insight into

circumstances of others and opportunities for community members to express their

social empathy and find a basis for conversation and interaction. In addition to emotion,

relational content is demonstrated when posts pose a question to the audience in order to

stimulate conversation. Similarly, a quiz or game embedded within the content is

109 | P a g e
Chapter 5: Research Design
expected to stimulate interaction amongst members. Mentions of holidays and events

are also relational in nature. Relational content also includes posts which ask for action

by the fans. These posts use terms such as ‘comment if__’, share if__’ and ‘like if__’

and are designed in order to facilitate conversation and sharing amongst fans. Images

within content which include people are also coded within this category.

The relational content category contains 15 codes, shown in Table 5.9. Further, the

custom dictionary applied to the social media content is provided.

Table 5.9 Relational Content Codes

Relational Content Dictionary for Text Analysis


Codes
1 Question [?], [question] [ask you] [what do you think] [can you suggest]
[suggestions] [ideas] [help]
2 Congratulations [congrats], [congratulations], [well done], [thanks], [thank you].
and thanking fans
3 Quiz/Game [Quiz], [game], [test], [guess], [challenge]

4 Holiday/Event/Da [Birthday, [Christmas], [Easter], [Boxing Day], [New Year], [Australia


y Day], [Good Friday], [Anzac Day], [Queen’s Birthday], [Labour Day],
[holiday], [public holiday], [Melbourne Cup], [April fool], [Father’s Day],
[Mother’s Day], [Monday], [Tuesday], [Wednesday], [Thursday], [Friday],
[Saturday], [Sunday], [festive season]
5 Affection – x and [xo], [xx], [x]
o
6 Ask for action [comment if], [like if], [share if]

7 Child/baby image Image contains a picture of a child or baby

8 Inspirational/ Image contains an inspirational or motivational quote, wordplay or text


motivational
quote
9 Customer image Image contains a single customer or group of customers

10 Employee image Image contains a single employee or group of employees

11 Community Image contains a reference to community involvement through local events,


involvement charities and causes
image
12 Friends and fans [friends], [fans], [customers], [supporters]

13 Employee name Post includes a name of employee, customer or pet


[Tim], [Nigel], [Rebecca], [Christie], [Emily], [Tony], [Rachel], [Marc],
[Pamela], [George], [Glen], [Claire], [Adam], [Travis], [Steve], [James],
[Liam], [Eric], [Johann], [Charles], [Wendy], [Michael], [Jeremy],
110 | P a g e
Chapter 5: Research Design
[Corrina], [Brioni], [Kieran], [Don], [D’arry], [Chester], [Jack], [Jay],
[Smithy], [Robert], [Dan], [Paul], [Sam], [Hayley], [Mel], [Ryan],
[Andreas], [Prue], [Justine].
14 Emotion 1 - Delighted, ebullient. ecstatic, elated, energetic, enthusiastic, euphoric,
Happy excited, exhilarated, overjoyed, thrilled, tickled pink, turned on, vibrant,
zippy, aglow, buoyant, cheerful, elevated, gleeful, happy, in high spirits,
jovial, light-hearted, lively, merry, riding high, sparkling, up.
14 Emotion 2- Caring Adoring, ardent, cherishing , compassionate, crazy about, devoted, doting,
fervent, idolizing, infatuated, passionate, wild about, worshipful, zealous,
admiring, affectionate, attached, fond, fond of, huggy, kind, kind-hearted,
loving, partial, soft on, sympathetic, tender, trusting, warm-hearted,
appreciative, attentive, considerate, friendly, interested in, kind, like,
respective, thoughtful, tolerant, warm toward, yielding.
14 Emotion 3- Alienated, barren, beaten, bleak, bleeding, dejected, depressed, desolate,
Depression despondent, dismal, empty, gloomy, grieved, grim, hopeless, in despair,
woeful, worried, awful, blue, crestfallen, demoralized, devalued,
discouraged, dispirited, distressed, downcast, downhearted, fed up, lost,
melancholy, miserable, regretful, rotten, sorrowful, tearful, upset, weepy,
blah, disappointed, down, funk, glum, low, moody, morose, sombre,
subdued, uncomfortable, unhappy
14 Emotion 4 - Blemished, blotched, broken, crippled, damaged, false, feeble, finished,
Inadequateness flawed, helpless, impotent, inferior, invalid, powerless, useless, washed up,
whipped, worthless, zero, defeated, deficient, dopey, feeble, helpless,
impaired, imperfect, incapable, incompetent, incomplete, ineffective, inept,
insignificant, meagre, puny, tenuous, tiny, uncertain, unconvincing, unsure,
weak, wishful, lacking, lame, overwhelmed, small, substandard,
unimportant
14 Emotion 5 – Fear Alarmed, appalled, desperate, distressed, frightened, horrified, intimidated,
panicky, paralysed, petrified, shocked, terrified, terror-stricken, wrecked,
afraid, apprehensive, awkward, defensive, fearful, fidgety, fretful, jumpy,
nervous, scared, shy, skittish, spineless, taut, threatened, troubled, wired,
anxious, careful, cautious, disquieted, goose-bumpy, shy, tense, timid,
uneasy, unsure, watchful, worried.
14 Emotion 6 – Baffled, befuddled, chaotic, confounded, confused, dizzy, flustered, rattled,
Confusion reeling, shocked, shook up, speechless, startled, stumped, stunned, taken-
aback, thrown, thunderstruck, adrift, ambivalent, bewildered, puzzled,
blurred, disconcerted, disordered, disorganised, disquieted, disturbed,
foggy, frustrated, misled, mistaken, misunderstood, mixed up, perplexed,
troubled, distracted, uncertain, uncomfortable, undecided, unsettled, unsure
14 Emotion 7 – Hurt Abused, aching, anguished, crushed, degraded, destroyed, devastated,
discarded, disgraced, forsaken, humiliated, mocked, punished, rejected,
ridiculed, ruined, scorned, stabbed, tortured, annoyed, belittled, cheapened,
criticised, damaged, depreciated, devalued, discredited, distressed, impaired,
injured, maligned, marred, miffed, mistreated, resentful, troubled, used,
wounded, let down, minimised, neglected, put away, put down, rueful,
tender, unhappy
14 Emotion 8 – Affronted, belligerent, bitter, burned up, enraged, fuming, furious, heated,
Anger incensed, infuriated, intense, outraged, provoked, seething, storming,
truculent, vengeful, vindictive, wild, aggravated, annoyed, antagonistic,
crabby, cranky, exasperated, fuming, grouchy, hostile, ill-tempered,
indignant, irate, irritated, offended, ratty, resentful, sore, spiteful, testy,
ticked off, bugged, chagrined, dismayed, galled, grim, impatient, irked,
petulant, resentful, sullen, uptight.

111 | P a g e
Chapter 5: Research Design

14 Emotion 9 – Abandoned, black, cut off, deserted, destroyed, empty, forsaken, isolated,
Loneliness marooned, neglected, ostracised, outcast, rejected, shunned, alienated,
alone, apart, cheerless, companionless, dejected, despondent, estranged,
excluded, left out, leftover, lonely, oppressed, uncherished, blue, detached,
discouraged, distant, insulated, melancholy, remote, separate, withdrawn
14 Emotion 10 - Abashed, debased, degraded, delinquent, depraved, disgraced, evil, exposed,
Remorse humiliated, judged, mortified, shamed, sinful, wicked, wrong, ashamed,
contrite, culpable, demeaned, downhearted, flustered, guilty, penitent,
regretful, remorseful, repentant, shamefaced, sorrowful, sorry, blushing,
chagrined, chastened, crestfallen, embarrassed, hesitant, humble, meek,
regretful, reluctant, sheepish.
15 Family [brother], [sister], [daughter], [cousin], [grandfather], [grandpa], [pop], [pa],
[nan], [grandmother], [grandma], [mum], [mother], [generation], [father],
[dad], [papa], [family].

Social media engagement behaviour operationalisation

The dependent variable of the study, SMEB merges both positively- and negatively-

valenced expressions of engagement, in addition to considering the intensity of

engagement. The review of the literature on customer engagement behaviour, internet

user typologies and social media user categorisations (See Chapter 2) provides a useful

foundation for organising the framework and its corresponding operationalisation of

measures. The construct of SMEB consists of three positively-valenced, one neutral and

two negatively-valenced behaviours as demonstrated in Table 5.10.

Table 5.10 Social Media Engagement Behaviour Operationalisation

SMEB Valence Definition Dictionary / Measure


Creating +ve Users engage with brands and other Number of comments made on the
users by creating positively-valenced post
content on social media platforms.
Contributing +ve Users contribute to existing content in Number of Post ‘Likes’
social media platforms. Contributing Number of post ‘Shares’
users exhibit a moderate level of
positively-valenced SMEB.

Consuming +ve Users passively consume content Total number of clicks


without any form of active Clicks to play video
reciprocation or contribution.
Link clicks
Consuming users demonstrate a
minimum level of positive, passive Other clicks
SMEB. Photo views
Dormancy Neutral A temporary state of inactive, passive Dormancy = (Post Reach –

112 | P a g e
Chapter 5: Research Design
engagement by users who may have Engaged users)
previously interacted with the focal Post reach: Post reach is the
brand. number of people who have seen
the post. The post counts as
reaching someone when it's shown
in News Feed. Figures are for the
first 28 days after a post was
created and include people viewing
the post on desktop and mobile.
Engaged users: total number of
customers who commented, liked,
shared, clicked, detached (hide
post, hide all post, unlike page) and
destructed (report as spam).
Detaching -ve Users take action to remove content of Negative Feedback: Actions users
the brand appearing in their news-feed can take to tell Facebook that they
or equivalent home page. Detaching do not want to see certain content.
users exhibit a moderate level of Measures: Hide post, hide all posts,
negatively-valenced SMEB. unlike page
Destructing -ve Negative, active contributions to Negative Feedback: Actions users
existing content on social media can take to tell Facebook that they
platforms are created by destructive do not want to see certain content
users. Destructive users represent a (report as spam).
highest level of negatively-active
SMEB.

Media richness operationalisation

Within social media content, specifically delivered by brands on Facebook, media

richness is categorised into three levels as shown in Table 5.11. This operationalisation

is derived from De Vries et al. (2012) and Cvijikj and Michahelles (2013): 1) Low

media richness for status updates as they are in the form of written text; 2) medium

richness for photos and images as they include pictorial content; 3) high vividness for

videos as they offer sound and pictorial content.

Table 5.11 Media Richness Operationalisation

Media Richness Operationalisation


Level
Low Status updates (text only)
Medium Photos and images (imagery, no audio)
High Videos (text, imagery and audio)

113 | P a g e
Chapter 5: Research Design
Congruity operationalisation

Social media content varies in its degree of congruity with the focal brand (Shamdasani,

Stanaland, & Tan, 2001). Three levels of congruity are used in this study to determine

the relevancy of the content, referring to the degree to which a piece of information

contributed to the identification of the primary message communicated by the brand.

The levels used in this study are presented in Table 5.12.

Table 5.12 Congruity Operationalisation

Congruity Operationalisation
Low Post is not related to the category, the brand or a product.
Medium Post relates in a general sense to the category (wine)
High Post relates to the specific brand or product of the brand

Community size operationalisation

Community size was categorised as small (less than 1,500 fans), medium (1500-10,000

fans) and large (over 10,000 fans). The categorisation of community size was derived

from the Mastermind (2015) report containing average community size statistics of

Australian and New Zealand brand Facebook pages.

Table 5.13 Community Size Operationalisation

Congruity Operationalisation
Small Less than 1,500 fans
Medium 1,500 to 10,000 fans
Large Over 10,000 fans

Step 4a. Human Coding Schemes

Human coding was required in the study in order to code non-textual data in which the

custom dictionaries could not be applied. This non-textual content includes photos and

videos relevant to social media content. NVivo 10 was used to complete coding of

content according to the codes in Table 5.14.

114 | P a g e
Chapter 5: Research Design
Table 5.14 Image Coding Scheme

Image Codes Image content present


Informational Content
Product Image contains a picture of the product: wine bottle, wine label, glass of wine
Vineyard Image contains a picture of the vineyard
Winery Image contains a picture of the winery: winery facilities, production
Venue Image contains a picture regarding the venue and facilities available
Review/Award Image contains a picture of a review or award: medal, wine review screenshot
or newspaper/magazine clipping, trophy
Entertaining Content
Interesting/fun Image contains an interesting artefact, relates to the history of the brand or
fact/historic image provides a fun fact
Scenic Image is a scenic photo
Occasion image Image includes customers or staff at an event, special occasion or party
hosted by the brand
Food and produce Image includes pictures of food, produce and recipes used by the brand
image
Meme Image or picture, typically humorous in nature, often in cartoon or pictorial
form
Animal Image contains a picture of an animal or pet
Remunerative Content
Competition image Image contains details and instructions about a competition/contest and/or
prize
Sale/Promotion image Image contains details about a sale, discount, promotion or special price.
Relational Content
Child/baby image Image contains a picture of a child or baby
Inspirational/ Image contains an inspirational or motivational quote, wordplay or text
motivational quote
Customer image Image contains a single customer or group of customers
Employee image Image contains a single employee or group of employees
Community Image contains a reference to community involvement through local events,
involvement image charities and causes

Step 4b. Computer Coding Schemes

In recent years, the advance in computer technology has allowed developments in

quantitative content analysis techniques, eliminating coding errors and enabling the

analysis of large volumes of written communication. Through the creation of computer

content coding schemes, coding rules may be formalised, permitting perfect coding

reliability to be obtained (Weber, 1990).

115 | P a g e
Chapter 5: Research Design
Computer software assisted in the QCA process of this research in two ways. Firstly, it

assisted in the sorting, analysing and reporting of research data including the coding and

notations made by the researchers and the construction of tables, charts and graphs.

Secondly, computer software was used for the automatic scanning of texts and

identification of words and phrases. Notwithstanding the thematic analysis of

community member comments and images embedded within the social media content,

automation of the entire process of coding and analysis was achieved.

In consideration of the nature of the sample, the research questions of the study and the

conditions as described by Holsti (1969), a predominantly computerised approach to

content analysis was required for this study. The electronic archive of the data necessary

for this study provides advanced searching capabilities, minimising the time required

for routine counting, increasing levels or reliability and creating greater degrees of

reproducibility which should result in a reduction of researcher bias (Neuendorf, 2002).

The custom dictionaries developed for computer coding have been presented in the

preceding sections. The process of computer coding post content can be conducted

using the SEARCH IF function in Microsoft Excel. This allows each row of text (social

media content) to be analysed for key words or phrases as specified within the custom

dictionary. The output then gives results on the occurrence (1) or non-occurrence (0) of

that term within the row (post), resulting in binary coding of all independent variable

codes. For example, the Excel formula was developed using the custom dictionary

developed for the code ‘product variety’ within informational posts:

=IF(ISNUMBER(SEARCH("Chardonnay",$D4)),"1",IF(ISNUMBER(SEARCH("PinotGrigio",$D4)),"1"
,IF(ISNUMBER(SEARCH("Riesling",$D4)),"1",IF(ISNUMBER(SEARCH("SauvignonBlanc",$D4)),"1"
,IF(ISNUMBER(SEARCH("Viogner",$D4)),"1",IF(ISNUMBER(SEARCH("CheninBlanc",$D4)),"1",IF
(ISNUMBER(SEARCH("Gewurtztraminer",$D4)),"1",IF(ISNUMBER(SEARCH("Semillon",$D4)),"1",I
F(ISNUMBER(SEARCH("Verdehlo",$D4)),"1",IF(ISNUMBER(SEARCH("CabernetSauvignon",$D4)),

116 | P a g e
Chapter 5: Research Design
"1",IF(ISNUMBER(SEARCH("PinotNoir",$D4)),"1",IF(ISNUMBER(SEARCH("Tempranillo",$D4)),"1
",IF(ISNUMBER(SEARCH("Carmanere",$D4)),"1",IF(ISNUMBER(SEARCH("Durif",$D4)),"1",IF(IS
NUMBER(SEARCH("Grenache",$D4)),"1",IF(ISNUMBER(SEARCH("Shiraz",$D4)),"1",IF(ISNUMB
ER(SEARCH("Sangiovese",$D4)),"1",IF(ISNUMBER(SEARCH("Zinfandel",$D4)),"1",IF(ISNUMBER
(SEARCH("Mouvedre",$D4)),"1","0"))))))))))))))))))).

To illustrate further, the following formula was developed using the custom dictionary

developed for the code ‘holiday/event’ within relational posts:

=IF(ISNUMBER(SEARCH("birthday",$D4)),"1",IF(ISNUMBER(SEARCH("christmas",$D4)),"1",IF(IS
NUMBER(SEARCH("easter",$D4)),"1",IF(ISNUMBER(SEARCH("boxingday",$D4)),"1",IF(ISNUMB
ER(SEARCH("newyear",$D4)),"1",IF(ISNUMBER(SEARCH("australiaday",$D4)),"1",IF(ISNUMBER(
SEARCH("goodfriday",$D4)),"1",IF(ISNUMBER(SEARCH("anzacday",$D4)),"1",IF(ISNUMBER(SE
ARCH("queen'sbirthday",$D4)),"1",IF(ISNUMBER(SEARCH("labourday",$D4)),"1",IF(ISNUMBER(S
EARCH("holiday",$D4)),"1",IF(ISNUMBER(SEARCH("publicholiday",$D4)),"1",IF(ISNUMBER(SEA
RCH("melbournecup",$D4)),"1",IF(ISNUMBER(SEARCH("aprilfool",$D4)),"1",IF(ISNUMBER(SEAR
CH("father'sday",$D4)),"1",IF(ISNUMBER(SEARCH("mother'sday",$D4)),"1",IF(ISNUMBER(SEARC
H("monday",$D4)),"1",IF(ISNUMBER(SEARCH("tuesday",$D4)),"1",IF(ISNUMBER(SEARCH("wedn
esday",$D4)),"1",IF(ISNUMBER(SEARCH("thursday",$D4)),"1",IF(ISNUMBER(SEARCH("friday",$
D4)),"1",IF(ISNUMBER(SEARCH("saturday",$D4)),"1",IF(ISNUMBER(SEARCH("sunday",$D4)),"1",
"0"))))))))))))))))))))))).

In total, 55 formulas were written in order to apply the custom dictionaries to each

variable. The formulas are available in Appendix C.

Step 5. Sampling

Step five of the content analysis process involves determining the selection of content to

analyse. For this analysis, social media content and corresponding SMEB data was

extracted from 12 Australian wine brands. A total of 2,236 Facebook posts were

analysed. Convenience sampling was used in this study as a result of the unavoidable

difficulty in obtaining social media message content and SMEB from a random sample

of sources. The convenience sample was limited to social media and SMEB data from

wine brands which the researcher was granted access to. In order to collect the data, the

researcher sent a preliminary email to Australian wine brands with active Facebook

accounts. Twelve brands responded and confirmed their participation in the study.

Though there are limits to the reliability of convenience sampling (Fricker & Schonlau,

2002), a long term goal of this research is to build a framework and method within

117 | P a g e
Chapter 5: Research Design
which additional social media content and behavioural data can be continually added.

The large size of the data set obtained from the convenience sample, in addition to the

cross-section of brands including large corporations, family-owned, regional, and

boutique brands was expected to dissipate such limitations.

To limit the scope of the project, only Australian wine brands were chosen for analysis.

However, there are numerous wine brands located internationally with a heavy social

media presence. This study was limited to Australian wine brands with social media

pages registered within Australia. The social media content selected for the sample

(number of posts=2,236) was created after January 1st, 2013 and prior to December 31st,

2013. This 12 month window of social media content allowed for further narrowing of

the project scope, whilst maintaining the ability to observe and account for possible

seasonal effects across a one year time frame.

Step 6. Training and Initial Reliability

Training and initial reliability is required for human coding of social media content. A

rigorous, scientific approach to media content analysis in order to gain maximum

reliability requires the use of two or more coders. Two or more coders can be used to

analyse the entire content, or a sample of the content, termed the ‘sub-sample’ in order

to ensure that obtained ratings and scores are not the idiosyncratic results of one coder’s

subjective judgement (Tinsley & Weiss, 1975).

A training session was held in which the two coders, the researcher and a research

assistant worked together on a sub-set of the data. This allowed for discussion regarding

agreement and disagreement on the coding of variables. Following this, the two coders

conducted an independent coding test on a sample of 100 posts. A number of statistical

118 | P a g e
Chapter 5: Research Design
formulas have been developed in recent literature for measuring inter-coder reliability.

Cohen’s Kappa coefficient () is a statistical measure of the inter-rater (coder)

reliability which many researchers regard as more useful than percentage agreement

reliability tests (Neuendorf, 2002). NVivo10 calculates the Kappa coefficient

individually for each combination of node and source.

In NVivo, The Kappa coefficient is calculated as follows (Carletta, 1996; QSR

International, 2015):

Calculate the expected frequency by which the agreement between users could have

occurred by chance (ΣEF), by summing:

1. The number of units of the source’s content coded at the node by user A,

multiplied by the number of units coded at the node by user B, divided by the

total number of units in the source (EF1)

2. The number of units of the source’s content not coded at the node by user A,

multiplied by the number of units not coded at the node by user B, divided by

the total number of units in the source (EF2)

3. Expected frequency (EF) of the agreement occurring by chance = EF1 + EF2

Calculate the Kappa coefficient (K) as equal to:

1. Total units of agreement between the two users (TA) minus the expected

frequency (ΣEF) of the agreement occurring by chance, divided by the total

units (TU) within the source minus the expected frequency (ΣEF) of the

agreement occurring by chance: K = (TA – ΣEF) ÷ (TU – ΣEF)

119 | P a g e
Chapter 5: Research Design
In the case where both users are in complete agreement as to how the source’s content

should be coded at the node, then the value of Kappa will equal 1.

Kappa statistics range in value from -1.0 to 1.0 with results closer to 1.0 suggesting

agreement beyond chance, whereas results close to zero suggest that agreement

occurred due to chance. Landis and Koch (1977) provide a more detailed analysis of

strength of agreement defined by kappa statistics, as per Table 5.15, which continues to

be used as a means of analysing strength of agreement (McGinn, Guyatt, Cook,

Korenstein, & Meade, 2008; Viera & Garrett, 2005) and thus is utilised in this study.

Table 5.15 Kappa Value Interpretation

Kappa value Interpretation


<0 Less than chance agreement
0.20 Slight agreement
0.21 – 0.40 Fair agreement
0.41 – 0.60 Moderate agreement
0.61 – 0.80 Substantial agreement
0.81 – 0.99 Almost perfect agreement

The median kappa value for inter-coder reliability regarding coding of images was 0.77,

indicating substantial agreement among the two independent coders (results shown in

Appendix D).

Step 7. Coding

The custom dictionaries are applied to the entire content to generate per-unit

frequencies and observations for each dictionary item.

Step 8. Tabulation and reporting

The final stage of the content analysis process is tabulation, reporting and analysis of

the results identified through the content analysis. For the purpose of this study,

120 | P a g e
Chapter 5: Research Design
relationships between content analysis results and other measures within the theoretical

model such as SMEB (H1-H7) will be explored through binary logistic regression and

process modelling (Hayes, 2013) in IBM SPSS Statistics 22. The following section,

descriptive analysis, provides a detailed description of the key results identified within

the content analysis process.

5.6 Descriptive Results


5.6.1 Social Media Content

Social media content was coded with informational, entertaining, remunerative and

relational custom dictionaries. Descriptive results were generated using the analyse,

descriptive statistics function of SPSS Statistics v22.

The total number of posts coded as containing any form of informational content was

1847, 82.6% of total posts (n=2236). The total number of posts that contained a form of

entertaining content was 1135, 50.7% of total posts. The total number of posts coded as

containing remunerative content was 313, 13.9% of total posts. The total number of

posts coded as containing relational content was 1545, 69% of total posts. These results

are presented in Table 5.16. The table further explicates the results of the specific types

of informational, entertaining, remunerative and relational content coded within each

content category.

Table 5.16 Post Content Categories

Informational Content No. of Observations No. Observation as % of


total posts (n=2236)
Brand name 281 12.6
General information 61 2.7
Product image 397 17.7
Commercial partner image 2 0.1
Vineyard image 230 10.3
Winery image 282 12.6

121 | P a g e
Chapter 5: Research Design
Website 423 18.9
Price 81 3.6
Venue image 72 3.2
Review/Award image 94 4.2
Tasting and sampling 124 5.5
Product variety 276 12.3
Product details 369 16.5
Region 323 14.4
Winemaking/processing details 190 8.5
Vineyard/location 167 7.5
Opening hours 106 4.7
Year made 361 16.1
Contact details 64 2.9
Brand Fact/News 49 2.2
Service 108 4.8
Wine show, awards and reviews 160 7.2
Event 164 7.3
Product description 188 8.4

Entertaining Content No. of Observations No. Observation as % of


total posts (n=2236)
Food/recipe 276 12.3
Emoticon 1 .0
Weather 213 9.5
Humour 213 9.5
Interesting fact image 19 0.8
Scenic image 122 5.4
Event image 388 17.3
Food image 226 10.1
Celebrity image 21 .9
Meme image 67 3.0
Animal image 101 4.5
Animal 27 1.2
Slang 63 2.8
Remunerative No. of Observations No. Observation as % of
total posts (n=2236)
Deal/special offer 132 5.9
Purchase instructions 83 3.7
Competition image 53 2.4
Sale image 63 2.8
Competition details 56 2.8
Relational No. of Observations No. Observation as % of
total posts (n=2236)
Ask question 360 16.1
Congrats/thanks 101 4.5
Quiz/game 24 1.1
Holiday, event, special day 324 14.5
Affection 9 0.4

122 | P a g e
Chapter 5: Research Design
Ask for action 1 0.0
Child/baby image 15 0.7
Inspirational/motivational quote 5 0.2
Customer image 1 0.0
Employee image 7 0.3
Community involvement image 54 2.4
Friends and fans 55 2.5
Human name 236 10.6
Emotion 480 24.3
Family 634 23.9

5.6.2 Social Media Engagement Behaviour

Table 5.17 provides descriptive statistics for SMEBs: creating, contributing, consuming,

dormancy, detaching and destructing. It can be observed that the average number of

comments made on a post is 2.22 comments. The table also shows that the average

number of likes received on a post is 21.68. The average number of times a post is

shares is 1.58 times. The table also shows the four types of consuming behaviour

exhibited by users. The mean for consuming behaviour in the form of ‘other clicks’ is

highest, with an average of 26.76 ‘other clicks’ made on posts. The average dormancy

rate is high, at 90%. The mean scores for detaching behaviour are quite low, indicating

that this behaviour occurs less frequently compared to the positively-valenced SMEBs.

Similarly, the number of cases in which destructing behaviour occurred was very low at

just 8 cases within the total of 2,236 posts. For this reason, the destructing behaviour

component of SMEB did not meet the requirements for the dependent variable case size

required within binary logistic regression (Harrell, 2013). Consequently, the binary

logistic regression results presented in Chapter 6 do not include testing of the effect

social media content on destructing SMEB.

123 | P a g e
Chapter 5: Research Design
Table 5.17 Descriptive Statistics for Social Media Engagement Behaviour
SMEB: Creating Contributing Consuming Dormancy Detachment Destru
cting

MEASURES: Number Likes Total Clicks Link Other Photo Dormancy Hide Hide 1 Unlike X Report
of shares to Play Click Click View (%) all post page Button as
Comments posts spam

Mean 2.22 21.68 1.58 0.13 1.68 26.76 17.03 0.90 0.16 0.04 0.05 0.24 0.01

Std. Error of Mean 0.13 0.86 0.12 0.03 0.12 0.88 0.65 0.00 0.01 0.00 0.01 0.01 0.00

Median 0.00 10.00 0.00 0.00 0.00 13.00 7.00 0.91 0.00 0.00 0.00 0.00 0.00

Mode 0.00 0.00 0.00 0.00 0.00 3.00 0.00 0.92 0.00 0.00 0.00 0.00 0.00

Std. Deviation 6.05 40.60 5.61 1.48 5.63 41.40 30.92 0.08 0.55 0.20 0.24 0.66 0.09

Variance 36.63 1648.53 31.47 2.19 31.70 1713.63 956.08 0.01 0.30 0.04 0.06 0.43 0.01

Minimum 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.00 0.00 0.00 0.00 0.00

Maximum 121.00 629.00 105.00 36.00 76.00 491.00 473.00 1.00 6.00 2.00 3.00 8.00 2.00

Sum 4960.00 48467.00 3532.00 294.00 3759.00 59831.00 38074.00 2010.53 362.00 81.00 109.00 531.00 8.00

124 | P a g e
Chapter 5: Research Design

5.6.3 Moderating Variables

Richness

Social media content was categorised according to post type in order to create the media

richness construct. The post types were (1) status, link, (2) photo, (3) video. Distribution

of each of these types over the dataset of 2,236 social media posts is presented in Table

5.18.

Table 5.18 Media Richness


Richness level Post type Frequency Percent
Low Status/link 687 30.7
Medium Photo 1500 67.1
High Video 49 2.2
Total 2236 100.0

It can be seen that 67.1% of the posts (1,500 posts) include a photo, followed by the

status and link media type (30.7%, 687). A total of 49 videos were included in the posts

(2.2%).

Congruity

Social media content with low congruity is categorised as content which is not

specifically related to the category (wine), brand or product. Content with medium

congruity is relevant to the product category but does not focus on a specific brand or

product. Social media content with high congruity includes details of the brand and a

specific product/s. This content is highly contextually relevant to the social media brand

page. The congruity variable therefore has three categories; low, medium and high.

Distribution of each category of congruity is presented in Table 5.19.

Table 5.19 Congruity


Congruity Level Frequency Percent
Low 960 43
Medium 433 19.4

125 | P a g e
Chapter 5: Research Design
High 843 37.7
Total 2236 100.0

Community Size

This study included community size as a moderating variable, measured by the number

of fans on the brand page. Community size was categorised as small (less than 1,500

fans), medium (1500-10,000 fans), and large (over 10,000 fans). A total of 525 (23.5%)

of posts were made within small community sizes. The majority of posts (1306 posts,

58.4%) were made in medium community sizes. 405 (18.1%) posts were made in large

community sizes.

Table 5.20 Community Size

Community Size No. Fans No. Posts Percent


Small < 1,500 525 23.5
Medium 1501 – 10,000 1306 58.4
Large >10,000 405 18.1
Total 2236 100

5.6.4 Control Variables


The testing of the relationship between social media content and SMEB controls for the

effect of three variables related to social media content scheduling. These variables are:

the day of the week in which the post is made, the month of the year in which the post is

made, and finally, the hour of the day in which the post is made, as discussed in the

following sections.

Day of Post

Previous studies of temporal interaction patterns have shown that most of the user

activities on Facebook are undertaken during workdays (Golder et al., 2007). Further,

studies have shown that click through rates of online advertisements decrease

significantly on weekend days, and people perform less internet searching during

weekends than on weekdays (Rutz & Bucklin, 2011). Within Facebook, brand fans may
126 | P a g e
Chapter 5: Research Design
visit brand fan pages more during the weekends than on weekdays, or vice versa.

Hence, this study takes into account the day of the week in which the brand post is

delivered to the fans.

It can be seen that the lowest level (4.8%) of activity in terms of number of shared posts

over a seven day period occurred on Saturdays with 107 occurrences, while the highest

number of posts were shared on Thursdays with 476 occurrences (21.2%). Table 5.21

contains the distribution of posts by day of the week.

Table 5.21 Post Distribution by Week


Post Day No. Posts Percent of posts
Monday 408 18.2
Tuesday 357 16.0
Wednesday 341 15.2
Thursday 472 21.1
Friday 162 7.2
Saturday 107 4.8
Sunday 389 17.4
Total 2236 100

The distribution of posts by month is presented in Table 5.22. The number of posts per

month remains consistent throughout the calender year with a slight decrease in January

(6.7%) and an increase in March (9.9%) and April (9.8%).

Table 5.22 Post Distribution by 12 Months


Month Frequency Percent
January 150 6.7
February 172 7.7
March 221 9.9
April 219 9.8
May 194 8.7
June 201 9.0
July 196 8.8
August 167 7.5
September 203 9.1
October 156 7.0
November 194 8.7
December 163 7.3
Total 2236 100

127 | P a g e
Chapter 5: Research Design
The distribution of posts by time is also included as a control variable in the study.

Social media users have been found to engage least during the morning and early

afternoon, with increased interaction toward the evening, reaching a steady, high level

during the night (Golder et al., 2007). Hence, if a post is created in the period when

Facebook fans are more active, there is a greater possibility for the brand post to be seen

on the wall, resulting in greater potential engagement. Based on this reasoning, this

study also controls for the effect of posting time.

Table 5.23 shows that the lowest level of activity in terms of the number of posts shared

across the day occurs at 10am (.0%, 1 occurrence). The highest number of posts were

shared between 7pm and 8pm.

Table 5.23 Post Distribution by Hour


Hour (o’clock) Frequency Percent
12 am 118 5.3
1 am 74 3.3
2 am 74 3.3
3 am 40 1.8
4 am 24 1.1
5 am 19 .8
6 am 6 .3
7 am 4 .2
9 am 2 .1
10 am 1 .0
11 am 3 .1
12 pm 17 .8
1 pm 16 .7
2 pm 61 2.7
3 pm 108 4.8
4 pm 154 6.9
5 pm 251 11.2
6 pm 228 10.2
7 pm 257 11.5
8 pm 189 8.4
9 pm 218 9.7
10 pm 202 9.0
11pm 170 7.6
Total 2236 100.0

128 | P a g e
Chapter 5: Research Design

5.7 Hypothesis Testing


Completion of the quantitative content analysis and resultant descriptive analysis as

outlined in this chapter allowed the data to be prepared for hypothesis testing.

5.7.1 Binary Logistic Regression

Hypotheses 1 to 5 were analysed using binary logistic regression using IBM SPSS

Statistics (Version 20).

Logistic regression is an appropriate statistical technique when the dependent variable is

a categorical (nominal or nonmetric) variable, and the independent variables are metric

or nonmetric variables (Hair, Black, Babin, Anderson, & Tatham, 2006). Logistic

regression is a preferred statistical technique in this study as assumptions such as

multivariate normality and equal variance-covariance matrices across groups were not

met. Logistic regression does not face these strict assumptions and is more robust when

these assumptions are not met (Hair et al., 2006).

Assumptions of logistic regression

Logistic regression does not assume a linear relationship between the dependent and

independent variables. The dependent variable must be a dichotomy (2 categories). The

independent variables are not required to be interval, normally distributed, linearly

related, or of equal variance within each group. The categories must be mutually

exclusive and exhaustive; for example a case can only be in one group and every case

must be a member of one of the groups. Finally, larger samples are needed than for

linear regression because maximum likelihood coefficients are large sample estimates.

A minimum of 50 cases per predictor is recommended (Hosmer Jr & Lemeshow, 2004)

Variable coding for logistic regression

129 | P a g e
Chapter 5: Research Design
The measurement of social media content presence for H1 to H4 was a dichotomy

(1=content present within post, 0 = content not present within post). The level of social

media content used for H1a to H4a was a categorical independent variable. The

dependent variables of the study (SMEB) were coded as a dichotomy in order to predict

the likelihood of the behaviour occurring (1 = behaviour occurred e.g. ‘comment made’,

0 = behaviour did not occur). A summary of the coding of the independent variables can

be found in Table 5.24.

Table 5.24 Independent Variable Coding

Predictor Coded as
Informational content 1 for content present, 0 for content absent
Entertaining content 1 for content present, 0 for content absent
Remunerative content 1 for content present, 0 for content absent
Relational content 1 for content present, 0 for content absent
Informational content level A number between 1 and 24
Entertaining content level A number between 1 and 13
Remunerative content level A number between 1 and 5
Relational content level A number between 1 and 15

A summary of the coding for the dependent variables can be found in Table 5.25.

Table 5.25 Dependent Variable Coding

SMEB Variable Coded as


Creating 1 for comment occurred, 0 for no comment
Contributing (shares) 1 for share occurred, 0 for no share
Contributing (likes) 1 for like occurred, 0 for no like
Consuming 1 for consuming behaviour, 0 for no consuming
Dormancy 1 for high dormancy (>90%) 0 for low dormancy (<90%)
Detaching 1 for detaching occurred, 0 for no detaching
Destruction 1 for destruction occurred, 0 for no destruction

A summary of the coding for the control variables can be found in Table 5.26.

Table 5.26 Control Variable Coding

Variable Coded as
Time of Day 1 for AM, 0 for PM

130 | P a g e
Chapter 5: Research Design

Post Month 1 for January, 2 for February, 3 for March, 4 for April, 5 for
May, 6 for June, 6 for July, 8 for August, 9 for September, 10
for October, 11 for November, 12 for December
Post Day 1 for Monday, 2 for Tuesday, 3 for Wednesday, 4 for
Thursday, 5 for Friday, 6 for Saturday, 7 for Sunday

5.7.2 Process Analysis

To test H6, H7 and H8 concerning moderation effects, Process Model 2 was applied.

Hayes (2013) mediation, moderation and conditional process analysis is a regression

based approach to statistical testing. This was conducted through ‘PROCESS’, a

computational tool for path analysis-based moderation and mediation analysis, in

addition to their integration in the form of a conditional process model (Hayes, 2013).

Hayes ‘PROCESS’ tool allows for estimation of unstandardised model coefficients,

standard errors, t and p-values and confidence intervals using OLS regression and/or

maximum likelihood logistic regression.

The hypotheses developed in Chapter 4 demonstrate a conceptual model in which three

variables (richness, congruity and community size) are estimated as moderating a single

focal predictor’s (social media content) effect. PROCESS (Hayes, 2013) has the ability

to estimate outcomes in this type of model, implementing the necessary computations

for probing the interaction and visualising more complex models (Hayes, 2013).

Statistical analysis was carried out using IBM SPSS Statistics 20 using the PROCESS

macro (Hayes, 2013, Model 2). The program was written by Andrew Hayes (2013). A

variety of models were examined separately, with social media content as predictors,

measures of congruity, richness and community size as moderators, and SMEB as the

outcome variable.

The statistical model used is presented in Figure 5.2 and Figure 5.3. The independent

131 | P a g e
Chapter 5: Research Design
variables are informational content, entertaining content, remunerative content and

relational content, the moderator variables are media richness, congruity, and

community size, and the dependent variables are SMEB typologies of creating,

contributing, consuming, dormancy and detaching.

Figure 5.2 PROCESS Model 2 Conceptual Diagram

Figure 5.3 PROCESS Model 2 Statistical Diagram

PROCESS moderation model with a three category moderator


In Figure 5.2, ‘X’ refers to the social media content as an independent variable. This

content can be informational, entertaining, remunerative or relational. ‘Y’ is the

outcome variable which is represents one of the six SMEBs; creating, contributing,

consuming, dormancy and detaching. ‘M’ and ‘W’ are the moderator variables. This

132 | P a g e
Chapter 5: Research Design
model shows ‘M’ and ‘W’ as two distinct moderating variables. However, by adopting

Hayes (2015) approach to adjusting this model, it was possible to convert M and W to

represent two categories of a single moderator, e.g. ‘low’ and ‘high’.

In order to construct this model, indicator coding of the moderator variables was

required. When k = 3 (three categories of the moderating variable), D1 is coded as 1 for

all cases in group 1, and 0 for all other cases. Similarly, D2 is coded as 1 for cases in

group 2, and 0 in all cases. The third group is identified when both D1 and D2 are coded

as 0. When M is a multi-categorical variable, in order to assess its role as a moderator of

the effect of X on Y, M is subsequently converted into two dichotomous variables (M

and W). To illustrate, in the case of media richness as the moderating variable, D1 refers

to low media richness D2 refers to high media richness. In the case that D1 and D2 were

both coded as ‘0’, the media richness would be medium. The full guidelines for

conducting a moderation analysis with PROCESS Model 2 and three category

moderator variables are outlined in Hayes (2015).

The results of the conditional process analysis and consequent hypothesis testing are

presented in Chapter 6.

5.8 Chapter Summary


Taking on a positivist perspective, a quantitative research design was adopted,

comprised of two phases deemed appropriate in order to test the model and hypotheses

developed in chapter three. The design and implementation of the quantitative content

analysis was described, including key considerations of the theory and rationale for the

study, conceptualisation decisions, operationalisation of measures, development of

coding schemes, sampling, training and initial reliability testing, coding, tabulation and

133 | P a g e
Chapter 5: Research Design
reporting. The descriptive characteristics and results of the quantitative content analysis

were presented prior to hypothesis testing. In order to test the hypotheses set out in the

previous chapter, binary logistic regression and Hayes (2013) process modelling was

applied. Chapter Six will outline the results of the study.

134 | P a g e
Chapter 6: Results

CHAPTER 6. Results
6.1 Introduction
The chapter begins with an examination of the effect of social media content presence

on SMEB through the application of binary logistic regression. It then reports on the

effect of social media content levels on SMEB, again using binary logistic regression

procedures. The interaction effects between informational, entertaining, relational and

remunerative social media content types on SMEB are also tested. A series of adapted

PROCESS Model 2 analyses were conducted to examine the moderating effects of

media richness, congruity and community size. These effects are subsequently

discussed. The chapter concludes with an overview of the hypotheses and summary of

results.

6.2 Social Media Content


Binary logistic regression is an appropriate statistical technique for examining the

impact of the presence of each type (informational, entertaining, remunerative and

relational) of social media content on specific SMEB’s (creating, contributing,

consuming, dormancy, detaching and destructing) independently.

Each model controls for the effect of post time (AM or PM), post day of the week, and

post month of the year. When conducting binary logistic regression with categorical

independent variables such as social media content, day of the week, month of the year

and time of the day, it is necessary to specify which predictor variables are categorical

within SPSS (Field, 2013).

The control variables, day of the week, month of the year and time of the day are

categorical variables. These variables were transformed into dummy variables in order

135 | P a g e
Chapter 6: Results
to represent groups using only zeros and ones. When creating dummy variables, the

number of variables required is one less that the number available. One variable is

selected as the baseline against which all other groups will be compared. For post day of

the week, Wednesday was selected as the baseline against which all other groups are

compared. Field (2013) suggests that when no obvious baseline group is identifiable,

researchers can select the most interesting baseline group. Consequently, the effects of

the day of the week were tested multiple times, each time selecting a different day as the

baseline. The results for Wednesday as the baseline category provided the most

interesting insight and are hence reported within this thesis. The same method was

adopted for the selection of the month ‘June’ as the baseline category for the month of

the year dummy variables. Post time of the day was also dummy coded into a

dichotomous (AM/PM) variable. ‘PM’ was selected as the baseline category, and hence

all posts made in the morning were coded 1, and posts made in the afternoon were

coded 0. The results for AM are therefore comparative to the baseline category 0, which

represents PM, and are presented in each table of results.

The results for Hypotheses 1 to 4 are presented in the following sections, with

statistically significant predictors presented in bold.

6.2.1 Informational Content Presence

Informational content may contain details on price, availability, location and product

names (Lee et al., 2013). Further, informational content may contain explanatory

images referring to the brand’s location, facilities and products. Information may also

relate to the brand’s contact details such as the provision of contact phone numbers,

email addresses, links to a website, and opening hours. A full description of the

informational content operationalisation is provided in Chapter 5, Section 5.5.3.


136 | P a g e
Chapter 6: Results
Drawing from empirical studies exploring the link between informational content and

consumer response, it was hypothesised that informational social media content will

facilitate positively-valenced SMEB (H1) as discussed in Chapter 4. Customers who

seek informational value from social media content are more likely to consume content

rather than interact through comments and discussion (Ko et al., 2005). As such,

informational content is further surmised to have a significant relationship with

consuming, passive SMEB;

H1. The presence of informational content facilitates passive, positively-valenced social

media engagement behaviour.

In order to test this hypothesis, binary logistic regression was conducted, demonstrating

the effect of the presence of informational content on the likelihood that each of the six

SMEBs would occur. Tables 6.1 through to 6.7 in the following sections show the

results for informational content and creating, contributing, consuming, dormant and

detaching behaviour respectively.

Each table includes the results regarding the effect control variables post day of the

week, post month of the year, and post time of the day (AM or PM). Statistically

significant results are presented in bold.

Informational content and SMEB - creating

Creating behaviour occurs when users engage with brands and other users by creating

positively-valenced content on social media platforms in the form of making a

comment. Creating epitomises a highly active level of SMEB. Table 6.1 shows that the

presence of informational content is a statistically significant predictor of the occurrence

of the SMEB type: creating. The odds ratio of 1.41 for creating behaviour shows that

137 | P a g e
Chapter 6: Results
those posts which had informational content included were 1.4 times more likely to

receive a comment, than posts which did not have informational content.

The day of the week, month of the year and time of the post were added as control

variables in the logistic regression. For post day of the week, posts made on a Thursday

or Sunday decreased the odds of SMEB in the form of creating occurring, compared to

the baseline category of Wednesday. Thursday posts were 26% (Exp(B)=0.74) less

likely to result in creating behaviour compared to Wednesday posts, and Sunday posts

were 37% (Exp(B)=0.63) less likely to result in creating behaviour compared to a

Wednesday post. Compared to the baseline category of June, posts made in July,

August, September and October reduced the likelihood of SMEB in the form of creating

occurring.

Table 6.1 Logistic Regression Showing Effect of Informational Content Presence


on Creating Behaviour
SMEB IV B Sig Exp(B)
Creating Informational .34 .00 1.41
Behaviour Monday -.21 .15 .81
Tuesday -.18 .23 .83
Thursday -.30 .04 .74
Friday .02 .92 1.02
Saturday -.01 .96 .99
Sunday -.46 .00 .63
January -.33 .14 .72
February -.19 .38 .83
March -.49 .01 .61
April -.34 .09 .72
May -.17 .40 .84
July -.50 .01 .60
August -.44 .04 .65
September -.52 .01 .59
October -.51 .02 .60
November -.27 .19 .76
December -.24 .26 .79
AM -.06 .64 .95

138 | P a g e
Chapter 6: Results
Informational content and contributing behaviour (shares)

Contributing behaviour occurs when users contribute to existing content in social media

platforms by disseminating further pre-existing content through likes and shares.

Contributing users exhibit a moderate level of positively-valenced SMEB by ‘liking’

and ‘sharing’ brand related content. The binary logistic regression predicting the

occurrence of contributing behaviour was conducted separately for liking and sharing

contributing behaviours

Table 6.2 presents the results of the logistic regression examining the presence of

informational content and the SMEB: contributing, in the form of sharing. The presence

of informational content is a statistically significant predictor of sharing behaviour, with

an Exp(B) of 1.91 indicating that when posts have informational content, they are 1.9

times more likely to be shared.

The control variables; day of the week, time of the day and month of the year in which

the post were made are also presented in Table 6.2. Posts made on Mondays (30% less

likely), Thursdays (32% less likely) and Sundays (35% less likely) reduced the

likelihood of contributing behaviour in the form of shares occurring, compared to the

baseline category of Wednesday. The results also indicate that posts made in the

morning are 36% less likely to be shared, indicating that in order to facilitate this

behaviour, posts made after midday are preferable. The month of the year in which the

post was made did not have a statistically significant effect on contributing behaviour in

the form of shares.

139 | P a g e
Chapter 6: Results
Table 6.2 Logistic Regression Showing Effect of Informational Content Presence
on Contributing Behaviour (Shares)

SMEB IV B Sig Exp(B)


Contributing Informational .65 .00 1.91
Behaviour Monday -.35 .02 .70
(Shares) Tuesday -.11 .48 .90
Thursday -.38 .01 .68
Friday -.21 .28 .81
Saturday -.29 .21 .75
Sunday -.43 .00 .65
January .09 .70 1.09
February -.11 .61 .90
March -.18 .37 .84
April -.11 .58 .90
May -.15 .48 .86
July -.14 .49 .87
August -.24 .26 .79
September -.28 .17 .76
October -.18 .42 .84
November .05 .82 1.05
December -.18 .40 .83
AM -.45 .00 .64

Informational content and contributing behaviour (likes)

Table 6.3 shows that the presence of informational content is a statistically significant

predictor of the occurrence of contributing SMEB in the form of likes. The odds ratio of

2.46 for contributing behaviour (likes) indicates that posts which have informational

content are almost 2.5 times more likely to receive a ‘like’, compared to posts with no

informational content. This appears relatively higher than the odds ratio of 1.41 for

creating behaviour as shown in Table 6.1. Additionally, this odds ratio is higher than the

odds ratio for shares of 1.91 provided in Table 6.2. The control variables also had a

statistically significant effect on the likelihood of contributing behaviour in the form of

liking. Posts made on a Tuesday are 52% less likely to be liked, and posts made on a

Sunday are 60% less likely to be liked, compared to posts on Wednesdays. Posts made

in the morning (AM) significantly and negatively predicted the occurrence of liking.

This finding indicates that in order to facilitate contributing behaviour in the form of
140 | P a g e
Chapter 6: Results
likes, afternoon posts are preferable.

Table 6.3 Logistic Regression Showing Effect of Informational Content Presence


on Contributing Behaviour (Likes)

SMEB IV B Sig Exp(B)


Contributing Informational .90 .00 2.46
Behaviour Monday -.48 .11 .62
(Likes) Tuesday -.73 .01 .48
Thursday -.41 .16 .66
Friday .25 .58 1.28
Saturday -.06 .90 .95
Sunday -.91 .00 .40
January -.70 .11 .50
February -.73 .08 .48
March -1.07 .01 .34
April -.42 .31 .66
May .01 .98 1.01
July -.42 .32 .66
August -.59 .17 .56
September .25 .61 1.28
October -.40 .38 .67
November -1.16 .00 .31
December -.27 .55 .76
AM -.41 .05 .66

Informational content and consuming behaviour

Consuming behaviour is defined as the passive consumption of brand related content

through reading reviews, discussion and comments, viewing photos, watching videos

and clicking on content and links. Consuming reflects the minimum level of positively-

valenced SMEB. The behaviour is passive, denoting a level of participation without

active contribution to or creation of content. Table 6.4 shows the effect of informational

content presence on consuming SMEB. Informational content is a statistically

significant predictor of consuming behaviour (Exp(B) =2.27). Posts which contain

informational content are 2.27 times more likely to result in consuming behaviour,

compared to posts with no informational content. The control variables had no

significant effect on the occurrence of consuming SMEB.

141 | P a g e
Chapter 6: Results
Table 6.4 Logistic Regression Showing Effect of Informational Content Presence
on Consuming Behaviour

SMEB IV B Sig Exp(B)


Consuming Informational .82 .03 2.27
Behaviour Monday -1.42 .07 .24
Tuesday -.73 .39 .48
Thursday -1.04 .19 .35
Friday -.64 .53 .53
Saturday 16.25 1.00 11405526.56
Sunday -.95 .25 .39
January .50 .69 1.65
February .00 1.00 1.00
March .92 .46 2.50
April .21 .84 1.23
May .77 .53 2.17
July .13 .90 1.14
August -.45 .63 .64
September -.33 .72 .72
October -.81 .36 .44
November -1.20 .14 .30
December -1.02 .23 .36
AM -.01 .99 .99

Informational content and dormant behaviour

A dormant user is one who has made zero active or passive contributions to the

community in relation to the post. Users do not engage with the brand through

consuming, contributing to or creating content. Dormancy is measured in this study as

the post reach, minus the number of engaged users. Post reach is the number of people

who have seen the post, while engaged users refers to the total number of customers

who commented, liked, shared, clicked, detached (hide post, hide all post, unlike page)

and destructed (report as spam).

Table 6.5 indicates that informational content presence is not a statistically significant

predictor of dormant behaviour. The control variable day of the week provided some

interesting results regarding the prediction of dormant behaviour. The positively

weighted beta for Thursdays (1.41) and Sundays (1.47) indicates that dormant

behaviour is 1.41 and 1.47 times more likely to occur on these two days, compared to
142 | P a g e
Chapter 6: Results
the baseline category Wednesday. Posts made on a Saturday are 35% less likely to

result in dormant behaviour, indicating that social media users may be more active on

Saturdays, compared to Wednesdays. The likelihood of dormant behaviour occurring

increased in April (1.56 times more likely) and July (1.90 times more likely) compared

to the baseline month of June.

Table 6.5 Logistic Regression Showing Effect of Informational Content Presence


on Dormant Behaviour

SMEB IV B Sig Exp(B)


Dormant Informational -.07 .54 .93
Behaviour Monday .25 .10 1.28
Tuesday .21 .19 1.23
Thursday .34 .02 1.41
Friday .02 .91 1.02
Saturday -.43 .06 .65
Sunday .38 .01 1.47
January -.80 .00 .45
February -.11 .60 .89
March .36 .08 1.43
April .44 .03 1.56
May .12 .56 1.13
July .64 .00 1.90
August .07 .76 1.07
September .05 .82 1.05
October -.07 .74 .93
November .05 .80 1.06
December -.19 .39 .83
AM -.06 .61 .94

Informational content and detaching behaviour

Detaching behaviour reflects users who privately remove themselves from the social

media page. This is measured by the level of ‘negative feedback’ on a post, reflective of

the actions users can take to tell Facebook that they do not want to see certain

content. The measures are captured from Facebook insights data and include hide post,

hide all posts and unlike the page.

Table 6.6 shows the results of the binary logistic regression predicting the occurrence of

detaching behaviour. Informational content presence is not a significant predictor of


143 | P a g e
Chapter 6: Results
detaching behaviour. Detaching behaviour is less likely to occur in the later months of

the year, specifically September (55% less likely), October (60% less likely), November

(68% less likely) and December (74% less likely) compared to the baseline month,

June. The post time of day did not have a statistically significant impact on the

likelihood of SMEB in the form of detaching occurring.

Table 6.6 Logistic Regression Showing Effect of Informational Content Presence


on Detaching Behaviour.

SMEB IV B Sig Exp(B)


Detaching Informational .29 .06 1.33
Behaviour Monday .06 .77 1.06
Tuesday .31 .11 1.37
Thursday .19 .31 1.21
Friday .13 .61 1.14
Saturday .20 .48 1.22
Sunday .13 .50 1.14
January .18 .47 1.20
February -.31 .24 .74
March .20 .38 1.22
April .12 .60 1.13
May .04 .87 1.04
July -.09 .71 .91
August -.03 .91 .97
September -.80 .00 .45
October -.91 .00 .40
November -1.15 .00 .32
December -1.36 .00 .26
AM -.14 .36 .87

Informational content and social media engagement behaviour comparison

Table 6.7 shows the summarised results for the main effects between informational

content presence and SMEB. Informational content presence significantly and

positively predicted the occurrence of creating, contributing (likes and shares) and

consuming behaviour. There were no significant relationships between informational

content and dormant and detaching behaviour.

Therefore, H1: The presence of informational content facilitates passive, positively-

valenced social media engagement behaviour is supported.


144 | P a g e
Chapter 6: Results
The results show a significant relationship between the presence of informational

content, creating, contributing, and consuming behaviour. Contributing behaviour in the

form of likes is most likely to occur (Exp(B)=2.46) when informational content is

present.

Table 6.7 Summarised Logistic Regression Results for H1

Content SMEB b Sig Exp(B)


Information Creating .34 .00 1.41
Contributing (shares) .65 .00 1.91
Contributing (likes) .90 .00 2.46
Consuming .82 .03 2.27
Dormancy -.07 .54 .93
Detaching .29 .06 1.33

6.2.2 Informational Content Level

The summarised results in Table 6.7 show that informational content presence is a

significant predictor of creating, contributing (likes), contributing (shares) and

consuming SMEB. In this section, this effect is explored in greater detail by

investigating the level of informational content within a post, and how this may alter the

effect on occurrence of each of the behaviours.

The informational content category contains 24 items, shown in Table 5.6 (Chapter 5).

In order to test the effect of the amount of informational content, binary independent

variables were created for each content level, with 1 used to indicate content level

presence and 0 for absence. Binary logistic regression is used to assess how the amount

of information present within a post predicts the occurrence of positively-valenced

SMEB as per H1a. Each model controlled for the effects of post day, month and time of

the day, with the results mirroring those presented in the binary logistic regressions

previously conducted. The control variable effects are therefore omitted in the following

tables for ease of interpretation

145 | P a g e
Chapter 6: Results
Hypothesis 1a suggested that high levels of informational content weaken the

relationship with positively-valenced social media engagement behaviours. As receivers

of content have a limited cognitive processing capacity, when information overload

occurs it is likely that a reduction in decision quality and attention will occur (Speier et

al., 1999).

Binary logistic regression was conducted with informational content as a categorical

independent variable. The informational content level has values ranging from 1 to 11,

with 11 being the highest number of informational elements included in any post in this

study. In the following section, the results regarding informational content level and its

effect on the occurrence of creating, contributing and consuming SMEB are presented.

Table 6.8 Logistic Regression Showing Effect of Informational Content Level on


SMEB
SMEB Content Level b Sig Exp(B)
Creating behaviour Information(1) 0.23 .09 1.25
Information(2) 0.30 .03 1.35
Information(3) 0.32 .04 1.37
Information(4) 0.43 .01 1.53
Information(5) 0.65 .00 1.91
Information(6) 0.81 .01 2.25
Information(7) 1.11 .00 3.04
Information(8) 0.66 .34 1.93
Information(9) 0.64 .50 1.89
Information(10) -0.33 .79 .72
Information(11) -20.82 1.00 .00
Contributing behaviour Information(1) 0.41 0.00 1.51
(shares) Information(2) 0.45 0.00 1.57
Information(3) 0.73 0.00 2.08
Information(4) 1.18 0.00 3.25
Information(5) 1.11 0.00 3.03
Information(6) 0.92 0.00 2.51
Information(7) 1.54 0.00 4.69
Information(8) 1.08 0.12 2.94
Information(9) 0.28 0.76 1.33
Information(10) 21.92 1.00 3314822935.28
Information(11) -20.41 1.00 0.00

146 | P a g e
Chapter 6: Results
Contributing behaviour Information(1) 0.53 0.01 1.70
(likes) Information(2) 0.58 0.01 1.79
Information(3) 1.27 0.00 3.55
Information(4) 2.01 0.00 7.43
Information(5) 3.13 0.00 22.90
Information(6) 19.36 1.00 255281052.68
Information(7) 19.48 1.00 287964457.35
Information(8) 19.66 1.00 343750485.70
Information(9) -0.77 0.51 0.46
Information(10) 19.43 1.00 275572346.08
Information(11) 19.25 1.00 229651909.46
Consuming behaviour Information(1) 0.51 0.26 1.66
Information(2) 0.45 0.35 1.57
Information(3) 1.06 0.11 2.90
Information(4) 1.84 0.08 6.30
Information(5) 17.54 1.00 41314885.39
Information(6) 17.35 1.00 34252930.95
Information(7) 17.58 1.00 43154110.52
Information(8) 17.78 1.00 52473614.36
Information(9) 16.68 1.00 17606880.70
Information(10) 17.48 1.00 39180206.73
Information(11) 17.34 1.00 33952131.85

Hypothesis 1a suggested that high levels of informational content weaken the

relationship with positively-valenced social media engagement behaviours. Positive and

significant relationships were found between increasing levels of informational content

and positively-valenced SMEB in the form of creating and contributing. As the level of

informational content within a post increased from 2 to 7, the likelihood of creating

behaviour occurring also increased. No statistical significance was found to suggest that

posts with 8 or greater elements of information can facilitate the occurrence of creating

behaviour, supporting the notion of information overload. Similarly, when predicting

contributing behaviour in the form of shares, the odds of sharing behaviour increased in

line with an increase in the level of informational content, to a maximum level of 7

elements of information. As the level of informational content within a post increased

from 1 to 5, the odds of contributing behaviour in the form of likes dramatically

147 | P a g e
Chapter 6: Results
increased. For informational content levels of 6 or greater, there was no significant

effect on contributing behaviour in the form of likes.

H1a is therefore partially supported. The results in Table 6.8 show that a maximum

level of informational content is reached when predicting the occurrence of active,

positively-valenced SMEBs of creating and contributing in the form of likes.

6.2.3 Entertaining Content Presence

Entertaining content was hypothesised to facilitate positively-valenced SMEB. If a

brand post is entertaining, user motivations to engage with the content are met. Hence,

users may exhibit a more positive response toward entertaining brand posts compared to

non-entertaining brand posts. Based on this foundation, it can be hypothesised that

entertaining brand posts will facilitate active, positive SMEB;

H2. The presence of entertaining content facilitates active, positively-valenced

engagement behaviour.

The results of the binary logistic regression, predicting likelihood of SMEB based on

entertaining content, are presented in Table 6.9 through to Table 6.15. Each table

includes the results regarding the effect of control variables post day of the week, post

month of the year, and post time of the day (AM or PM). Statistically significant results

are presented in bold.

Entertaining content and creating behaviour

The results of the logistic regression in Table 6.9 indicate that the presence of

entertaining content increases the odds of creating behaviour occurring by 1.41 times,

compared to a post with no entertaining content. The control variables for post day of

the week did have an influence on whether or not creating behaviour occurred. It can be
148 | P a g e
Chapter 6: Results
seen in Table 6.9 that posts made on Thursdays and Sundays are 26% and 37%

respectively less likely to result in the occurrence of creating behaviour. This result is

consistent with the findings regarding control variables in Table 6.1, which indicated

that informational content is also less likely to facilitate creating behaviour on

Thursdays and Sundays. It can therefore be suggested that when posting informational

or entertaining content, Thursday and Sunday are not preferable days if the goal is to

facilitate creating behaviour in the form of comments by users.

Post months of March, April, July, August, September and October significantly

reduced the likelihood of creating behaviour occurring relative to June. This effect was

greatest for the post month October, in which the likelihood of creating behaviour

occurring was reduced by 43%. The post time of the day did not have a statistically

significant impact on the likelihood of creating behaviour occurring.

Table 6.9 Logistic Regression Showing Effect of Entertaining Content Presence on


Creating Behaviour
SMEB IV B Sig Exp(B)
Creating Entertaining .34 .00 1.41
Behaviour Monday -.21 .15 .81
Tuesday -.18 .23 .83
Thursday -.30 .04 .74
Friday -.03 .89 .97
Saturday -.07 .75 .93
Sunday -.47 .00 .63
January -.35 .11 .71
February -.19 .36 .82
March -.49 .01 .61
April -.39 .05 .68
May -.21 .30 .81
July -.53 .01 .59
August -.52 .02 .60
September -.55 .01 .58
October -.56 .01 .57
November -.33 .11 .72
December -.28 .20 .76
AM -.05 .65 .95

149 | P a g e
Chapter 6: Results
Entertaining content and contributing behaviour (likes)

Table 6.10 shows the results of the logistic regression predicting the odds of

contributing behaviour in the form of likes occurring. The results show that when

entertaining content is present, the odds of contributing behaviour (likes) occurring

increase by 1.24 times.

The day of the week did not have a statistically significant impact on the occurrence of

SMEB in the form of contributing (likes). Compared to the baseline category of June,

posts made in the months of October, November and December reduced the likelihood

of contributing behaviour in the form of likes occurring for posts with entertaining

content. Posts made in September are more likely to facilitate the occurrence of

contributing behaviour in the form of likes occurring (Exp(B)=.44). The time of the post

(AM) did not have a significant effect on the odds of contributing behaviour in the form

of likes occurring.

Table 6.10 Logistic Regression Showing Effect of Entertaining Content Presence


on Contributing (Likes) Behaviour
SMEB IV B Sig Exp(B)
Contributing Entertaining .54 .05 1.24
Behaviour Monday -.49 .75 1.06
(Likes) Tuesday -.74 .10 1.37
Thursday -.39 .29 1.22
Friday .19 .67 1.12
Saturday -.18 .57 1.18
Sunday -.92 .49 1.15
January -.74 .51 1.18
February -.78 .23 .73
March -1.07 .38 1.22
April -.54 .71 1.09
May -.06 .96 1.01
July -.49 .67 .90
august -.77 .74 .92
September .19 .00 .44
October -.56 .00 .39
November -1.32 .00 .30
December -.33 .00 .25
AM -.45 .36 .87

150 | P a g e
Chapter 6: Results
Entertaining content and contributing behaviour (shares)

Table 6.11 presents the results of the logistic regression concerning entertaining content

presence and contributing behaviour in the form of sharing. Entertaining content

presence is a statistically significant predictor of sharing behaviour, with an Exp(B) of

1.41 indicating that when posts have entertaining content, they are 1.41 times more

likely to be shared.

The control variables Monday, Thursday and Sunday had a statistically significant and

negative effect on the likelihood of contributing behaviour in the form of shares

occurring, compared to the baseline category of Wednesday. Posts made on Mondays,

Thursdays and Sundays are 30%, 31% and 35% respectively, less likely to be shared.

There is no indication that post on other days had statistical significance. Posts made in

the morning (AM) decrease the odds of contributing behaviour in the form of shares

occurring by 37%, compared to posts made in the afternoon.

Table 6.11 Logistic Regression Showing Effect of Entertaining Content Presence


on Contributing (Shares) Behaviour
SMEB IV B Sig Exp(B)
Contributing Entertaining .34 .00 1.41
Behaviour Monday -.35 .02 .70
(Shares) Tuesday -.11 .47 .89
Thursday -.37 .01 .69
Friday -.25 .21 .78
Saturday -.37 .11 .69
Sunday -.44 .00 .65
January .06 .80 1.06
February -.14 .52 .87
March -.19 .34 .83
April -.18 .36 .83
May -.19 .34 .82
July -.18 .39 .84
August -.35 .10 .70
September -.31 .13 .73
October -.26 .24 .77
November -.05 .81 .95
December -.22 .30 .80
AM -.46 .00 .63

151 | P a g e
Chapter 6: Results
Entertaining content and consuming behaviour

Table 6.12 shows that the presence of entertaining content was not found to be a

statistically significant predictor of consuming behaviour. Consuming behaviour is

defined as a passive positively-valenced SMEB, in which users read posts, click on

posts, view photos and view videos. Table 6.12 shows that the presence of entertaining

content does not significantly predict the occurrence of consuming behaviour. This

finding suggests that users are more likely to consume content through clicking links,

reading posts, clicking on posts, viewing photos and viewing videos when the content is

informational. Comparatively, when the content is entertaining, there is no evidence to

suggest that users are any more or less likely to consume the content.

The control variables day of the week, month of the year and time of the day in which

the post was made did not significantly affect whether or not the post resulted in the

occurrence of consuming behaviour.

Table 6.12 Logistic Regression Showing Effect of Entertaining Content Presence


on Consuming Behaviour
SMEB IV B Sig Exp(B)
Consuming Entertaining -.01 .97 .99
Behaviour Monday -1.46 .06 .23
Tuesday -.82 .33 .44
Thursday -1.03 .19 .36
Friday -.60 .56 .55
Saturday 16.17 1.00 10568580.81
Sunday -1.03 .21 .36
January .46 .71 1.58
February -.07 .95 .93
March .85 .49 2.33
April .14 .89 1.15
May .74 .55 2.10
July .06 .95 1.06
August -.57 .54 .56
September -.36 .69 .70
October -.93 .29 .39
November -1.35 .10 .26
December -1.05 .22 .35
AM -.06 .91 .94

152 | P a g e
Chapter 6: Results
Entertaining content and dormant behaviour

Table 6.13 shows the results of the logistic regression predicting the occurrence of

dormant SMEB. As the Exp(B) is statistically significant, the presence of entertaining

content reduces the odds of dormant behaviour occurring by 35%. This finding reflects

the results for active SMEBs of creating and contributing in the forms of likes and

comments, which were increased as a result of entertaining content presence. It is

therefore expected that the likelihood of dormancy occurring should be decreased.

The control variables for day of the week provide some interesting results regarding the

prediction of dormant behaviour. It can be seen that posts made on Thursdays and

Sundays are 1.43 and 1.47 times more likely to result in dormant behaviour occurring,

compared to the baseline category Wednesday.

Table 6.13 Logistic Regression Showing Effect of Entertaining Content Presence


on Dormant Behaviour
SMEB IV B Sig Exp(B)
Dormant Entertaining -.43 .00 .65
Behaviour Monday .25 .11 1.28
Tuesday .20 .20 1.22
Thursday .35 .02 1.43
Friday .09 .66 1.09
Saturday -.39 .09 .68
Sunday .38 .01 1.47
January -.79 .00 .45
February -.13 .53 .87
March .33 .11 1.40
April .49 .02 1.62
May .16 .44 1.18
July .65 .00 1.91
August .12 .57 1.13
September .07 .72 1.08
October -.05 .83 .95
November .08 .70 1.08
December -.15 .48 .86
AM -.08 .54 .93

153 | P a g e
Chapter 6: Results
Entertaining content and detaching behaviour

Table 6.14 shows the results of the logistic regression predicting the occurrence of

detaching behaviour as a consequence of the presence of entertaining content. The

presence of entertaining content increases the odds of detaching behaviour occurring by

1.24 times. The inclusion of the control variables for post month of the year show

negatively weighted and statistically significant beta scores for September, October and

November and December, compared to the baseline category of June. This finding

indicates that for posts made later in the year, the likelihood of detaching behaviour is

reduced. Posts with entertaining content present are less likely to result in detaching

behaviour occurring when they are made in September (56% less likely), October (61%

less likely), November (70% less likely), and December (75% less likely), compared to

June. The day of the post and the time of the day in which the post is made do not

significantly increase or decrease the likelihood of negatively-valenced SMEB in the

form of detaching, occurring.

Table 6.14 Logistic Regression Showing Effect of Entertaining Content Presence


on Detaching Behaviour
SMEB IV B Sig Exp(B)
Detaching Entertaining .216 .05 1.24
Behaviour Monday .061 .75 1.06
Tuesday .318 .10 1.37
Thursday .196 .29 1.22
Friday .110 .67 1.12
Saturday .164 .57 1.18
Sunday .136 .49 1.15
January .165 .51 1.18
February -.314 .23 .73
March .201 .38 1.22
April .085 .71 1.09
May .012 .96 1.01
July -.104 .67 .90
August -.083 .74 .92
September -.817 .00 .44
October -.950 .00 .39
November -1.193 .00 .30
December -1.385 .00 .25
AM -.144 .36 .87

154 | P a g e
Chapter 6: Results
Entertaining content and social media engagement behaviour comparison

The summarised binary logistic regression results for each SMEB are displayed in

Table 6.15. Entertaining content presence is a statistically significant predictor of the

occurrence of active, positively-valenced SMEBs, creating and contributing. The

statistically significant Exp(B) of .65 indicates that presence of entertaining content

within a post reduces the likelihood that dormant behaviour will occur by 35%.

Interestingly, the results also support a relationship between the presence of entertaining

content and detaching SMEB. The presence of entertaining content increases the

likelihood of users detaching from the content by 1.24 times, compared to when there

was no entertaining content in the post.

Therefore, H2 The presence of entertaining content facilitates active, positively-

valenced social media engagement behaviour is supported.

Table 6.15 Summarised Logistic Regression Results for H2


Content SMEB b Sig Exp(B)
Entertaining Creating .34 .00 1.41
Contributing (shares) .34 .00 1.41
Contributing (likes) .54 .05 1.24
Consuming -.01 .97 .99
Dormancy -.43 .00 .65
Detaching .216 .05 1.24

6.2.4 Entertaining Content Level

The entertaining content level variable had values ranging from 1 to 5, with 5 being the

highest number of entertaining elements included in a post in this study. Therefore, five

dichotomous categorical dummy variables were created, with 1 indicating the presence

of the level of content and 0 indicating the absence.

Hypothesis 2a suggested that high levels of entertaining content weaken the relationship

155 | P a g e
Chapter 6: Results
with positively-valenced social media engagement behaviours. An optimal level of

entertaining context was expected to exist, beyond which information overload would

occur and the user is presented with too many cues to successfully process (Hiltz &

Turoff, 1985)

Binary logistic regression was conducted with entertaining content levels as categorical

independent variables. The entertaining content level has values ranging from 1 to 5,

with 5 being the highest number of entertaining elements in a post identified in this

study. In the following section, results regarding entertaining content level and its effect

on the occurrence of SMEB are discussed.

Table 6.16 Logistic Regression Showing Effect of Entertaining Content Level on


SMEB

SMEB Content Level b Sig Exp(B)


Creating behaviour Entertaining(1) 0.23 0.02 1.25
Entertaining(2) 0.52 0.00 1.68
Entertaining(3) 0.68 0.01 1.98
Entertaining(4) 1.11 0.04 3.02
Entertaining(5) 21.46 1.00 2081088462.23
Contributing behaviour Entertaining(1) 0.23 0.02 1.26
(shares) Entertaining(2) 0.45 0.00 1.57
Entertaining(3) 1.06 0.00 2.88
Entertaining(4) 0.52 0.29 1.68
Entertaining(5) 21.51 1.00 2205451191.74
Contributing behaviour Entertaining(1) 0.33 0.06 1.39
(likes) Entertaining(2) 1.31 0.00 3.70
Entertaining(3) 0.52 0.33 1.69
Entertaining(4) 0.51 0.63 1.66
Entertaining(5) 19.60 1.00 326477632.82
Consuming behaviour Entertaining(1) 0.04 0.91 1.04
Entertaining(2) -0.20 0.70 0.82
Entertaining(3) -0.15 0.89 0.86
Entertaining(4) 17.08 1.00 26105017.90
Entertaining(5) 16.98 1.00 23774943.61

156 | P a g e
Chapter 6: Results
Hypothesis 2a suggested that high levels of entertaining content weaken the relationship

with positively-valenced engagement behaviour. Positive and significant relationships

were found between increasing levels of informational content and active, positively-

valenced SMEB in the form of creating and contributing (shares). For creating

behaviour, the level of 4 for entertaining content had the greatest impact on the

likelihood of creating behaviour occurring. For entertaining content levels of 5 or

greater, there was no significant relationship with creating behaviour. Similarly, when

predicting the occurrence of contributing behaviour in the form of shares, the maximum

number of entertaining content elements within a post was 3, increasing the odds ratio

of 2.88. For posts with greater than 3 elements of entertaining content, there was no

statistical significance to suggest a prediction of contributing behaviour in the form of

shares occurring. This suggests that there is a maximum level of entertaining content

within a post, beyond which the occurrence of active, positively-valenced engagement

behaviours cannot be predicted. Two entertaining content elements within a post were

found to predict the occurrence of contributing behaviour in the form of likes. H2a, high

levels of entertaining content weaken the relationship with positively-valenced

engagement behaviour is therefore partially supported.

6.2.5 Remunerative Content Presence

The level of remuneration offered to consumers has been studied as a driver of

consumer decisions to contribute to online communities (Muntinga et al., 2011).

Remunerative content has been previously found to have no effect on post shares

(Cvijikj & Michahelles, 2013). It is expected that a low level of engagement would

occur as a result of a post containing a reward or offer, for example ‘consuming’ rather

than contributing or co-creating (Muntinga et al., 2011). Hence it may be argued that

157 | P a g e
Chapter 6: Results
content which provides remunerative or economic benefit to the user leads to less active

expressions of SMEB,

H3. The presence of remunerative content facilitates passive, positively-valenced social

media engagement behaviour.

Tables 6.17 through 6.22 show the results for entertaining content and creating,

contributing, consuming, dormant and detaching behaviour. Additionally, each table

includes the results regarding the effect of the control variables post day of the week,

post month of the year, and post time of the day (AM or PM). Statistically significant

results are presented in bold.

Remunerative content and creating behaviour

Table 6.17 shows that remunerative content presence is a statistically significant

predictor of the occurrence of creating SMEB. The odds ratio of 1.29 for creating

behaviour shows that those posts which had remunerative content included were 1.29

times more likely to receive a comment than posts which did not have remunerative

content.

The occurrence of creating behaviour was influenced by the control variable, day of the

week. Posts made on Thursdays and Sundays were 25% and 37% respectively less

likely to facilitate the occurrence of creating behaviour, compared to the baseline day of

Wednesday.

The results also identify an effect of the control variable, post month of the year. The

statistically significant Exp(B) values for the months of March, July, August, September

and October indicate that for posts made in these months, the odds of creating behaviour

occurring are reduced compared to posts made in June. These results indicate that June
158 | P a g e
Chapter 6: Results
is a preferable month for posts to be made, compared to March, July, August,

September and October, in order to facilitate active, positively-valenced SMEB in the

form of creating.

Table 6.17 Logistic Regression Showing Effect of Remunerative Content Presence


on Creating Behaviour
SMEB IV B Sig Exp(B)
Creating Remunerative 0.25 0.04 1.29
Behaviour Monday -0.22 0.14 0.80
Tuesday -0.19 0.21 0.83
Thursday -0.29 0.04 0.75
Friday 0.03 0.86 1.04
Saturday -0.03 0.89 0.97
Sunday -0.47 0.00 0.63
January -0.33 0.14 0.72
February -0.21 0.32 0.81
March -0.49 0.01 0.61
April -0.35 0.07 0.70
May -0.17 0.41 0.85
July -0.51 0.01 0.60
August -0.47 0.03 0.62
September -0.52 0.01 0.59
October -0.55 0.01 0.58
November -0.31 0.13 0.73
December -0.25 0.23 0.77
AM -0.05 0.68 0.95

Remunerative content and contributing behaviour (shares)

Table 6.18 presents the results of the logistic regression concerning remunerative

content presence and contributing behaviour in the form of sharing. Informational

content presence is a statistically significant predictor of sharing behaviour, with an

Exp(B) of 1.77 indicating that when posts have remunerative content, they are 1.77

times more likely to be shared.

Remunerative posts were less likely to facilitate the occurrence of contributing

behaviour in the form of shares on Mondays (30% less likely), Thursdays (31% less

likely) and Sundays (36% less likely), compared to the baseline post day of Wednesday.

This finding suggests that in order to facilitate active SMEB, specifically contributing

159 | P a g e
Chapter 6: Results
through post shares, posting remunerative content on a Wednesday is preferable

compared to Mondays, Thursdays and Sundays. The results indicate that the likelihood

of sharing behaviour is decreased by 35% for posts made in the morning, compared to

remunerative posts made in the afternoon (Exp(B)=0.65). Remunerative social media

content is therefore more likely to be ‘shared’ in the afternoon.

Table 6.18 Logistic Regression Showing Effect of Remunerative Content Presence


on Contributing Behaviour (Shares)
SMEB IV B Sig Exp(B)
Contributing Remunerative 0.57 0.00 1.77
Behaviour Monday -0.36 0.02 0.70
(shares) Tuesday -0.12 0.42 0.88
Thursday -0.37 0.01 0.69
Friday -0.18 0.36 0.83
Saturday -0.32 0.16 0.73
Sunday -0.44 0.00 0.64
January 0.08 0.71 1.09
February -0.15 0.49 0.86
March -0.16 0.43 0.85
April -0.14 0.49 0.87
May -0.13 0.53 0.88
July -0.16 0.45 0.86
August -0.31 0.14 0.73
September -0.28 0.17 0.76
October -0.26 0.23 0.77
November -0.02 0.92 0.98
December -0.21 0.33 0.81
AM -0.43 0.00 0.65

Remunerative content and contributing behaviour (likes)

Table 6.19 shows that remunerative content presence is a statistically significant

predictor of the occurrence of contributing SMEB in the form of likes. The odds ratio of

1.96 for contributing behaviour (likes) indicates that posts which have remunerative

content are almost 2 times more likely to receive a ‘like’, compared to posts with no

remunerative content.

The variable for post day of the week influenced the likelihood of active, positively-

valenced SMEB occurring in the form of contributing (likes). Remunerative posts were

160 | P a g e
Chapter 6: Results
less likely to facilitate the occurrence of contributing behaviour in the form of likes on

Tuesdays (53% less likely) and Sundays (61% less likely). The variable for post month

of the year also influenced the likelihood of contributing behaviour in the form of likes

occurring. Posts made in February, March and November were less likely to be ‘liked’

compared to the baseline month June. Posts made in the morning (AM) significantly

and negatively predicted the occurrence of liking. This finding indicates that in order to

facilitate contributing behaviour in the form of likes, the time of the day is a significant

consideration, and afternoon posts are preferable.

Table 6.19 Logistic Regression Showing Effect of Remunerative Content Presence


on Contributing Behaviour (Likes)
SMEB IV B Sig Exp(B)
Contributing Remunerative 0.67 0.02 1.96
Behaviour Monday -0.50 0.09 0.61
(likes) Tuesday -0.76 0.01 0.47
Thursday -0.38 0.20 0.68
Friday 0.30 0.49 1.35
Saturday -0.10 0.83 0.90
Sunday -0.93 0.00 0.39
January -0.72 0.10 0.49
February -0.82 0.05 0.44
March -1.07 0.01 0.34
April -0.49 0.23 0.61
May 0.01 0.98 1.01
July -0.46 0.27 0.63
August -0.72 0.09 0.49
September 0.24 0.63 1.27
October -0.58 0.20 0.56
November -1.30 0.00 0.27
December -0.31 0.49 0.73
AM -0.44 0.03 0.65

Remunerative content and consuming behaviour

Table 6.20 presents the results of the binary logistic regression predicting consuming

SMEB when remunerative content is present.

Table 6.20 indicates that there is no statistical support for a relationship between the

presence of remunerative content and consuming behaviour. These results indicate that

161 | P a g e
Chapter 6: Results
this form of content is not a significant predictor of SMEB in the form of consuming

behaviour, including reading posts, viewing photos, watching videos, clicking on the

post and viewing photos. The control variables post day of the week, month of the year

and time of the day also had no statistically significant effect on the likelihood of

consuming SMEB occurring.

Table 6.20 Logistic Regression Showing Effect of Remunerative Content Presence


on Consuming Behaviour
SMEB IV B Sig Exp(B)
Consuming Remunerative 0.66 0.28 1.94
Behaviour Monday -1.47 0.06 0.23
Tuesday -0.83 0.32 0.44
Thursday -1.06 0.19 0.35
Friday -0.60 0.55 0.55
Saturday 16.17 1.00 10493884.30
Sunday -1.04 0.21 0.35
January 0.48 0.70 1.61
February -0.05 0.96 0.95
March 0.90 0.47 2.46
April 0.16 0.87 1.18
May 0.77 0.53 2.15
July 0.08 0.93 1.09
August -0.57 0.54 0.56
September -0.34 0.71 0.71
October -0.95 0.28 0.39
November -1.34 0.10 0.26
December -1.07 0.21 0.34
AM -0.01 0.98 0.99

Remunerative content and dormant behaviour

Table 6.21 presents the results of the logistic regression, testing the effect of

remunerative content on the occurrence of dormant behaviour.

The results show that the presence of remunerative content significantly and negatively

predicts the occurrence of dormant behaviour. The Exp(B) of 0.75 indicates that

dormant behaviour is 25% less likely to occur when a post contains remunerative

content, compared to when a post does not contain any remunerative content. These

findings suggest that the presence of remunerative content in a post can decrease the

162 | P a g e
Chapter 6: Results
odds of users remaining dormant in their behaviour.

The odds of dormant SMEB occurring increase when posts are made on Thursdays

(1.42 times more likely), and Sundays (1.47 times more likely), compared to posts made

Wednesdays. This finding suggest that users are less active in their SMEB on Thursdays

and Sundays, regardless of the presence of informational, entertaining or remunerative

content.

Significant effects were observed regarding the effect of month of the year on the

likelihood of dormant behaviour occurring. Dormant behaviour was less likely to occur

when the post was made in January (55% less likely). However it was more likely to

occur when the post was in April (1.55 times more likely) and July (1.89 times more

likely), compared to the June. This finding suggests that users are more active in their

engagement behaviour in January, as the odds of dormant behaviour occurring are

decreased, when compared to the middle of the year (June). Comparatively, users are

less active (more dormant) in the months of April and July, as the odds of dormant

behaviour occurring are increased, compared to June.

The variable for post time of the day did not have a significant effect on the prediction

of dormant behaviour.

163 | P a g e
Chapter 6: Results
Table 6.21 Logistic Regression Showing Effect of Remunerative Content on
Dormant Behaviour
SMEB IV B Sig Exp(B)
Dormant Remunerative -0.29 0.02 0.75
Behaviour Monday 0.25 0.10 1.29
Tuesday 0.21 0.18 1.24
Thursday 0.35 0.02 1.42
Friday 0.01 0.95 1.01
Saturday -0.43 0.06 0.65
Sunday 0.39 0.01 1.47
January -0.81 0.00 0.45
February -0.11 0.60 0.89
March 0.33 0.10 1.40
April 0.44 0.03 1.55
May 0.11 0.61 1.11
July 0.64 0.00 1.89
August 0.07 0.74 1.08
September 0.04 0.84 1.04
October -0.05 0.80 0.95
November 0.06 0.79 1.06
December -0.18 0.41 0.84
AM -0.08 0.52 0.92

Remunerative content and detaching behaviour

Table 6.22 indicates that the presence of remunerative content is not a statistically

significant predictor of negatively-valenced engagement behaviour in the form of

detachment. The control variables for post month of the year indicate that detaching

behaviour is less likely to occur in the later months of the year. Remunerative posts are

less likely to result in detaching behaviour when the post is made in September (55%

less likely), October (61% less likely), November (69% less likely) and December (75%

less likely). This finding indicates that remunerative content posts made later in the year

are preferable in order to mitigate the likelihood of users detaching from the content,

compared to the baseline month of June.

There was no significant effect for post day of the week or post time of the day in

predicting the likelihood of detaching behaviour occurring.

164 | P a g e
Chapter 6: Results
Table 6.22 Logistic Regression Showing Effect of Remunerative Content on
Detaching Behaviour

SMEB IV B Sig Exp(B)


Detaching Remunerative 0.17 0.26 1.19
Behaviour Monday 0.06 0.77 1.06
Tuesday 0.31 0.11 1.37
Thursday 0.20 0.29 1.22
Friday 0.15 0.57 1.16
Saturday 0.19 0.51 1.21
Sunday 0.13 0.49 1.14
January 0.18 0.48 1.19
February -0.32 0.21 0.72
March 0.20 0.37 1.22
April 0.11 0.64 1.11
May 0.04 0.86 1.04
July -0.10 0.69 0.91
August -0.06 0.81 0.94
September -0.80 0.00 0.45
October -0.95 0.00 0.39
November -1.18 0.00 0.31
December -1.37 0.00 0.25
AM -0.14 0.38 0.87

Remunerative content and social media engagement behaviour comparison

The summarised binary logistic regression results for each SMEB are displayed in

Table 6.23. The presence of remunerative content significantly predicts the occurrence

of active, positively-valenced SMEBs, creating and contributing. The greatest odds ratio

was for contributing behaviour in the form of likes, which increased in likelihood of

occurrence by almost 2 times when remunerative content was present within a post. The

presence of remunerative content was found to have a significant negative impact on

dormant behaviour. This means that when remunerative content is present within a

social media posts, users are more active in their expressions of engagement behaviour.

H3: The presence of remunerative content facilitates passive, positively-valenced social

media engagement behaviour is therefore not supported.

This result suggests that remunerative content can facilitate the occurrence of active,

positively-valenced SMEB in the form of creating and contributing. However, there is

165 | P a g e
Chapter 6: Results
no significant relationship between remunerative content and consuming behaviour.

This could be explained by the presence of a call-to-action with in remunerative posts,

for example ‘share/like to win’ or ‘share/like for discount/promotion’ etc., causing the

increase in contributing behaviour. There is no evidence to suggest that remunerative

details content facilitate increased consumption of the content. Finally, the presence of

remunerative content within a social media posts decreased the odds that users would

remain dormant.

Table 6.23 Summarised Logistic Regression Results for H3

Content SMEB b Sig Exp(B)


Remunerative Creating 0.25 0.04 1.29
Contributing (shares) 0.57 0.00 1.77
Contributing (likes) 0.67 0.02 1.96
Consuming 0.66 0.28 1.94
Dormancy -0.29 0.02 0.75
Detaching 0.17 0.26 1.19

6.2.6 Remunerative Content Level

Remunerative content may include monetary incentives, special offers, giveaways, prize

drawings, monetary compensations (Füller, 2006), contests and sweepstakes (Cvijikj &

Michahelles, 2013). In this study, the remunerative content variable contains 5 specific

types of remunerative content (see Table 5.8, Section 5.5.3, Chapter 5). In line with H1a

and as previously discussed, it was hypothesised that with increased levels of

remunerative content, users would experience information overload. Hence,

H3a: High levels of remunerative content weaken the relationship with positively-

valenced social media engagement behaviours.

Binary logistic regression was conducted with remunerative content as a categorical

independent variable. The categorical variable for remunerative content level has values

166 | P a g e
Chapter 6: Results
ranging from 1 to 4, with 4 being the highest number of remunerative elements a post in

this study had. A total of four categorical, dichotomous independent variables were

creating by dummy coding in order to indicate the presence (1) or absence (0) of each

specific level of content.

In the following section, the results regarding remunerative content level and its effect

on the occurrence of creating, contributing (likes), contributing (shares) and consuming

SMEB are discussed, as presented in Table 6.24.

Table 6.24 Logistic Regression Showing Effect of Remunerative Content Level on


SMEB

SMEB Content Level b Sig Exp(B)


Creating behaviour Remunerative (1) 0.27 0.05 1.31
Remunerative (2) 0.27 0.30 1.31
Remunerative (3) -0.81 0.51 0.45
Remunerative (4) -20.76 1.00 0.00
Contributing behaviour Remunerative (1) 0.52 0.00 1.68
(shares) Remunerative (2) 0.86 0.00 2.37
Remunerative (3) -21.11 1.00 0.00
Remunerative (4) 21.63 1.00 2478549027.84
Contributing behaviour Remunerative (1) 0.43 0.49 1.53
(likes) Remunerative (2) 17.12 1.00 27243354.93
Remunerative (3) 15.87 1.00 7822914.41
Remunerative (4) 15.95 1.00 8453525.37
Consuming behaviour Remunerative (1) -0.26 0.06 0.77
Remunerative (2) -0.43 0.09 0.65
Remunerative (3) 0.46 0.72 1.58
Remunerative (4) 20.23 1.00 611397334.86

H3a suggested that high levels of remunerative content weaken the relationship with

positively-valenced SMEBs. The results presented in Table 6.24 support a relationship

between lower levels of remunerative content and active, positively-valenced

engagement behaviours in the form of creating and contributing (shares). For creating

behaviour, the lowest level remunerative content (1) has a significant and positive

167 | P a g e
Chapter 6: Results
effect. However, when the level of remunerative content within the post increases from

1 to 2, 3 or 4, there is no support to suggest a significant effect on creating behaviour.

Similarly, for contributing behaviour in the form of shares, a maximum level of

remunerative content appears evident, beyond which there is no significant effect on

contributing behaviour in the form of shares. Remunerative content levels of 1 and 2

significantly and positively predict an increase in the odds of a post being shared.

However when the remunerative content level is increased to 3 items or more, there is

no statistical significance to support a relationship with positively-valenced engagement

behaviour. Hypothesis 3a: high levels of remunerative content weaken the relationship

with positively-valenced SMEBs is therefore partially supported.

6.2.7 Relational Content Presence

Users are motivated by social and relational needs when creating user generated content

online, as they find it a comfortable space in which to reveal feelings and share views

and opinions (Leung, 2009). Within Facebook, users are strongly motivated by

socialising needs when participating in specific interest groups (Park et al., 2009). The

socialisation benefits gained include gaining support from other members, meeting new

and interesting people and a developing sense of belonging to the community (Park et

al., 2009). Customers who are highly motivated by socialisation needs frequently

participate in human-to-human interactions, as defined by Ko et al. (2005). It is

hypothesised that relational social media content which stimulates interaction amongst

customers will be successful in facilitating positively-valenced SMEB;

H4. The presence of relational content facilitates active, positively-valenced social

media engagement behaviour.

168 | P a g e
Chapter 6: Results
Tables 6.25 through 6.30 show the results for relational content and creating,

contributing, consuming, dormant and detaching behaviour. Additionally, each table

includes the results regarding the effect of the control variables post day of the week,

post month of the year, and post time of the day (AM or PM). Statistically significant

results are presented in bold.

Relational content and creating behaviour

Table 6.25 shows the results of the binary logistic regression predicting the occurrence

of creating behaviour as a consequence of relational social media content. Relational

content is a significant predictor of creating behaviour. The odds of creating behaviour

occurring are increased by 1.41 times, compared to posts with no relational content.

This model also controls for the post day of the week, post time of the year (month) and

post time of the day (AM or PM). Relational social media content is less likely to

predict the occurrence of creating SMEB when it is posted on a Thursday (24% less

likely) or Sunday (37% less likely), compared to posts made on Wednesdays. The

month of the year in which the relational social media post is made has an effect on the

likelihood of active, positively-valenced SMEB in the form of creating occurring. The

odds of relational content resulting in creating behaviour occurring are significantly

reduced when the post is made in March (39% less likely), April (31% less likely), July

(41% less likely), August (38% less likely), September (40% less likely), and October

(42% less likely). The time of day in which the relational post is made does not

significantly impact on the odds of creating behaviour occurring.

169 | P a g e
Chapter 6: Results
Table 6.25 Logistic Regression Showing Effect of Relational Content Presence on
Creating Behaviour

SMEB IV B Sig Exp(B)


Creating Relational 0.34 0.00 1.41
behaviour Monday -0.21 0.15 0.81
Tuesday -0.16 0.30 0.85
Thursday -0.28 0.05 0.76
Friday 0.03 0.87 1.03
Saturday -0.01 0.97 0.99
Sunday -0.46 0.00 0.63
January -0.35 0.11 0.70
February -0.19 0.36 0.82
March -0.50 0.01 0.61
April -0.37 0.06 0.69
May -0.20 0.32 0.82
July -0.53 0.01 0.59
August -0.48 0.02 0.62
September -0.51 0.01 0.60
October -0.55 0.01 0.58
November -0.33 0.10 0.72
December -0.25 0.24 0.78
AM -0.07 0.59 0.94

Relational content and contributing behaviour (shares)

Contributing behaviour in the form of shares is significantly predicted by the presence

of relational content. The results of the binary logistic regression in Table 6.26 show

that when relational content is present within a post, the odds of the post being ‘shared’

are increased by 1.30 times compared to when the posts has no relational content

present. Relational posts are less likely to be shared on Mondays (30% less likely),

Thursdays (30% less likely) and Sundays (35% less likely) compared to Wednesdays.

The month in which the relational content post was made did not significantly influence

the likelihood of contributing behaviour in the form of shares occurring. The likelihood

of sharing behaviour occurring is decreased by 37% when the post is made in the

morning (AM).

170 | P a g e
Chapter 6: Results
Table 6.26 Logistic Regression Showing Effect of Relational Content Presence on
Contributing Behaviour (Shares)

SMEB IV B Sig Exp(B)


Contributing Relational 0.26 0.01 1.30
behaviour Monday -0.35 0.02 0.70
(shares) Tuesday -0.09 0.54 0.91
Thursday -0.36 0.01 0.70
Friday -0.19 0.33 0.83
Saturday -0.31 0.17 0.73
Sunday -0.43 0.00 0.65
January 0.05 0.80 1.06
February -0.14 0.50 0.87
March -0.20 0.31 0.82
April -0.17 0.40 0.85
May -0.18 0.38 0.84
July -0.18 0.38 0.84
August -0.32 0.14 0.73
September -0.27 0.18 0.76
October -0.24 0.27 0.79
November -0.05 0.81 0.95
December -0.20 0.36 0.82
AM -0.46 0.00 0.63

Relational content and contributing behaviour (likes)

Table 6.27 demonstrates that relational content presence is a significant predictor of

contributing behaviour in the form of likes. When a post contains relational content, the

likelihood that contributing behaviour in the form of likes will occur is increased by 2.2

times, compared to when it does not contain relational content.

Relational content posts made on Tuesdays and Sundays significantly reduce the odds

of contributing behaviour in the form of likes occurring, by 48% and 60% respectively

compared to posts made on Wednesdays. The variables for post month of the year

indicate that contributing behaviour in the form of likes is less likely to occur when the

post is made in March (66% less likely) and November (74% less likely) show that

when posts are made in March and November. When relational posts were made in the

morning, the odds of contributing behaviour in the form of likes occurring decreased by

37%.

171 | P a g e
Chapter 6: Results
Table 6.27 Logistic Regression Showing Effect of Relational Content Presence on
Contributing Behaviour (Likes)
SMEB IV B Sig Exp(B)
Contributing Relational 0.79 0.00 2.20
behaviour Monday -0.48 0.11 0.62
(likes) Tuesday -0.66 0.03 0.52
Thursday -0.34 0.25 0.71
Friday 0.29 0.51 1.34
Saturday -0.06 0.89 0.94
Sunday -0.91 0.00 0.40
January -0.76 0.08 0.47
February -0.78 0.06 0.46
March -1.10 0.00 0.33
April -0.55 0.18 0.58
May -0.06 0.90 0.94
July -0.46 0.27 0.63
August -0.73 0.09 0.48
September 0.26 0.59 1.30
October -0.53 0.23 0.59
November -1.35 0.00 0.26
December -0.30 0.50 0.74
AM -0.47 0.02 0.63

Relational content and consuming behaviour

Consuming behaviour characterises a passive, positively-valenced form of SMEB.

Users click on photos, read posts, click on links and watch videos when they engage in

consuming SMEB.

The presence of relational content is a significant predictor of the occurrence of

consuming behaviour, with the odds ratio of 3.73 indicating that for posts with

relational content present, the likelihood of consuming behaviour occurring increases by

3.7 times, compared to posts without relational content. The control variables included

in this model of: day, month and time, did not have a statistically significant effect on

the occurrence of SMEB in the form of consuming.

172 | P a g e
Chapter 6: Results
Table 6.28 Logistic Regression Showing Effect of Relational Content Presence on
Consuming Behaviour

SMEB IV B Sig Exp(B)


Consuming Relational 1.32 0.00 3.73
behaviour Monday -1.41 0.07 0.24
Tuesday -0.60 0.48 0.55
Thursday -1.00 0.21 0.37
Friday -0.53 0.60 0.59
Saturday 16.27 1.00 11633903.20
Sunday -0.98 0.23 0.37
January 0.46 0.71 1.59
February 0.05 0.96 1.05
March 0.95 0.44 2.59
April 0.11 0.91 1.12
May 0.71 0.57 2.03
July 0.10 0.92 1.10
August -0.55 0.56 0.58
September -0.28 0.77 0.76
October -0.91 0.30 0.40
November -1.40 0.09 0.25
December -1.04 0.22 0.35
AM 0.00 0.99 1.00

Relational content and dormant behaviour

Relational content presence within social media is not a statistically significant predictor

of dormant behaviour, as shown in Table 6.29. Dormant behaviour is characterised by

users who are exposed to the social media post (which contains relational content) and

do not take any actions to consume, contribute to or create content.

The control variable, post day of the week, increased the odds of dormant behaviour

occurring. Dormant behaviour was more likely to occur when a relational post was

made on a Thursday (1.4 times more likely) and Sunday (1.47 times more likely),

compared to posts made on a Wednesday. In contrast, when relational content is posted

on a Saturday, the odds of dormant behaviour occurring are reduced by 35%.

Relational social media content posted in April and July increases the odds of dormant

SMEB occurring by 1.57 and 1.91 times respectively, compared to the baseline month

173 | P a g e
Chapter 6: Results
of June. Relational content posts made in January reduce the likelihood of dormant

behaviour occurring by 55%. The post time of the day did not significantly impact on

the likelihood of dormant behaviour occurring.

Table 6.29 Logistic Regression Showing Effect of Relational Content Presence on


Dormant Behaviour
SMEB IV B Sig Exp(B)
Dormant Relational -0.11 0.26 0.90
behaviour Monday 0.25 0.10 1.28
Tuesday 0.20 0.21 1.22
Thursday 0.34 0.02 1.40
Friday 0.02 0.93 1.02
Saturday -0.44 0.05 0.65
Sunday 0.38 0.01 1.47
January -0.80 0.00 0.45
February -0.11 0.60 0.89
March 0.36 0.08 1.43
April 0.45 0.03 1.57
May 0.13 0.53 1.14
July 0.65 0.00 1.91
August 0.08 0.73 1.08
September 0.04 0.84 1.04
October -0.06 0.77 0.94
November 0.07 0.74 1.07
December -0.18 0.39 0.83
AM -0.06 0.62 0.94

Relational content and detaching behaviour

Relational content presence is not a statistically significant predictor of detaching

SMEB, as shown in Table 6.30. The only type of social media content that predicted the

occurrence was therefore entertaining content, which was shown in Table 6.14 to

increase the odds of detaching behaviour occurring by 1.24 times.

The control variables for post month of the year indicate that detaching behaviour is less

likely to occur in the later months of the year. The odds of detaching behaviour are

reduced in September (55% less likely), October (61% less likely), November (69% less

likely) and December (75% less likely), compared to June. This finding is consistent

174 | P a g e
Chapter 6: Results
with the effects of post month of the year and detaching behaviour for informational

content (Table 6.6), entertaining content (Table 6.14) and remunerative content (Table

6.22).This finding indicates that relational content posts made later in the year are

preferable in order to mitigate the likelihood of users detaching from the content,

compared to the baseline month of June. Post day of the week and time of the day did

not have a statistically significant effect on the likelihood of detaching behaviour

occurring.

Table 6.30 Logistic Regression Showing Effect of Relational Content Presence on


Detaching Behaviour
SMEB IV B Sig Exp(B)
Detaching Relational -0.03 0.81 0.97
Behaviour Monday 0.05 0.78 1.05
Tuesday 0.31 0.11 1.37
Thursday 0.20 0.29 1.22
Friday 0.14 0.59 1.15
Saturday 0.18 0.53 1.20
Sunday 0.13 0.50 1.14
January 0.17 0.49 1.19
February -0.33 0.21 0.72
March 0.18 0.42 1.20
April 0.10 0.66 1.10
May 0.03 0.89 1.03
July -0.10 0.67 0.90
August -0.06 0.81 0.94
September -0.80 0.00 0.45
October -0.94 0.00 0.39
November -1.18 0.00 0.31
December -1.37 0.00 0.25
AM -0.15 0.35 0.86

Relational content and social media engagement behaviour comparison

Table 6.31 shows that the presence of relational content significantly predicts an

increase in the likelihood that positively-valenced, active and passive SMEB occurs.

The highest odds ratio for consuming behaviour of 3.73 indicates that for posts which

have relational content present, the odds that the content will be consumed increase by

almost 3.8 times, compared to posts which do not have relational content.

175 | P a g e
Chapter 6: Results
Significant and positive relationships are also found between relational content presence

and creating behaviour (1.4 times more likely to occur), contributing behaviour in the

form of likes (2.2 times more likely to occur) and contributing behaviour in the form of

shares (1.3 times more likely to occur).

H4: The presence of relational content facilitates active, positively-valenced social

media engagement behaviour is supported. However the likelihood of passive,

positively-valenced SMEB occurring as a result of relational content presence is much

stronger, with an odds ratio of 3.73.

Table 6.31 Summarised Logistic Regression Results for H4

Content SMEB b Sig Exp(B)


Relational Creating 0.34 0.00 1.41
Contributing (likes) 0.79 0.00 2.20
Contributing (shares) 0.26 0.01 1.30
Consuming 1.32 0.00 3.73
Dormancy -0.11 0.26 0.90
Detaching -0.03 0.81 0.97

6.2.8 Relational Content Level


The summarised results in Table 6.31 show that relational content presence significantly

predicts the occurrence of creating, contributing (shares and likes) and consuming

behaviour. In this section, this effect is explored further by applying binary logistic

regression to test how the specific levels of relational content effects each of the

behaviours. H4a suggests that high levels of relational content weaken the relationship

with positively-valenced social media engagement behaviours.

The relational content category contained 15 relational elements (see Table 5.9, Section

5.5.3, Chapter 5). The results showed that the maximum number of relational elements

in any one post within the study was 7 elements. The relational content categorical

176 | P a g e
Chapter 6: Results
variables used to test the effect of each level therefore range in values from 1 (minimum

level of relational content) to 7 (maximum level of relational content). In order to input

these into the model as independent variables, seven dummy coded dichotomous

variables were created with ‘1’ indicating presence of the relevant level of content, and

‘0’ indicating the absence. The results for the effect of the 7 levels of relational content

and the consequent effects on SMEB are summarised in Table 6.32.

Table 6.32 Logistic Regression Showing Effect of Relational Content Level on


SMEB

SMEB Content Level b Sig Exp(B)


Creating Behaviour Relational (1) 0.25 0.02 1.29
Relational (2) 0.37 0.00 1.45
Relational (3) 0.58 0.00 1.79
Relational (4) 0.39 0.14 1.48
Relational (5) 1.00 0.03 2.71
Relational (6) -0.56 0.52 0.57
Relational (7) 21.72 1.00 2696854057.59
Contributing behaviour Relational (1) 0.22 0.04 1.25
(shares) Relational (2) 0.23 0.06 1.25
Relational (3) 0.50 0.00 1.65
Relational (4) 0.44 0.10 1.55
Relational (5) -0.14 0.76 0.87
Relational (6) -0.38 0.66 0.68
Relational (7) 21.60 1.00 2401099102.04
Contributing behaviour Relational (1) 0.57 0.00 1.77
(likes) Relational (2) 0.92 0.00 2.51
Relational (3) 1.34 0.00 3.81
Relational (4) 1.09 0.07 2.96
Relational (5) 19.41 1.00 270115201.31
Relational (6) 19.08 1.00 192637748.97
Relational (7) 20.44 1.00 754128574.71
Consuming behaviour Relational (1) 0.99 0.02 2.68
Relational (2) 1.65 0.01 5.22
Relational (3) 1.67 0.11 5.32
Relational (4) 17.77 1.00 52090846.02
Relational (5) 17.99 1.00 64966239.42
Relational (6) 17.10 1.00 26704715.45
Relational (7) 18.91 1.00 163570271.02

177 | P a g e
Chapter 6: Results
Hypothesis 4a suggested that high levels of relational content weaken the relationship

with positively-valenced engagement behaviour. In Table 6.32 it can be seen that as the

level of relational content increases, so do the odds ratios for creating, contributing

(shares), contributing (likes) and consuming SMEBs. However, for each of these

behaviours, the effect is significant to a certain level of relational content, beyond which

there is no statistical support to predict the occurrence of the behaviour. For creating

behaviour, this point is at the level of 5. For higher relational content levels of 6 and 7,

there is no statistically significant effect on creating behaviour. Relational content levels

of 3 and under significantly increase the odds of contributing behaviour in the form of

shares and likes. For higher levels of relational content (4 to 7) there is no statistical

support for a relationship with active, positively-valenced engagement behaviour. The

prediction of passive, positively-valenced engagement behaviour is also limited by the

level of relational content within the post. Table 6.32 indicates that while 1 and 2

elements of relational content within a post significantly increase the odds of consuming

behaviour occurring, there is no statistically significant relationship between high levels

(3 to 7) of relational content and consuming behaviour. H4a is therefore supported.

6.2.9 Social Media Content Presence Summary

The effects of each social media content type on each SMEB are summarised in Table

6.33. In this table, comparison of each engagement behaviour and the relative effect of

each social media content type are presented. Statistically significant content types are

presented in bold.

178 | P a g e
Chapter 6: Results
Table 6.33 Binary Logistic Regression Results for Social Media Content and Social
Media Engagement Behaviour
SMEB Content Type b Sig Exp(B)
Creating Informational .34 .00 1.41
Entertaining .34 .00 1.41
Remunerative 0.25 0.04 1.29
Relational 0.34 0.00 1.41
Contributing Informational .90 .00 2.46
(Likes) Entertaining .54 .05 1.24
Remunerative 0.67 0.02 1.96
Relational 0.79 0.00 2.20
Contributing Informational .65 .00 1.91
(Shares) Entertaining .34 .00 1.41
Remunerative 0.57 0.00 1.77
Relational 0.26 0.01 1.30
Consuming Informational .82 .03 2.27
Entertaining -.01 .97 .99
Remunerative 0.66 0.28 1.94
Relational 1.32 0.00 3.73
Dormant Informational -.07 .54 .93
Entertaining -.43 .00 .65
Remunerative -0.29 0.02 0.75
Relational -0.11 0.26 0.90
Detaching Informational .29 .06 1.33
Entertaining .216 .05 1.24
Remunerative 0.17 0.26 1.19
Relational -0.03 0.81 0.97

Table 6.33 shows that the presence of informational, entertaining, remunerative and

relational content within a social media post can facilitate the occurrence of active,

positively-valenced SMEB in the form of creating behaviour. The highest odds ratios

are for informational content, relational content and entertaining content. Similarly, the

presence of all four content types can facilitate the occurrence of active, positively-

valenced engagement behaviour in the form of contributing behaviour. This is measured

through shares and likes.

For contributing behaviour in the form of likes, whilst the presence of all four content

types support a significant and positive relationship, the behaviour is most likely to

179 | P a g e
Chapter 6: Results
occur when informational content is present (Exp(B) =2.46). When predicting the

occurrence of contributing behaviour in the form of shares, the results are similar. The

presence of all four of the content types significantly and positively increases the odds

that SMEB in the form of contributing (shares) will occur. However sharing is most

likely to occur when informational content is present (Exp(B) = 1.91).

With regards to passive, positively-valenced engagement behaviour (consuming), the

presence of entertaining and remunerative social media content were unable to

significantly predict occurrence. However, the presence of informational content within

a post, and the presence of relational content within a post, significantly and positively

increased the odds of consuming SMEB occurring. The greatest odds ratio was

observed for relational content (Exp(B)= 3.73) indicating that the presence of relational

content is the strongest predictor of passive, positively-valenced SMEB.

The presence of entertaining social media content and remunerative social media

content significantly reduced the odds of dormant SMEB occurring. This effect was

greater for entertaining content, which reduced the odds of dormant behaviour occurring

by 35%.

6.2.10 Social Media Content Level Summary

This section presented the results of binary logistic regression which detailed the effects

of specific levels of informational (H1a), entertaining (H2a), remunerative (H3a) and

relational content (H4a) on SMEB.

The level of informational, entertaining, remunerative, and relational content present

within a social media post had a significant effect on the likelihood of active, positively-

valenced SMEB in the form of creating occurring. For each of these content types, it

180 | P a g e
Chapter 6: Results
was found that creating behaviour was only significantly and positively predicted at

certain levels of content. Regarding informational content, as the level of informational

content within a post increased from 2 to 7, the likelihood of creating behaviour

occurring also increased. Lower levels of entertaining content were required in order to

facilitate creating behaviour, with higher levels of entertaining content showing an

information overload effect. The level of 4 for entertaining content had the greatest

impact on the likelihood of creating behaviour occurring. An even lower level of

remunerative content was required in order for creating behaviour to be significantly

predicted. The results showed that only one element of remunerative content within a

post has a significant and positive effect. For relational content, the maximum level of

content that significantly predicted creating behaviour was 5. These findings suggest

that users are more tolerant of higher levels of informational (up to 7 elements) and

relational (up to 5 elements) content in regards to the prediction of creating behaviour.

A much lower level of remunerative content (1 element) is required in order to

significantly predict an increase in the occurrence creating behaviour.

The level of informational, entertaining, remunerative and relational content within a

post was a significant determinant of the occurrence of SMEB in the form of

contributing. When predicting the occurrence of contributing behaviour in the form of

shares, the notion of information overload was supported. A maximum level of content

was determined for each type of social media content in order to significantly and

positively predict the likelihood of the post being shared. The level was greatest for

informational content (7 elements) and lower for entertaining content (3 elements),

relational content (3 elements), and remunerative content (2 elements).

181 | P a g e
Chapter 6: Results
The prediction of contributing behaviour in the form of likes was also dependent on the

level of informational, entertaining and relational content present within a post.

Informational content with up to 5 elements significantly and positively predicted the

occurrence of contributing behaviour in the form of likes occurring, while the level of

entertaining content (maximum of 2 elements) and relational content (3) required to

predict contributing behaviour in the form of likes was lower. The level of remunerative

content within a post did not significantly impact on the occurrence of contributing

behaviour in the form of likes. Consistent with the predictions regarding informational

content level and creating behaviour as well as contributing behaviour in the form of

shares, a higher level of informational content was tolerated by users when predicting

the odds of the post being ‘liked’.

When predicting consuming SMEB, the level of informational, entertaining and

remunerative content did not have a statistically significant effect. . Table 6.32 indicated

that 1 and 2 elements of relational content within a post can significantly increase the

odds of consuming behaviour occurring. However, there is no statistically significant

relationship between higher levels (3 to 7) of relational content and consuming

behaviour, indicating a point of information overload occurring. The level of relational

content within a post did however have an effect on the likelihood of consuming

behaviour occurring.

6.3 Interaction Effects


Within social media, there is potential for content to possess multiple cues and thus

deliver content which may simultaneously contain entertaining, informational,

remunerative and relational content. The four social media content categories outlined

and tested in the previous sections are therefore not mutually exclusive.
182 | P a g e
Chapter 6: Results
It was expected that highly utilitarian social media content (informational or

remunerative content) presented simultaneously with highly hedonic social media

content (entertaining or relational content) would cause information overload and a

conflict of processing styles, resulting in a lack of attention or content avoidance by the

consumer. Thus:

H5a: The simultaneous presence of informational and entertaining content facilitates


negatively-valenced social media engagement behaviours.

H5b: The simultaneous presence of informational and relational content facilitates


negatively-valenced social media engagement behaviours.

H5c: The simultaneous presence of informational and remuneration content facilitates


positively-valenced social media engagement behaviours.

H5d: The simultaneous presence of entertaining and remunerative content facilitates


negatively-valenced social media engagement behaviours.

H5e: The simultaneous presence of entertaining and relational content facilitates


positively-valenced social media engagement behaviours.

H5f: The simultaneous presence of relational and remunerative content facilitates


negatively-valenced social media engagement behaviour.

The results regarding the interaction effects are summarised in Table 6.34 and described

throughout the following sections.

183 | P a g e
Chapter 6: Results
Table 6.34 Interaction Effects Summary

SMEB Interactions (amount of content) b Sig Exp(B)


Entertaining (1) by Informational (1) -0.77 0.01 0.46
Creating
Entertaining (1) by Informational (4) -1.19 0.00 0.30
Entertaining (1) by Informational (5) -1.43 0.00 0.24
Entertaining (1) by Informational (8) -3.29 0.04 0.04
Entertainment (2) By Relational (2) -.94 .01 .39
Entertainment (3) by Relational (1) 1.86 .01 6.44
Contributing Entertaining (1) by Informational (4) -.90 .02 .40
(Shares) Informational (1) by Relational (2) -0.72 0.04 0.49
Contributing
(Likes) Relational (1) by Remunerative (1) 1.55 0.04 4.70
Entertaining (1) by Informational (3) 1.13 0.00 3.10
Dormancy
Entertaining (1) by Informational (5) 1.27 0.01 3.56
Entertaining (1) by Informational (6) 1.67 0.03 5.30
Informational (1) by Relational (2) 0.83 0.03 2.30
Informational (4) by Relational (2) 1.31 0.01 3.71
Informational (4) by Relational (4) 3.69 0.02 40.16
Informational (6) by Relational (2) 2.96 0.05 19.32
Informational (6) by Relational (4) 4.18 0.04 65.48
Entertaining (1) by Remunerative (1) 0.95 0.00 2.58
Relational (3) by Remunerative (1) 1.22 0.02 3.39
Entertaining (1) by Informational (4) -1.13 0.03 0.32
Detaching
Entertainment (3) by Relational (2) 1.74 0.04 5.68

6.3.1 Informational and Entertaining Content Interaction

In this section, the results of informational content interactions with entertaining content

are presented. The full results are presented in Appendix E and significant relationships

are summarised in Table 6.34. The results support a negative interaction effect between

informational and entertaining content on the prediction of creating behaviour. The

results indicate that the odds of creating behaviour occurring when the post contains one

element of informational and one element of entertaining content decreased by 54%.

Similarly, 1 element of entertaining content, combined with 4 and 5 elements of

informational content decreased the odds of creating behaviour occurring by 70% and

76% respectively. A higher level of informational content within a post (8 elements)


184 | P a g e
Chapter 6: Results
combined with 1 element of entertaining content also reduces the odds of creating

behaviour occurring by 96%.

These findings suggest a negative effect of combined conflicting content types such as

hedonic (entertaining) and utilitarian (informational) on creating behaviour. The

presence of informational and entertaining content within a post were previously found

to independently increase the likelihood of creating behaviour occurring. However,

when combined, the likelihood of creating behaviour occurring significantly decreased.

There was no statistical significance to suggest that the simultaneous presence of

informational and entertaining content interactions predicted the occurrence of

contributing behaviour in the form of shares or likes.

The interaction between entertaining content and informational content was a

significant and positive predictor of the occurrence of dormant behaviour. The results in

Table 6.34 indicate that the simultaneous presence of one element of entertaining

content with greater levels of informational content increase the odds of dormant

behaviour occurring by 3.1 times (3 elements of informational content, 1 element of

entertaining), 3.56 times (5 elements of informational and 1 element of entertaining)

and 5.3 times (6 elements of informational content, 1 element of entertaining). These

results indicate that combining entertaining content with informational content at certain

levels facilitates the occurrence of dormant behaviour. As the level of informational

content that is presented simultaneously with entertaining content increases, the effect

size increases.

Table 6.34 shows that when informational content (specifically 4 elements) is presented

simultaneously with one element of entertaining content, the likelihood of detaching

185 | P a g e
Chapter 6: Results
behaviour occurring is decreased by 68%.

H5a: The simultaneous presence of informational and entertaining content facilitates

negatively-valenced SMEBs is therefore not supported. While the presence of

informational and entertaining content increased the likelihood of occurrence of neutral

SMEB in the form of dormancy occurring, the interaction reduced the odds of

negatively-valenced engagement behaviour in the form of detaching occurring.

However, there is evidence to suggest that the simultaneous presence of informational

and entertaining content has a negative effect on the likelihood of active, positively-

valenced engagement behaviour in the form of creating and contributing (shares)

occurring.

6.3.2 Informational and Relational Content Interaction

H5b predicted that the simultaneous presence of informational and relational content

facilitates negatively-valenced SMEBs. The results indicated that the simultaneous

presence of informational and relational content within a post did not have a statistically

significant effect on the likelihood of positively-valenced engagement behaviour

(creating, contributing and consuming) occurring. The simultaneous presence of

informational and relational content increased the odds of the neutral state, dormancy

behaviour occurring. Table 6.34 shows that when one element of informational content

is presented simultaneously with 2 elements of relational content, the odds of dormant

behaviour occurring are increased by 2.3 times. When the level of informational content

increases to 4 elements and is presented simultaneously with 2 elements of relational

content, the odds of dormant behaviour occurring increase by 3.71 times. This level of

informational content (4 elements) presented simultaneously with a greater level of

relational content (4 elements) has a significant effect on dormant behaviour, increasing


186 | P a g e
Chapter 6: Results
the odds of dormant behaviour occurring by 40.16 times. Further, a greater amount of

informational content (6 elements) combined with relational content (4 elements)

increased the odds of dormant behaviour occurring by 65.48 times. There was no

significant effect on negatively-valenced SMEB occurring, as a result of the combined

informational and relational content presence. Therefore, H5b is not supported. It is

interesting to note however, that the simultaneous presence of these types of content

does significantly increase the odds of users remaining dormant in their engagement

behaviour.

6.3.3 Informational and Remunerative Content Interaction

Hypothesis 5c predicted that as informational and remunerative content are both

‘utilitarian’ content types, there will be no conflicting demands placed on the user. The

user can therefore adequately process the information. Hence, it was predicted that; the

presence of informational and remunerative content facilitates positively-valenced

engagement behaviours. The results (Appendix E) show no significant effects of the

simultaneous presence of informational and remunerative content on the prediction of

SMEB. H5c is therefore not supported. This demonstrates that there is no benefit of

combining informational content with remunerative content. It is possible that contrary

to the initial expectation, the remunerative and informative components of the post do in

fact pose conflicting demands on the reader and hence are not adequately processed.

6.3.4 Entertaining and Remunerative Content Interaction


H5d predicted that the simultaneous presence of entertaining and remunerative content

facilitates negatively-valenced social engagement behaviours.

The results in Appendix E show that there were no significant effects regarding the

187 | P a g e
Chapter 6: Results
simultaneous presence of entertaining content and remunerative content on positively-

valenced SMEBs.

The results support an interaction effect between 1 element of entertaining content and 1

element of remunerative content. As shown in Table 6.34, the entertaining (1) by

remunerative (1) variable significantly and positively predicts the occurrence of

dormant behaviour. The odds ratio of 2.58 indicates that when a post includes 1 element

of remunerative and entertaining content, the likelihood that dormant behaviour will

occur is increased by 2.58 times. There were no significant effects of the simultaneous

presence of entertaining content and remunerative content on negatively-valenced

SMEB. H5d: The simultaneous presence of entertaining and remunerative content

facilitates negatively-valenced social media engagement behaviours is therefore not

supported. This finding indicates that it is not necessarily detrimental to the brand to

combine entertaining content with remunerative content.

6.3.5 Entertaining and Relational Content Interactions

Entertaining and relational content are both hedonic types of social media content.

Therefore it was predicted that their effect on social media engagement would be

positive, Hence, H5e predicted that the simultaneous presence of entertaining and

relational content facilitates positively-valenced social media engagement behaviours.

The results, presented in Table 6.34, indicate a significant effect of the simultaneous

presence of entertaining and relational content on active, positively-valenced

engagement behaviour in the form of creating. The results are specific to the level of

content within the post. For example, when a post contains an equal amount of

entertaining content (2 elements) and relational content (2 elements), the odds of

188 | P a g e
Chapter 6: Results
creating behaviour occurring are decreased by 61%. However, when the post contains

slightly more entertaining content (3 elements) and slightly less relational content (1

element), the result is very different. In this scenario, the likelihood of creating

behaviour occurring is increased by 6.44 times.

The simultaneous presence of entertaining content and relational content within a post

did not have a significant effect on positively-valenced SMEB in the form of

contributing or consuming. Similarly, there was no significant effect of the

simultaneous presence of entertaining content and relational content on the likelihood of

dormant and behaviour occurring.

The simultaneous presence of entertaining and relational content within a social media

post significantly increased the odds of detaching behaviour occurring at specific levels

of content. Table 6.34 shows that the presence of 3 elements of entertaining content

within a post, presented simultaneously with 2 elements of relational content increases

the odds of detaching behaviour occurring by 5.68 times.

H5e: the simultaneous presence of entertaining and relational content facilitates

positively-valenced social media engagement behaviours is therefore partially

supported. The simultaneous presence of entertaining and relational content

significantly predicted an increased in the odds of creating behaviour occurring.

However this effect was dependent on the specific levels of entertaining and relational

content that were simultaneously presented.

6.3.6 Relational and Remunerative Content Interaction


Hypothesis 5f predicted that the simultaneous presence of relational and remunerative

content facilitates negatively-valenced social media engagement behaviour.

189 | P a g e
Chapter 6: Results
The simultaneous presence of relational and remunerative content did not have a

statistically significant impact on the likelihood of creating, contributing or detaching

behaviour occurring. The results in Table 6.34 show that there is an effect on dormant

behaviour, but only at one specific level of the entertaining content and remunerative

content. The results show that when three elements of relational content are presented

simultaneously with 1 element of remunerative content, the odds of dormant behaviour

occurring increase by 3.39 times.

As there were no significant effects identified regarding the simultaneous presence of

relational and remunerative content on negatively-valenced engagement behaviour

(detaching), H5f was not supported.

6.3.7 Interaction Effects Summary

This section tested the effects of social media content interactions on SMEBs.

Significant effects were found for three SMEBs; creating, contributing (shares) and

dormancy. Table 6.34 provided a summary of the results. For the purpose of this

summary, only statistically significant predictors of each SMEB were presented in this

table.

It was expected that utilitarian social media content (informative and remunerative)

presented simultaneously with hedonic social media content (entertaining and

relational) would cause information overload and a conflict of processing styles,

resulting in a lack of attention or content avoidance by the consumer. Thus:

H5a: The presence of informational and entertaining content facilitates negatively-


valenced social media engagement behaviours.

190 | P a g e
Chapter 6: Results
H5b: The presence of informational and relational content facilitates negatively-
valenced social media engagement behaviours.

H5c: The presence of informational and remuneration content facilitates positively-


valenced social media engagement behaviours.

H5d: The presence of entertaining and remunerative content facilitates negatively-


valenced social media engagement behaviours.

H5e: The presence of entertaining and relational content facilitates positively-valenced


social media engagement behaviours.

H5f: The presence of relational and remunerative content facilitates negatively-


valenced social media engagement behaviours.

The results in Table 6.34 indicate partial support for H5e, while H5a, H5b, H5c, H5d

and H5f were not supported.

6.4 Moderation
Moderation process analysis is a regression based approach to statistical testing. The

analysis was conduct through ‘PROCESS’ which is a computational tool for path

analysis-based moderation and mediation analysis (Hayes, 2013). In order to test the

hypotheses of the study concerning moderation effects, Hayes (2015) method of

modifying the PROCESS model to estimate a simple moderation model with a three-

category moderator was applied.

The hypotheses specified in Chapter 4 demonstrate a conceptual model in which three

variables; media richness, congruity of the post and community size are estimated as

moderating a single focal predictor’s (in this study, social media content) effect on

SMEB. The ‘PROCESS’ computational tool enables estimation in this type of model,

by implementing the necessary computations for probing the interaction and visualising

191 | P a g e
Chapter 6: Results
the results.

The following section applies Hayes PROCESS Model 2 in order to test the moderation

effects of community size, media richness and congruity of the post on the relationship

between social media content and SMEBs. PROCESS Model 2 was chosen as the

moderating variables were categorical variables, as discussed in the following section.

6.4.1 Hayes PROCESS Moderation Model with Three Category Moderator

Hayes (2013) PROCESS macro for SPSS provides a means for estimating a model with

moderation of a variable X’s (social media content) on Y (SMEB) by moderator M

(media richness, congruity and community size). Hayes (2015) method of ‘hacking’

PROCESS Model 2 was applied to estimate a simple moderation model with a

moderator that is multi-categorical with three levels was applied in order to test the

moderation effects. This process was outlined in Chapter 5, Section 5.7.2. The

following sections present the results of the moderation analysis for media richness,

congruity and community size, followed by a summary of the moderating effects.

6.4.2 Media Richness

Within social media content, specifically delivered by brands on Facebook, media

richness is categorised into three levels as shown in Table 6.35. This operationalisation

is derived from De Vries et al. (2012) and Cvijikj and Michahelles (2013); 1) Low

media richness for status updates as they are in the form of written text, 2) medium

richness for photos and images as they include pictorial content, 3) high vividness for

videos as they offer sound and imagery.

192 | P a g e
Chapter 6: Results
Table 6.35 Richness Operationalisation

Media Richness Level Operationalisation


Low Status updates (text only)
Medium Photos and images (imagery, no audio)
High Videos (text, imagery and audio)

It was proposed that the richness of social media content (low, medium and high)

moderates the relationships between social media content and SMEB as outlined in

hypothesis 6; H6: The strength of the relationship between social media content and

social media engagement behaviour is directly related to media richness.

Using the SPSS code to construct PROCESS Model 2 for a multi-categorical

moderating variable (Hayes, 2015), tests of the moderation of the effect of

informational, entertaining, remunerative and relational content on the varying types of

SMEB, by media richness were achieved. Among the tests of moderation, richness was

found to moderate one relationship: the effect of informational content on contributing

behaviour in the form of shares. The results of this moderation effect are presented in

the following section.

Moderation of the effect of informational content on contributing behaviour (shares) by


richness

Using the SPSS code to construct PROCESS Model 2 for a multi-categorical

moderating variable (Hayes, 2015), a test of the moderation of the effect of

informational content level on creating behaviour by richness was achieved. The level

of informational content used as the independent variable in this case is derived from

the results in Section 6.2.2. Table 6.8 showed that informational content levels between

1 and 7 have a statistically significant, positive effect on the likelihood of contributing

behaviour in the form of sharing occurring. Table 6.36 shows the test of interaction

between media richness and consuming SMEB.


193 | P a g e
Chapter 6: Results

Table 6.36 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Shares) by Media
Richness

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (information x status/link) 0.00 0.36 1.00 1823.00 0.55
int_2 (information x video) 0.00 7.43 1.00 1823.00 0.01
Both 0.00 3.99 2.00 1823.00 0.02

The “Both” row in Table 6.36 provides a test of the null hypothesis that media richness

does not moderate the effect of informational content presence on contributing

behaviour in the form of shares. The null hypothesis can be rejected, F (2, 1823) =3.99,

p <0.05. In other words, the regression slope quantifying contributing behaviour

(shares) as a function of informational content depends on media richness.

In order to visualise the moderation effect, the plot option within the PROCESS code

generates a table (Table 6.37) of values of the moderator (D1 status/link and D2 video),

focal predictor (X, informational content) and estimated values of Y (contributing

behaviour, shares). The table contains estimates of contributing behaviour in the form of

shares in each of the three groups of informational content level. The three groups are

generated by PROCESS and defined as the sample mean (2.44) as well as a standard

deviation below (0.95) and a standard deviation (3.94) above the mean, shown in the

first column of Table 6.37.

Table 6.37 PROCESS Data for Visualising Conditional Effect of Informational


Content on Contributing Behaviour (Shares)

Information Level Video Status/link Contributing (shares)


0.95 0.00 0.00 0.95
2.44 0.00 0.00 1.06
3.94 0.00 0.00 1.16
0.95 0.00 1.00 0.38
2.44 0.00 1.00 0.45
3.94 0.00 1.00 0.51
0.95 1.00 0.00 0.14
194 | P a g e
Chapter 6: Results
2.44 1.00 0.00 0.97
3.94 1.00 0.00 1.80
0.95 1.00 1.00 -0.44
2.44 1.00 1.00 0.36
3.94 1.00 1.00 1.15

Given that the evidence suggests the relationship between informational content and

contributing behaviour (shares) varies as a function of media richness, the next step is to

probe the interaction by estimating the conditional effect of informational content in

each of the three groups. In this model, the conditional effect of X (informational

content) on Y (contributing behaviour, shares) depends on M (media richness) as shown

in Table 6.38.

Table 6.38 Conditional Effect of Informational Content on Contributing


Behaviour (Shares) At Values of the Moderator

Status
Video /link Effect se t p LLCI ULCI
0.00 0.00 0.07 0.02 3.60 0.00 0.03 0.11
0.00 1.00 0.04 0.04 1.23 0.22 -0.03 0.11
1.00 0.00 0.56 0.18 3.13 0.00 0.21 0.90

In the Table 6.38, the first row corresponds to video=0 and status/link=0, which

therefore acts as the medium richness moderator group, i.e. the photo group. This output

indicates that the conditional effect of informational content on contributing behaviour

(shares) is 0.07 with a standard error of 0.02. This is statistically different from zero, t =

3.60, p <0.05, or between 0.03 and 0.11 with 95% confidence.

Therefore, it can be reported that two medium richness posts which differ by one unit in

informational content level are estimated to differ by 0.07 units in contributing (shares).

This is the slope of the line for ‘medium’ in Figure 6.1. The remaining conditional

effects are presented in Table 6.38 and correspond to the slopes of the lines for their

respective groups in Figure 6.1. As can be seen in the output in Table 6.38, all

195 | P a g e
Chapter 6: Results
conditional effects are positive however only the effect for photo (row 1), and video

(row 3) are significantly different from zero. The conditional effects are not equal, as

implied by the claim that media richness moderates the effect of informational content

level on contributing behaviour in the form of shares. The effects can be visualised by

the varying slopes of the lines in Figure 6.1.

Figure 6.1 A Visual Representation of the Moderation of the Effect of


Informational Content on Contributing Behaviour (Shares) by Richness

Figure 6.1 demonstrates that the slope of the line for video (high richness) is greater

than that of lower richness categories photo and status/link. For posts with high richness

(video), the effect of informational content on contributing behaviour is greater (0.56, p

= 0.05), than for posts with medium richness (0.07, p = <0.05).

For posts with more than 2 pieces of information, the most effective media type in order

to facilitate contributing behaviour in the form of shares is video. Comparatively, for

informational posts of 1 to 2 pieces information, medium richness (photo) is the most

effective mode of delivery in order to facilitate the occurrence of post sharing.


196 | P a g e
Chapter 6: Results
Media richness moderation summary

This section demonstrates that media richness moderates the relationship between

informational content and contributing behaviour in the form of shares. The effect of

informational content on contributing behaviour (shares) was greater with higher

richness (categorised as video posts in this study). This finding indicates that while a

significant relationship was found in previous sections between informational content

presence and likelihood that a social media post would be ‘shared’, the effect is greater

when the post is of higher richness, such as in the form of a video. There were no

significant moderation effects concerning informational content and other types of

SMEB. Similarly, media richness did not moderate the relationships between

entertaining, remunerative and relational social media content and SMEB.

H6: The strength of the relationship between social media content and social media

engagement behaviour is directly related to media richness is therefore partially

supported. Media richness moderates the relationship between informational content

and contributing behaviour (shares).

6.4.3 Congruity

The extent to which social media content is congruent or incongruent to the brand may

enhance or mitigate different forms of SMEB. It is proposed that entertaining,

informational, remunerative and relational posts may vary in their degree of congruity

with the brand.

Social media content with low congruity is categorised as content which is not

specifically related to the category (wine), brand or product. Content with medium

congruity is relevant to the product category but does not focus on a specific brand or

197 | P a g e
Chapter 6: Results
product. Social media content with high congruity includes details of the brand and a

specific product(s), and hence is highly contextually relevant to the social media brand

page. As the congruity variable has three categories (Table 6.39), Hayes PROCESS

Model 2 was applied in order to test the moderation effect.

Table 6.39 Congruity Operationalisation

Congruity Operationalisation
Low Post is not explicitly related to the category, the brand or a product.
Medium Post relates to the category (wine)
High Post relates to the brand and/or specific products

It is proposed that the congruity of social media content (low, medium and high)

moderates the relationships between social media content and SMEB as outlined in

Hypothesis 7;

H7: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to congruity.

Using the SPSS code to construct PROCESS Model 2 for a multi-categorical

moderating variable (Hayes, 2015), tests of the moderation of the effect of

informational, entertaining, remunerative and relational content on the various types of

SMEB by congruity were achieved. Among the tests of moderation, congruity was

found to moderate two relationships; the effect of informational content on contributing

behaviour in the form of likes, and the effect of entertaining content on creating

behaviour. The results of these moderation effects are presented in the following

section.

Moderation of the effect of informational content on contributing behaviour (likes) by


congruity

The results in Section 6.2.2, the results in Table 6.8 showed that 1, 2, 3, 4 and 5
198 | P a g e
Chapter 6: Results
elements of information within a post significantly and positively predict the occurrence

of contributing SMEB in the form of likes. Therefore, informational content levels of 1

to 5 are used as the independent variable in this model which tests the moderation of the

effect of informational content on contributing behaviour (likes) by congruity. Hayes

(2015) PROCESS model was applied. Table 6.40 shows the test of interaction between

congruity and contributing behaviour in the form of likes.

Table 6.40 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Likes) by Congruity

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (information x low congruity) 0.00 1.66 1.00 1743.00 0.20
int_2 (information x high congruity) 0.00 6.76 1.00 1743.00 0.01
Both 0.00 3.57 2.00 1743.00 0.03

The “Both” row in Table 6.40 provides a test of the null hypothesis that congruity does

not moderate the effect of informational content presence on contributing behaviour in

the form of likes. The null hypothesis is not supported, F (2, 1743) =3.5709, p <0.05.

This indicates that the regression slope quantifying contributing behaviour (likes) as a

function of informational content depends on congruity.

In order to visualise the moderation effect, the plot option within the PROCESS code

generates a table (Table 6.41) of values of the moderator (D1 congruity low and D2

congruity high), focal predictor (X, informational content) and estimated values of Y

(contributing behaviour, likes). The table contains estimates of contributing behaviour

in the form of shares in each of the three groups of informational content level. The

three groups are generated by PROCESS and defined as the sample mean (2.26) as well

as a standard deviation below (1.01) and a standard deviation (3.52) above the mean,

shown in the first column of Table 6.41.

199 | P a g e
Chapter 6: Results
Table 6.41 PROCESS Data for Visualising Conditional Effect of Informational
Content on Contributing Behaviour (Likes)

Information Level Congruity- High Congruity - Low Contributing (Likes)


1.01 0.00 0.00 1.72
2.26 0.00 0.00 1.94
3.52 0.00 0.00 2.17
1.01 0.00 1.00 1.74
2.26 0.00 1.00 1.89
3.52 0.00 1.00 2.03
1.01 1.00 0.00 2.05
2.26 1.00 0.00 2.11
3.52 1.00 0.00 2.17
1.01 1.00 1.00 2.08
2.26 1.00 1.00 2.05
3.52 1.00 1.00 2.03

Given that the evidence suggests the relationship between informational content and

contributing behaviour (likes) varies as a function of congruity, the next step is to probe

the interaction by estimating the conditional effect of informational content in each of

the three groups. In this model, the conditional effect of X (informational content) on Y

(contributing behaviour, likes) depends on M (congruity) as shown in Table 6.42.

Table 6.42 Conditional Effect of Informational Content on Contributing


Behaviour (Likes) At Values of the Moderator

Congruity Congruity
High Low Effect se t p LLCI ULCI
0.00 0.00 0.18 0.04 4.32 0.00 0.10 0.26
0.00 1.00 0.11 0.03 4.02 0.00 0.06 0.17
1.00 0.00 0.05 0.03 1.60 0.11 -0.01 0.10

In the Table 6.42, the first row corresponds to the absence of high congruity (0) and low

congruity (0) which therefore acts as the medium congruity moderator group. This

output indicates that the conditional effect of informational content of contributing

behaviour (likes) is 0.18 with a standard error of 0.04. This is statistically different from

zero, t = 4.32, p=0.05, or between 0.10 and 0.26 with 95% confidence. Therefore, it can

be reported that two medium congruity posts which differ by one unit in informational

content level are estimated to differ by 0.18 units in contributing (likes). This is the
200 | P a g e
Chapter 6: Results
slope of the line for ‘Medium Congruity’ in Figure 6.2. The remaining conditional

effects are presented in Table 6.42 and correspond to the slopes of the lines for their

respective groups in Figure 6.1.

As can be seen in the output in Table 6.42, all conditional effects are positive. However

only the effect for medium congruity (row 1), and low congruity (row 2) are

significantly different from zero. The conditional effects are not equal, as implied by the

claim that congruity moderates the effect of informational content level on contributing

behaviour in the form of likes. The effects can be visualised by the varying slopes of the

lines in Figure 6.2.

Figure 6.2 Visual Representation of the Moderation of the Effect of Informational


Content on Contributing Behaviour (Likes) By Congruity.

Figure 6.2 demonstrates that the slope of the line is greater for medium congruity (effect

= 0.18, p = <0.05) than that of low congruity (effect = 0.11, p= < 0.05). The finding

indicates that the relationship between informational content and contributing behaviour

in the form of likes is positively moderated by congruity. Therefore, posts with

201 | P a g e
Chapter 6: Results
informational content that are related to the category (medium congruity) have a greater

effect on contributing behaviour in the form of likes, compared to posts with

informational content that are not related to the category, product or brand (low

congruity).

At low levels of informational content (1 and 2), a high level of congruity coincides

with a greater amount of contributing behaviour in the form of likes occurring.

However, as the level of information within a post increases, medium and low congruity

has a greater effect on contributing behaviour in the form of likes.

Moderation of the effect of entertaining content on creating behaviour by congruity

Using the SPSS code to construct PROCESS Model 2 for the multi-categorical variable;

congruity (Hayes, 2015), a test of the moderation of the effect of entertaining content

level on creating behaviour by congruity was achieved. The level of entertaining content

used as the independent variable in this model is derived from the results in Section

6.2.4.

Table 6.43 shows the test of interaction between congruity and creating SMEB.
Table 6.43 Partial Output from PROCESS Model 2 Examining the Moderation of
the Effect of Entertaining Content on Creating Behaviour by Congruity

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (entertainment x low congruity) 0.0009 .9849 1 1129 .32
int_2 (entertainment x high congruity) 0.0045 5.1103 1 1129 .02
Both 0.0052 2.985 2 1129 .05

The ‘Both’ row in Table 6.43 provides a test of the null hypothesis that congruity does

not moderate the effect of entertaining content presence on creating SMEB. The null

hypothesis is not supported, F (2, 1129) =2.985, p = 0.05. In other words, the regression

slope quantifying creating behaviour as a function of entertaining content depends on

202 | P a g e
Chapter 6: Results
congruity.

In order to visualise the moderation effect, the plot option within the PROCESS code

generates a table (Table 6.44) of values of the moderator. The table contains estimates

of creating behaviour in each of the three groups of entertaining content. The three

groups are generated by PROCESS and defined as the sample mean (1.43) as well as a

standard deviation below (0.74) and a standard deviation (2.13) above the mean, shown

in the first column of Table 6.44.

Table 6.44 PROCESS Data for Visualising Conditional Effect of Entertaining


Content on Creating Behaviour

Entertaining Congruity - High Congruity - Low Contributing (Likes)


0.74 0.00 0.00 1.14
1.43 0.00 0.00 1.10
2.13 0.00 0.00 1.05
0.74 0.00 1.00 1.04
1.43 0.00 1.00 1.10
2.13 0.00 1.00 1.16
0.74 1.00 0.00 0.96
1.43 1.00 0.00 1.16
2.13 1.00 0.00 1.36
0.74 1.00 1.00 0.86
1.43 1.00 1.00 1.17
2.13 1.00 1.00 1.48

Given that the evidence suggests the relationship between entertaining content and

creating behaviour varies as a function of congruity, the next step is to probe the

interaction by estimating the conditional effect of informational content in each of the

three groups. In this model, the conditional effect of X (entertaining content) on Y

(creating behaviour) depends on M (congruity) as shown in Table 6.45.

203 | P a g e
Chapter 6: Results
Table 6.45 Conditional Effect of Entertaining Content on Creating Behaviour at
Values of the Moderator

Congruity Congruity
High Low Effect se t p LLCI ULCI
0.00 0.00 -0.07 0.13 -0.50 0.62 -0.32 0.19
0.00 1.00 0.08 0.07 1.13 0.26 -0.06 0.23
1.00 0.00 0.29 0.09 3.33 0.00 0.12 0.46

In Table 6.45, the first row corresponds to high congruity (0) and low congruity (0)

which therefore acts as the medium congruity moderator group.

The results indicate that the conditional effect of entertaining content on creating

behaviour is 0.29 with a standard error of 0.09 for high congruity posts. This is

statistically different from zero, t = 3.33, p=<0.05. This is the slope of the line for high

congruity in Figure 6.3.

Figure 6.3 Visual Representation of the Moderation of the Effect of Entertaining


Content on Creating Behaviour by Congruity.

204 | P a g e
Chapter 6: Results
Figure 6.3 demonstrates that the slope of the line is the most positive for high congruity

(effect = 0.29, p = <0.05). The finding indicates that the relationship between

entertaining content and creating behaviour is positively moderated by congruity.

Therefore, posts with entertaining content that are related to the brand or specific

product (high congruity) have a greater effect on creating behaviour. As more

entertaining content is presented (up to 5 elements) the effect on creating behaviour is

significantly increased for posts which are highly congruent to the focal brand.

However, when the post was medium congruity (i.e. related to the product category in

general) there is a negative effect on creating behaviour as the amount of entertaining

content within the post increases towards 5 elements.

Congruity moderation summary

Congruity was found to moderate the relationship between informational content and

contributing behaviour in the form of likes, and the relationship between entertaining

content and creating behaviour. There were no statistically significant results to indicate

that congruity moderates the relationships between informational and entertaining

content and other SMEBs. Further, there were no statistically significant results to

indicate that congruity moderates the relationships between relational and remunerative

content and any of the SMEB types.

The relationship between informational content and contributing behaviour in the form

of likes was positively moderated by congruity. The results suggest that posts with

informational content that are related to the category (medium congruity) have a greater

effect on contributing behaviour in the form of likes, compared to posts with

informational content that are not related to the category, product or brand (low

congruity). The relationship between entertaining content and creating behaviour was
205 | P a g e
Chapter 6: Results
also positively moderated by congruity. The results suggested that posts with

entertaining content that are related to the brand or product (high congruity) have a

greater effect on creating SMEB than those with medium or low congruity.

H7: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to congruity is therefore partially supported.

The relationship between informational content and contributing behaviour in the form

of likes, and the relationship between entertaining content and creating behaviour were

moderated by congruity.

6.4.4 Community Size

The final moderation effect that was tested on the relationships between social media

content and SMEB was community size. This study included community size as a

moderating variable, measured by the number of ‘fans’ on the brand page. Community

size was categorised as small (less than 1,500 fans), medium (1500-10,000 fans) and

large (over 10,000 fans).

Community size has been found to negatively impact the level of interactions between

individuals (Simmel, 1950). Participation in smaller communities results in stronger

interpersonal relationship and therefore a greater intention for engagement (Dholakia et

al., 2004). Social media users are therefore likely to be more connected to a smaller

brand community, resulting in a higher level of SMEB. It is proposed that the

community size (small, medium and large) moderates the relationships between social

media content and SMEB as outlined in hypothesis 8;

H8: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to community size.

206 | P a g e
Chapter 6: Results
Using the SPSS code to construct PROCESS Model 2 for a multi-categorical

moderating variable (Hayes, 2015), tests of the moderation of the effect of

informational, entertaining, remunerative and relational content on SMEB by

community size were achieved. Among the tests of moderation, community size was

found to moderate three relationships: entertaining content on contributing behaviour

(likes), entertaining content on contributing behaviour (shares) and entertaining content

on dormant behaviour. The results of these moderation effects are presented in the

following sections.

Moderation of the effect of entertaining content on contributing behaviour (likes) by


community size

The presence of entertaining content within social media posts was found to be a

statistically significant and positive predictor of contributing behaviour in the form of

likes (Section 6.2.3, Table 6.10). The odds ratio of 1.71 showed that posts which had

entertaining content were 1.7 times more likely to facilitate the occurrence of

contributing behaviour (likes), compared to posts with no entertaining content.

PROCESS Model 2 was again applied to test how this relationship is moderated by

community size. Table 6.46 shows the test of interaction between community size and

contributing behaviour (likes).

Table 6.46 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Likes) by Community
Size

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (entertaining x small) 0.00 2.38 1.00 1129.00 0.12
int_2 (entertaining x large ) 0.00 2.32 1.00 1129.00 0.13
Both 0.00 2.83 2.00 1129.00 0.05

The “Both” line in Table 6.46 provides a test of the null hypothesis that community size

does not moderate the effect of entertaining content on contributing behaviour (likes).
207 | P a g e
Chapter 6: Results
The null hypothesis can be rejected, F (2, 1129) =2.83, p =0.05. In other words, the

regression slope quantifying contributing behaviour (likes) as a function of entertaining

content depends on community size. This means that community size significantly

moderates the relationship between entertaining content and contributing behaviour

(likes).

In order to visualise the moderation effect, the plot option within the PROCESS code

generates a table (Table 6.47) of values of the moderator (D1 small community size and

D2 large community size), focal predictor (X, entertaining content) and estimated values

of Y (contributing behaviour, likes). The table contains estimates of contributing

behaviour in the form of likes at three groups of entertaining content. The three groups

are generated by PROCESS and defined as the sample mean (1.43) as well as a standard

deviation below (0.74) and a standard deviation (2.13) above the mean, shown in the

first column of Table 6.47.

Table 6.47 PROCESS Data for Visualising Conditional Effect of Entertaining


Content on Contributing Behaviour (Likes)

Entertainment Large Small Contributing (shares)


0.74 0.00 0.00 2.02
1.43 0.00 0.00 2.10
2.13 0.00 0.00 2.17
0.74 0.00 1.00 1.81
1.43 0.00 1.00 1.74
2.13 0.00 1.00 1.68
0.74 1.00 0.00 2.15
1.43 1.00 0.00 2.33
2.13 1.00 0.00 2.52
0.74 1.00 1.00 1.93
1.43 1.00 1.00 1.98
2.13 1.00 1.00 2.03

Given that the evidence suggests the relationship between entertaining content and

contributing behaviour (likes) varies as a function of community size, the next step is to

probe the interaction by estimating the conditional effect of entertaining content in each
208 | P a g e
Chapter 6: Results
of the three groups. In this model, the conditional effect of X (entertaining content) on Y

(contributing behaviour, likes) depends on M (community size) as shown in Table 6.48.

Table 6.48 Conditional Effect of Entertaining Content on Contributing Behaviour


(Likes) at Values of the Moderator

Large Small Effect se t p LLCI ULCI


0.00 0.00 0.11 0.05 2.23 0.03 0.01 0.20
0.00 1.00 -0.09 0.12 -0.78 0.44 -0.33 0.14
1.00 0.00 0.26 0.09 2.85 0.00 0.08 0.45

In Table 6.48, the first line corresponds to Large=0 and Small=0, which therefore acts

as the medium community size moderator group. This output indicates that the

conditional effect of entertaining content on contributing behaviour (likes) is 0.11 with a

standard error of 0.05. This is statistically different from zero, t = 2.23, p <0.05, or

between 0.01 and 0.20 with 95% confidence. Thus, it can be reported that two medium

community size posts which differ by one unit in entertaining content level are

estimated to differ by 0.11 units in contributing (likes). This is the slope of the line for

‘Medium’ in Figure 6.4. The remaining conditional effects are presented in Table 6.48

and correspond to the slopes of the lines for their respective groups in Figure 6.4.

As can be seen in the output in Table 6.48, all conditional effects are positive. However

only the effect for medium (row 1), and large (row 3) are significantly different from

zero. The conditional effects are not equal, as implied by the claim that community size

moderates the effect of entertaining content on contributing behaviour (likes). The

effects can be visualised by the varying slopes of the lines in Figure 6.4.

209 | P a g e
Chapter 6: Results
Figure 6.4 Visual Representation of the Moderation of the Effect of Entertaining
Content on Contributing Behaviour (Likes) by Community Size.

Figure 6.4 demonstrates that the slope of the line is greater for large community sizes

(effect = 0.26, p = <0.05) than that of medium community sizes (effect =0.11, p= <

0.05). As the amount of entertaining content within the post increases, the number of

likes on the post also increases. This effect is greatest for large community sizes, but is

also positive and statistically significant for medium community sizes.

This finding indicates that the provision of higher levels of entertaining content is not an

effective strategy for brands with smaller community sizes (less than 1,500 fans) in

order to facilitate an increase in the number of likes received on the post.

Comparatively, when the community size has more than 1,500 fans, providing higher

levels of entertaining content is a successful social media strategy in order to facilitate

the occurrence of contributing behaviour in the form of likes.

210 | P a g e
Chapter 6: Results
Moderation of the effect of entertaining content on contributing behaviour (shares) by
community size

Using the SPSS code to construct PROCESS Model 2 for the multi-categorical variable:

community size (Hayes, 2015), a test of the moderation of the effect of entertaining

content level on contributing behaviour in the form of shares by richness was achieved.

This allows interpretation of how the effect of entertaining content on contributing

(shares) behaviour is altered depending on the community size (small, medium or

large). The levels of entertaining content (1, 2 and 3) used as the independent variable in

this model is derived from the results in Section 6.2.4.

Table 6.49 shows the test of interaction between community size and contributing

behaviour in the form of likes.

Table 6.49 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Shares) by Community
Size.

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (entertaining x small) 0.01 5.85 1.00 1109.00 0.02
int_2 (entertaining x large) 0.00 3.65 1.00 1109.00 0.06
Both 0.01 5.76 2.00 1109.00 0.00

The “Both” line in Table 6.49 provides a test of the null hypothesis that community size

does not moderate the effect of entertaining content on contributing behaviour in the

form of shares. The null hypothesis is not supported, F (2, 1109) =5.76, p =< 0.05. In

other words, the regression slope quantifying contributing behaviour (shares) as a

function of entertaining content depends on community size.

In order to visualise the moderation effect, the plot option within the PROCESS code

generates a table (Table 6.50) of values of the moderator (D1 small community size and

D2 large community size), focal predictor (X, entertaining content) and estimated values

211 | P a g e
Chapter 6: Results
of Y (contributing behaviour, shares).

Table 6.50 PROCESS Data for Visualising Conditional Effect of Informational


Content on Contributing Behaviour (Shares)

Entertaining Community size (large) Community size (small) Contributing (shares)


0.79 0.00 0.00 0.76
1.39 0.00 0.00 0.86
1.98 0.00 0.00 0.96
0.79 0.00 1.00 0.73
1.39 0.00 1.00 0.61
1.98 0.00 1.00 0.49
0.79 1.00 0.00 0.93
1.39 1.00 0.00 1.21
1.98 1.00 0.00 1.48
0.79 1.00 1.00 0.90
1.39 1.00 1.00 0.96
1.98 1.00 1.00 1.01

Given that the evidence suggests the relationship between entertaining content and

contributing behaviour (shares) varies as a function of community size, the next step is

to probe the interaction by estimating the conditional effect of entertaining as shown in

Table 6.51.

Table 6.51 Conditional Effect of Entertaining Content on Contributing Behaviour


(Shares) At Values of the Moderator

Large Small Effect se t p LLCI ULCI


0.00 0.00 0.17 0.07 2.56 0.01 0.04 0.29
0.00 1.00 -0.20 0.14 -1.47 0.14 -0.48 0.07
1.00 0.00 0.46 0.14 3.29 0.00 0.19 0.74

The output in the first row of Table 6.51 indicates that the conditional effect of

entertaining content on contributing behaviour (shares) is .17 with a standard error of

0.07. This is statistically different from zero, t = 2.56 p= <0.05, or between 0.04 and

0.29 with 95% confidence. Therefore, it can be reported that two medium community

size posts which differ by one unit in entertaining content level are estimated to differ

by 0.17 units in contributing (shares). This is the slope of the line for ‘Medium’ in
212 | P a g e
Chapter 6: Results
Figure 6.5.

The remaining conditional effects are presented in Table 6.51 and correspond to the

slopes of the lines for their respective groups in Figure 6.5. As can be seen in the output

in Table 6.51, the conditional effects for medium and large community sizes are

positive and significantly different from zero. The conditional effects are not equal, as

implied by the claim that community size moderates the effect of entertaining content

on contributing behaviour (shares). The effects can be visualised by the varying slopes

of the lines in Figure 6.5.

Figure 6.5 A Visual Representation of the Moderation of the Effect of Entertaining


Content on Contributing Behaviour (Shares) by Community Size.

Figure 6.5 demonstrates that the slope of the line is positive and greater for large

community sizes (effect = 0.46, p = <0.05) is greater than that of medium community

sizes (effect = 0.17, p= < 0.05). The slope of the line is negative for small community

sizes. Therefore, posts with entertaining content that are in a large community have a
213 | P a g e
Chapter 6: Results
greater effect on contributing behaviour in the form of shares, compared to posts with

entertaining in a medium sized community. This finding is consistent with the results

depicted in Figure 6.4 which showed that an increase in entertaining content was

detrimental to the number of likes received on a post when the community size was

small. These findings indicate the importance of building larger community sizes (more

than 1,500 fans) in order to ensure that the provision of highly entertaining content will

result in the content being shared and liked by users. Small community sizes (less than

1,500 fans) do not appear to be as successful in their provision of entertaining content to

users, which reduces the number of shares and likes made on the content.

Moderation of the effect of entertaining content on dormant behaviour by community


size

Table 6.52 shows the test of interaction between community size and contributing

behaviour in the form of likes. Table 6.52 shows the test of interaction between

community size and dormant behaviour.

Table 6.52 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Dormant Behaviour by Community Size

R-square increase due to interactions R2-chng F df1 df2 p


int_1 (entertainment x small) 0.01 7.39 1.00 1129.00 0.01
int_2 (entertainment x large) 0.00 0.03 1.00 1129.00 0.86
Both 0.01 3.90 2.00 1129.00 0.02

The “Both” line in Table 6.52 provides a test of the null hypothesis that community size

does not moderate the effect of entertaining content dormant SMEB. The null

hypothesis is not supported, F (2, 1129) =3.9, p =< 0.05. The regression slope

quantifying dormant behaviour as a function of entertaining content depends on

community size.

In order to visualise the moderation effect, the plot option within the PROCESS code
214 | P a g e
Chapter 6: Results
generates a table (Table 6.53) of values of the moderator (D1 small community size and

D2 large community size), focal predictor (X, entertaining content) and estimated values

of Y (dormant behaviour).

Table 6.53 PROCESS Data for Visualising Conditional Effect of Entertaining


Content on Dormant Behaviour

Entertaining Content Large Small Dormant behaviour


0.74 0.00 0.00 0.89
1.43 0.00 0.00 0.89
2.13 0.00 0.00 0.88
0.74 0.00 1.00 0.87
1.43 0.00 1.00 0.89
2.13 0.00 1.00 0.90
0.74 1.00 0.00 0.91
1.43 1.00 0.00 0.91
2.13 1.00 0.00 0.90
0.74 1.00 1.00 0.89
1.43 1.00 1.00 0.91
2.13 1.00 1.00 0.92

Given that the evidence suggests the relationship between entertaining content and

dormant behaviour varies as a function of community size, the next step is to probe the

interaction by estimating the conditional effect of entertaining as shown in Table 6.54.

Table 6.54 Conditional Effect of Entertaining Content on Dormant Behaviour at


Values of The Moderator

Large Small Effect se t p LLCI ULCI


0.00 0.00 -0.01 0.00 -1.84 0.07 -0.01 0.00
0.00 1.00 0.02 0.01 2.20 0.03 0.00 0.04
1.00 0.00 -0.01 0.01 -1.13 0.26 -0.02 0.01

In the Table 6.54, the first line corresponds to Large=0 and Small=0, which therefore

acts as the ‘Medium’ community size moderator group. This output indicated that the

conditional effect of entertaining content on dormant behaviour is -0.01 with a standard

error of 0.00. This is statistically different from zero, t = -1.84, p <0.05, or between -

0.01 and 0.00 with 95% confidence. Therefore, it can be reported that two medium
215 | P a g e
Chapter 6: Results
community size posts which differ by one unit in entertaining content are estimated to

differ by -0.01 units in dormant behaviour. This is the slope of the line for ‘Medium’ in

Figure 6.6. The remaining conditional effects are presented in Table 6.54 and

correspond to the slopes of the lines for their respective groups in Figure 6.6. The

conditional effect for medium community size is negative, while the conditional effect

for small community size is positive and significantly different from zero. The

conditional effect for large community size (row 3) is not statistically significant, but

slopes in a negative direction as can be seen in Figure 6.6. The conditional effects for

small and medium community size are not equal, as implied by the claim that

community size moderates the effect of entertaining content on dormant SMEB. The

effects can be visualised by the varying slopes of the lines in Figure 6.6.

Figure 6.6 A Visual Representation of the Moderation of the Effect of Entertaining


Content on Dormant Behaviour by Community Size.

216 | P a g e
Chapter 6: Results
Figure 6.6 demonstrates that the slope of the line is positive for the moderator group

‘small community size’. The positive slope indicates that for small community sizes, as

the level of entertaining content within a post increases towards 5 elements, the

percentage of dormant behaviour that occurs among users’ increases. Comparatively,

the slope of the line for the moderator group ‘medium community size’ is negative

(effect = -0.01 p = <0.05). This indicates that in a medium community size, for every

unit increase in entertaining content level, the level of dormant behaviour exhibited

among users is reduced.

This finding demonstrates a clear advantage for brands with a large following on social

media, indicating that when the community size is higher (in this study, 10,000 fans or

greater), entertaining content reduces dormant SMEB. Comparatively, for brands with a

small social media following, entertaining social media content appears to increase the

occurrence of dormant SMEB.

Community size moderation summary

Community size was found to moderate the relationships between entertaining content

and contributing behaviour (likes), contributing behaviour (shares) and dormant

behaviour. There were no statistically significant results to indicate that community size

moderates the relationships between informational, relational and remunerative content

and SMEB.

The relationship between entertaining content and contributing behaviour in the form of

likes was moderated by community size. The results suggest that entertaining posts

within a large community size have a greater effect on the number of likes compared to

the same posts in a medium or small community size. Similarly, the moderation of the

217 | P a g e
Chapter 6: Results
effect of entertaining content on contributing behaviour (shares) by community size

indicated a greater effect for large community sizes.

The moderation of the effect of entertaining content on dormant behaviour

demonstrated a benefit for large and medium community sizes, with the presence of

entertaining content within a post decreasing the occurrence of dormant SMEB.

Comparatively, for small community sizes, entertaining content within a post was

associated with an increase in dormant behaviour. These results demonstrate a clear

benefit for large and medium community sizes with regards to posting entertaining

content. It appears that in larger community sizes, the use of entertaining content has a

greater effect on SMEBs including an increase in post likes and shares, and a decrease

in dormancy.

H8: The strength of the relationship between social media content and social media

engagement behaviour is inversely related to community size was not supported.

Instead, the relationship between social media content and SMEB was positively

moderated by community size.

6.4.5 Moderation Effect Summary

In this section the moderation of the effect of social media content on SMEB by media

richness, congruity and community size was presented. Hayes (2015) PROCESS Model

2 modification technique was used in order to test the effect of multi-categorical

moderating variables.

The media richness of a post was found to positively moderate the effect of

informational content on contributing behaviour in the form of shares. This finding

indicates that when presenting informational content, the use of a highly rich delivery

218 | P a g e
Chapter 6: Results
method (such as a video) is preferable in order to stimulate sharing of the content.

The congruity of a post was found to moderate the effect of informational content on

contributing behaviour in the form of likes. The finding indicated that when presenting

informational content, the use of moderately congruent information (related to the

category) is preferable in order to facilitate contributing behaviour in the form of likes.

Congruity positively moderated the relationship between entertaining content and

creating behaviour, with highly congruent posts (related to the specific product/brand)

having the greatest effect on creating behaviour.

Community size was the final moderating variable tested in this section. Community

size was found to positively moderate the relationship between entertaining content and

contributing behaviour (likes and shares) and creating behaviour. The positive

moderation indicates that within larger community sizes, the effect of entertaining

content on contributing behaviour is greater, and the occurrence of dormant behaviour is

reduced.

219 | P a g e
Chapter 6: Results

6.5 Summary of Results


Table 6.55 provides a summary of the results concerning each hypothesis of the study.
Table 6.55 Summary of Hypotheses and Results

H# Hypothesis Supported/Not
supported
H1 The presence of informational content facilitates passive, positively- Supported
valenced social media engagement behaviour.
H1a High levels of informational content weaken the relationship with Partially supported
positively-valenced social media engagement behaviours,
H2 The presence of entertaining content facilitates active, positively- Partially supported
valenced social media engagement behaviour.
H2a High levels of entertaining content weaken the relationship with Partially supported
positively-valenced social media engagement behaviours,
H3 The presence of remunerative content facilitates passive, positively- Not supported
valenced social media engagement behaviour.
H3a High levels of remunerative content weaken the relationship with Partially supported
positively-valenced social media engagement behaviours.
H4 The presence of relational content facilitates active, positively-valenced Supported
social media engagement behaviour.
H4a High levels of relational content weaken the relationship with positively- Supported
valenced social media engagement behaviours.
H5a The simultaneous presence of informational and entertaining content Not supported
facilitates negatively-valenced social media engagement behaviours.
H5b The simultaneous presence of informational and relational content Not supported
facilitates negatively-valenced social media engagement behaviours.
H5c The simultaneous presence of informational and remunerative content Not supported
facilitates positively-valenced social media engagement behaviours.
H5d The simultaneous presence of entertaining and remunerative content Not supported
facilitates negatively-valenced social media engagement behaviours.
H5e The simultaneous presence of entertaining and relational content Partially supported
facilitates positively-valenced social media engagement behaviours.
H5f The simultaneous presence of remunerative and relational content Not supported
facilitates negatively-valenced social media engagement behaviours.
H6 The strength of the relationship between social media content and social Partially supported
media engagement behaviour is directly related to media richness.
H7 The strength of the relationship between social media content and social Partially supported
media engagement behaviour is inversely related to congruity.
H8 The strength of the relationship between social media content and social Not supported
media engagement behaviour is inversely related to community size.

220 | P a g e
Chapter 6: Results

6.6 Chapter Summary


This chapter presented the results of the statistical analysis performed in order to test the

hypotheses developed in Chapter 4. Hypothesis 1, stating that the presence of

informational content facilitates passive, positively-valenced SMEB was supported.

There was partial support found for Hypothesis 1a which suggested that high levels of

informational content weaken the relationship with positively-valenced engagement

behaviours. Hypothesis 2 which suggested that the presence of entertaining content

facilitates active, positively-valenced SMEB was partially supported. There was partial

support for Hypothesis 2a which suggested that high levels of entertaining content

weaken the relationship with positively-valenced engagement behaviours. Hypothesis 3,

which predicted that the presence of remunerative content would facilitate passive,

positively-valenced engagement behaviour, was not supported. Instead, the presence of

remunerative content facilitated active, positively-valenced SMEB. H3a was partially

supported, suggesting that high levels of remunerative content weaken the relationship

with positively-valenced engagement behaviours. Hypothesis 4, which suggested that

the presence of relational content facilitates active, positively-valenced SMEB, was

supported. Hypothesis 4a was also supported, indicating that high levels of relational

content weaken the relationship with positively-valenced SMEB. There was partial

support found for one of the interaction effects, Hypothesis 5e, which stated that the

simultaneous presence of entertaining and relational content facilitates positively-

valenced engagement behaviours. Finally, partial support was found for the moderation

hypotheses regarding media richness (Hypothesis 6) and congruity (Hypothesis 7). In

the next chapter, these results are discussed in detail with ensuing theoretical and

practical implications.

221 | P a g e
Chapter 7: Discussion and Conclusion

CHAPTER 7. Discussion and Conclusion


7.1 Introduction
This chapter addresses the main contributions of this thesis. These contributions

include: the development of the SMEB construct, the establishment of an empirical

relationship between social media content and social media engagement behaviour, the

use of a novel approach to social media data analytics, and the application of the UGT

theoretical perspective to online engagement. There are some limitations to the research

which are addressed within this chapter. These limitations lead to valuable areas for

further research, which are outlined. The important managerial implications ensuing

from the results of this thesis, such as the point at which information overload impacts

on the processing of social media, the interaction effects of conflicting content types

and the approach to understanding social media data are discussed. This chapter closes

with the concluding thoughts.

7.2 Contributions of the Research


7.2.1 Development of the SMEB Construct
This thesis developed and tested a new construct through which the engagement

behaviour of users with social media content could be conceptualised, defined and

measured. The conceptual development of this construct was explained in Chapter 3 and

empirically tested in Chapter 6.

Proponents of customer engagement have commonly argued for a three dimensional

construct, with cognitive, affective and behavioural components (e.g. Brodie et al.,

2011). This thesis focused exclusively on furthering the understanding of one dimension

of engagement; behavioural engagement. Customer engagement behaviour has been

defined in previous literature as “customer’s behavioural manifestations that have a


222 | P a g e
Chapter 7: Discussion and Conclusion
brand or firm focus, beyond purchase” (van Doorn et al., 2010 p.254). Despite this

definition and preliminary investigation into the nature of engagement behaviours (E.g.

Jaakkola & Alexander, 2014), there still remains much to understand about these

behavioural manifestations in different contexts. Further, with the growing prevalence

of social media, there has been an emergent focus from both academics and

practitioners on the concept of engagement within social media platforms (Brodie et al.,

2013). Scholars have agreed that social media platforms provide users with an

interactive avenue to create value and engage with the firm (Brodie et al., 2013;

Gummerus et al., 2012). However, prior to this research, there was little theoretical and

empirical evidence to explicate the specific user engagement behaviours that may occur

in a social media setting. Through the development of the SMEB construct, this thesis

identified exactly how social media engagement behaviour is manifested. In doing so,

six specific and discrete engagement behaviours were identified and defined. These

social media engagement behaviours contribute to engagement literature by

incorporating the important facets of engagement behaviour including valence

(Hollebeek & Chen, 2014; van Doorn et al., 2010) and engagement intensity (Malthouse

et al., 2013; Muntinga et al., 2011).

The development of the SMEB construct and corresponding six behaviours enhances

the understanding of engagement behaviours by theoretically and empirically

demonstrating the occurrence of engagement, which varies by intensity. Previous

scholars such as Muntinga et al., (2011) and Malthouse et al., (2013) have theorised that

engagement within social media may exist in low and high levels of intensity; however

this notion had not been empirically developed and tested. Through the development of

the SMEB construct within this research, six different types of engagement were

223 | P a g e
Chapter 7: Discussion and Conclusion
defined and measured. It is argued that simple categorisations of low and high

engagement (e.g. Malthouse et al., 2013) or levels of engagement (e.g. Muntinga et al.,

2011) are limited in their ability to fully understand the nature of engagement

behaviour, particularly within social media platforms. The SMEB construct incorporates

lower intensity and more passive engagement behaviours, such as dormancy and

consuming. Additionally, it recognises more active engagement behaviours with a

moderate intensity; detaching and contributing. Finally, the construct demonstrates the

occurrence of highly active engagement behaviour such as creation and destruction,

which represents behavioural engagement that impacts on others in the community as

well as on interaction with the brand.

In addition to enhancing the understanding of engagement behaviour through the

conceptualisation and measurement of various engagement intensity levels, the SMEB

construct provides an important contribution through its integration of both positively-

and negatively-valenced manifestations of engagement. A majority of the research

conducted on the engagement concept has focussed on the specific positively-valenced

expressions of engagement behaviour (Hollebeek, 2011b; Sprott, Czellar, &

Spangenberg, 2009; Vivek et al., 2012) . In the social media setting, scholars have also

theorised that engagement behaviours reflect positive user experiences (Muntinga et al.,

2011). This thesis contributes to the literature concerning negatively-valenced

engagement by empirically demonstrating that engaged users may experience

negatively-valenced engagement in addition to positively-valenced engagement

behaviour, both at various levels of intensity. Negatively-valenced engagement

behaviours of detaching and destructing were conceptualised and examined within this

thesis.

224 | P a g e
Chapter 7: Discussion and Conclusion
This development of the SMEB construct provides a further contribution through its

focussed level of investigation regarding the singular touch point of social media

content. Recent focus on the engagement construct in marketing has centred on

customer engagement with a brand (Hollebeek, 2011b; van Doorn et al., 2010).

Customers engage with a firm or brand through multiple touch-points and service

encounters (Vivek, Beatty, Dalela, & Morgan, 2014). Examples of this engagement

include interactions with staff, use of products, physical retail spaces, social media

pages and other forms of communication (Vivek et al., 2014). Authors have recognised

that there are various focal objects of customer engagement including product or service

offerings (Brodie et al., 2011), activities and events (Vivek et al., 2012) and media

(Calder et al., 2009). Together these interactions constitute the brand experience of the

customer. Engagement is interactive and therefore context-dependent and can only be

properly understood through an examination of each of these service experiences

(Brodie et al., 2011; Calder et al., 2009; Gummerus et al., 2012). However, there is little

research that examines customer engagement at this focused level. The examination of

social media engagement provided within this thesis focuses attention on a singular

touch-point in the service experience. Consistent with calls from previous researchers

(Brodie et al., 2011; Vivek et al., 2012), this in-depth examination within a context-

specific environment (e.g. social media) provides greater insight into the behavioural

manifestations of engagement.

7.2.2 Application of the UGT Perspective to Engagement


This thesis contributes to the body of literature concerning antecedents to engagement

by empirically demonstrating how social media content may act as a driver of user

decisions to engage, behaviourally, with social media content. This confirms the notion

225 | P a g e
Chapter 7: Discussion and Conclusion
of UGT which suggests that users are motivated by specific needs to actively select and

determine the content which they engage with in order to satisfy these needs. UGT was

shown to be an appropriate theoretical lens through which engagement can be explored

further, as it offers an insight into why and how individuals actively seek out and use

specific media to satisfy specific needs (Katz & Foulkes, 1962). Through the adoption

of the UGT perspective, this thesis explained how users are free to interact and engage

with specific types of content, as determined by their motivations, needs and

gratifications sought. UGT provides this research with an important theoretical

explanation of why and how individual users interact and engage with various forms of

social media content.

This research is one of only a handful of studies to apply UGT in a social media setting

(Chen, 2011; Ham, Lee, & Lee, 2014; Whiting & Williams, 2013). While UGT has

been applied in the context of television and electronic bulletins (Leung & Wei, 2000)

the rapid growth of the internet and social media platforms has created mediums in

which a higher level of interactivity from users is required (Ko et al., 2005; Ruggiero,

2000). This research demonstrates the importance of UGT and how this theory can be

evolved to consider prevalent emerging mediums such as social networking sites.

The constructs based on the motivations inherent in the UGT perspective, including the

need for social interaction, the need for entertainment, information seeking and sharing

needs, and the desire for reward or remuneration were adopted in this thesis in order to

develop the four categories of social media content positioned as an antecedent to

SMEB as depicted in the conceptual model developed within Chapter 4. Social media

content was categorised into four main groups, based on its level of information (Cvijikj

& Michahelles, 2013; De Vries et al., 2012), entertainment (De Vries et al., 2012;

226 | P a g e
Chapter 7: Discussion and Conclusion
Taylor et al., 2011), remunerative (Cvijikj & Michahelles, 2013; Lee et al., 2013) and

relational (Muntinga et al., 2011) content. Through empirical testing of the conceptual

model, it was shown that the delivery of these varying forms of content does effect

customer motives for social media use, hence resulting in various expressions of social

media engagement behaviours.

7.2.3 Establishment of the Relationship between Social Media Content and Social
Media Engagement Behaviour

This thesis demonstrates an empirical relationship between social media content

categories and SMEB. An association between content and user actions such as ‘liking’

and ‘commenting’ is discussed in previous literature, with conflicting and limited

empirical support (Cvijikj & Michahelles, 2013; De Vries et al., 2012; Lee et al., 2013).

In social media, it has been suggested that a brand’s overt goal is to attract an audience

by providing value, or gratification, through its content (Malthouse et al., 2013). This

thesis demonstrates an empirical relationship between social media content presence,

levels, and interactions, and SMEBs, demonstrating that content should be designed in a

way which encourages individual consumers to exhibit a greater level of engagement

(Malthouse et al., 2013). The determination of this link between content and

engagement provides fruitful avenues for further research, discussed later in this

chapter.

This thesis examined the role of each content type including analysis of the effect of the

presence of the content, the level of the content, interaction of the content with other

content types, and the moderating variables which affected the resultant expressions of

SMEB. The presence of informational, entertaining, remunerative and relational content

within social media posts was found to influence the occurrence of positively-valenced,

227 | P a g e
Chapter 7: Discussion and Conclusion
neutral and negatively-valenced SMEBs. These findings were dependent on the level of

social media content present, demonstrating an information overload effect. Further,

concepts of media richness, congruity of the social media content and the community

size in which the content were posted had significant effects on the relationships

between social media content and SMEB.

The presence of informational content within social media posts was found to predict

the occurrence of positively-valenced SMEBs. This finding was consistent with

previous literature which has stated that internet users increase their usage patterns as a

result of content gratifications such as information seeking, knowledge and learning

(Stafford et al., 2004). Similarly, the presence of entertaining content within a social

media post significantly predicted the occurrence of positively-valenced SMEBs

occurring. This finding is consistent with Cvijikj and Michahelles (2013) who

demonstrated that entertaining content was a significant factor in increasing the number

of likes, comments and shares made on social media posts. A significant and positive

relationship between remunerative content and active, positively-valenced SMEBs was

also identified. The presence of relational content had a significant effect on all

positively-valenced SMEBs. However the likelihood of passive, positively-valenced

SMEB in the form of consuming was the greatest. This denotes a passive participation

among users, which contrasts with Park et al. (2009) who found that socialising motives

predict active participation among users. Similarly, scholars have suggested that the

gratification of social and interaction motives causes users to create online content. The

findings in this thesis show that individuals are far more likely to passively and

individually consume this content, rather than create new content.

228 | P a g e
Chapter 7: Discussion and Conclusion
These results were explored in greater detail through assessing the specific amount of

content within the social media post. Supported by the notion of information overload,

this thesis enhances the understanding of the relationship between social media content

and SMEB. The findings demonstrate that for each type of content, the positive

relationship with the prediction SMEB occurring only exists at lower levels of the

content. At high levels of social media content, regardless of its informational,

entertaining, relational or remunerative nature, there were no significant effects on the

occurrence of SMEB. This is a very important finding, empirically demonstrating that

the specific amount of content delivered to users through social media posts is an

essential consideration, and has a significant impact on the resultant engagement

behaviour of users. Social media users do experience information overload, as has been

suggested for users of traditional media audiences (Eppler & Mengis, 2004; Hiltz &

Turoff, 1985).

This thesis also showed the importance of the concepts of media richness, congruity of

the social media content, and community size in determining the relationship between

social media content and SMEB. A high level of media richness (social media content

delivered in the form of a video) was found to be beneficial to the relationship between

informational social media content and contributing SMEB. As the level of

informational content delivered to users increased, high media richness levels were

preferable in order to facilitate an increase in users contributing behaviour. Media

richness did not have a significant effect on the relationships between the other three

types of social media content; entertaining, remunerative and relational, and SMEB.

The congruity of a post was found to moderate the effect of informational content on

contributing behaviour in the form of likes. The finding indicated that when presenting

229 | P a g e
Chapter 7: Discussion and Conclusion
informational content, the use of moderately congruent information (related to the

category) is preferable in order to facilitate contributing behaviour in the form of likes.

Congruity positively moderated the relationship between entertaining content and

creating behaviour, with highly congruent posts (related to the specific product/brand)

having the greatest effect on creating behaviour. This finding demonstrates the benefit

of providing congruent information, which is contrary to previous research which

suggested that incongruent information is more beneficial, leading to greater brand

recall (Hastie, 1980), enhanced arousal (Gardner et al., 1985), curiosity and interest

(Muehling & Laczniak, 1988), and increased message involvement (Lee, 2000).

Community size was the final moderating variable tested in this section. Community

size was found to positively moderate the relationship between entertaining content and

contributing behaviour (likes and shares) and creating behaviour. The positive

moderation indicates that within larger community sizes, the effect of entertaining

content on contributing behaviour is greater, and the occurrence of dormant behaviour is

reduced.

7.2.4 Social Media Data Analytics


This thesis adopted a novel approach to data collection and analysis to determine the

empirical relationships between social media content and SMEB. This thesis used data

from the social media brand pages of twelve Australian wine brands. This allowed

collection of actual behavioural data of 54,069 social media users, who were current

‘fans’ of the twelve wine brand pages. This rich source of data provided valuable

insights into the social media engagement behaviours demonstrated by actual Facebook

users, as opposed to self-reported data.

230 | P a g e
Chapter 7: Discussion and Conclusion
This thesis used two tools, Facebook Insights and NCapture to extract the behavioural

data from the twelve Facebook pages. The use of these tools shows how this data can

empirically enhance the understanding of actual SMEB actions. The metrics available

through these tools provide a rich and comprehensive insight into social media

engagement behaviour of users, compared to the limited measures used in previous

literature, such as the mere number of likes, shares of comments used as engagement

indicators by De Vries et al. (2012) and Cvijikj and Michahelles (2013).

While there is a vast quantity of social data available, the consequent challenge is to be

able to analyse the large volumes of user-generated content in order to gain meaningful

insights into the behaviour, opinions, sentiments, issues and trends among users

(Leskovec, 2011; Stieglitz & Dang-Xuan, 2013). This thesis contributes to knowledge

regarding the analytics of social media data through the use of three methods;

quantitative content analysis (QCA), binary logistic regression and Hayes (2013)

process analysis.

QCA is a suitable technique for analysing large quantities of social media data, as it

allows for a non-intrusive research method incorporating examination of a wide range

of data over an extensive time period (Neuendorf, 2002). QCA within social media

allowed inferences to be made about the social media content, including coding and

categorisation of all social media content according to the developed types of

informational, entertaining, remunerative, and relational content. Additionally, QCA

was used to make inferences about the audiences of content, or social media users

through the investigation of their relevant expressions of SMEB. This thesis provides a

detailed and sequential process for conducting QCA with social media data.

231 | P a g e
Chapter 7: Discussion and Conclusion
Marketing practitioners and academics are faced with the challenge of engineering

content effectively in order to facilitate engagement within social media platforms

(Malthouse et al., 2013). This thesis contributes to the understanding of how to

categorise, code and measure content through the development of the process for

automatically coding the presence of content attributes and elements. The codes and

corresponding dictionaries, shown in Chapter 5, enabled dichotomous measurement of

social media content categories, indicating the presence or non-presence of key terms,

words and phrases. These dictionaries could be used or adapted for future studies which

wish to similarly mechanically search for and code key terms within large quantities of

text-based content.

7.3 Limitations
The first limitation of this thesis is in regards to the social media content categories

determined. Through the process of QCA and application of the UGT perspective, four

types of social media content were categorised. The four categories of social media

content were derived through the literature review concerning the main gratifications

sought by users as per the UGT perspective. There are factors beyond the user

gratifications of the need for information, entertainment, remuneration and relational

interaction which were not included in this study. Factors such as personal identity

(Calder et al., 2009), affection, instrumentality, psychological reassurance,

fashion/status, mobility (Leung & Wei, 2000), relaxation, coordination for business,

status seeking (Grellhesl & Punyanunt-Carter, 2012), the need to vent negative feelings,

personal recognition (Leung, 2013), escape (Leung & Wei, 2000) and sensation seeking

(Zuckerman, 1979) have been suggested in UGT and related research, which could be

incorporated into further studies predicting user motivations for SMEB. Integration of a

232 | P a g e
Chapter 7: Discussion and Conclusion
richer collection of user motivations for social media engagement derived through UGT

would add value to the theory and further enhance knowledge regarding the individual

drivers of social media, and other online engagement behaviours. In order to achieve

this, a mixed methods approach to the research design may be required, including a

qualitative approach to more comprehensively identify, understand and explicate

individual motivations for online engagement.

The second limitation of thesis concerns the adoption of an exclusively behavioural

perspective of engagement. In order to more fully encapsulate the notion of social

media engagement, incorporation of the widely accepted dimensions of emotional and

cognitive customer engagement (Brodie et al., 2011) is required. Customer engagement

has been defined as a multidimensional concept subject to a context- and/or

stakeholder-specific expression of relevant cognitive, emotional and/or behavioural

dimensions (Brodie et al., 2011). Hence, the application of an exclusively behavioural

investigation of social media engagement within this thesis fails to consider users

expressions of relevant cognitive and emotional dimensions of engagement. The extent

to which social media content can facilitate users’ cognitive and emotional engagement

experiences provides a valuable area of further investigation. For example, it was found

that only informational and relational content facilitate consuming behaviour. It would

be interesting to further investigate whether this consuming behaviour influences the

amount of cognitive and affective engagement among users. It remains unknown

whether the expressions of SMEB in this thesis would occur simultaneously with users’

cognitive and affective aspects of engagement. One may posit that utilitarian content

such as informational and remunerative may facilitate cognitive customer engagement,

whereas the provision of hedonic content such as entertaining or relational content may

233 | P a g e
Chapter 7: Discussion and Conclusion
cause users to experience affective engagement. These questions pose interesting areas

for future research incorporating the multi-dimensional view of customer engagement.

Thirdly, within the data analysis process of the research, the method adopted did not

consider the complex nature and effect of the post content and engagement algorithms

that are determined by Facebook. Dependent on the algorithm employed by Facebook,

it is possible that certain posts types and post content characteristics are allocated more

‘weight’ and hence delivered to more users news feeds. For example, it has been

speculated that the Facebook news-feed algorithm is designed to ensure that when

brands post videos, they are rewarded with enhanced organic reach as opposed to the

use of status updates and photos (McGee, 2013). The extent to which social media

algorithms determine the content that consumers are exposed to within their news feed

or home pages is an important consideration for academic research within the social

media space. This research controlled for this effect to an extent, by considering the

engagement of users as a percentage of the total users who are exposed to the content

(measured as post reach).

It was identified in Chapter 5 that a very small number of destructing behaviour cases

were present within the data set. The number of cases in which destructing behaviour

occurred was very low at just 8 cases within the total of 2,236 posts. For this reason, the

destructing behaviour component of SMEB did not meet the requirements for the

dependent variable case size required within binary logistic regression (Harrell,

2013).This finding, along with the low mean scores for detaching behaviour indicates

that in the context of this research, the occurrence of negatively-valenced SMEB is less

prevalent. As a result, the binary logistic regression results presented in Chapter 6 did

not include testing of the effect social media content on destructing SMEB. In a more
234 | P a g e
Chapter 7: Discussion and Conclusion
controversial product category, it is likely that a greater amount of active, negatively-

valenced SMEB in the form of destruction may occur, warranting an interesting area for

further investigation.

A final main limitation of this thesis concerns the process adopted for the quantitative

content analysis of social media data. Following Neuendorf’s (2002) QCA process, the

development of the custom dictionaries was conducted by the researcher. Through the

development of custom dictionaries applied to word search formulas to identify key

characteristics of message content (see Chapter 5, Section 5.5.3), a highly mechanical

approach to social media content coding was conducted. This does not take into account

nuances within the language used in the social media content. Further, it does not

account for the full context and nature in which key words or phrases are used. Human

coding was also required for the categorisation of social media content in the form of

photographs and images as outlined in Chapter 5, Section 5.5.3. This process is also

subject to human error and misattribution of certain photos to incorrect social media

content categories.

7.4 Directions for Future Research


7.4.1 User Progression through SMEB

A valuable area for further research concerns the dynamic nature of the levels of

engagement behaviour in the SMEB construct. For example, analysis of user

progression through, or within the six behaviours is not considered. This could be

captured more accurately in future research through a longitudinal research design

which would provide a more comprehensive view of the development of SMEBs over

time. This research would identify the phases of SMEB through which users move over

time. For example, interesting research questions could address whether new members
235 | P a g e
Chapter 7: Discussion and Conclusion
of social media brand pages exhibit active, positively-valenced SMEB such as creating,

before transitioning to a phase of lower intensity SMEB such as consuming or

dormancy, comparative to older or existing members. Previous research has theorised

that the customer engagement process may vary for existing, compared to new

customers of a brand (Bowden, 2009), warranting further empirical investigation.

Examination of these user engagement behaviours across time would allow marketing

practitioners and academics to pin-point time periods in which users transition across

various intensities or valences of SMEB. This would provide strategic direction for

practitioners wanting to encourage consumers to transition from a passive state of

engagement to an active state, or from a negatively-valenced expression of engagement

to a positively-valenced expression. The extent to which the levels proposed occur in an

interactive, cyclical or unpredictable pattern warrants further research.

The notion of user progression through various stages of SMEB has parallels with

relationship marketing literature (Grönroos, 2004), particularly with regard to concepts

such as customer bonding (Szmigin, Canning, & Reppel, 2005), relationship ending

(Duck & Perlman, 1985), and customer exit processes such as relationship termination,

withdrawal, dissolution, discontinuation, uncoupling and break-up (Stewart, 1998).

Such bodies of literature can contribute to the understanding of SMEB. Further,

integration of relevant literature concerning customer loyalty (Gummerus, Liljander,

Pura, & Van Riel, 2004) would be valuable in further research exploring user

progression through stages of SMEB.

7.4.2 Identification of Further Antecedents to SMEB

The second area warranting further academic scrutiny arising from this thesis concerns

236 | P a g e
Chapter 7: Discussion and Conclusion
the drivers of SMEB. Firstly, the extent to which SMEBs are caused by factors beyond

social media content warrants further investigation. This thesis shows that SMEBs are a

consequence of the provision of social content. However, SMEB is likely to be caused

by a range of factors beyond social media content. For example, customer expressions

of negatively-valenced SMEB may arise as a result of a brand-related experience

outside of the social media platform. A negative product or service experience may

drive customers to create destructive content within social media platforms in order to

disseminate their dissatisfaction. Theoretical and empirical investigation of the factors

beyond social media content which may facilitate positively- and negatively-valenced

engagement behaviour will further the understanding of SMEB. Throughout recent

customer engagement literature, scholars have proposed a range of antecedents to

engagement related to the individual state of consumers, including identification

(Nambisan & Baron, 2007; Tsai et al., 2012), identity (Eisenbeiss et al., 2012; van

Doorn et al., 2010), hedonism (Gambetti et al., 2012; Nambisan & Baron, 2007),

interaction, (Hollebeek, 2011b), rapport (Hollebeek, 2011b), satisfaction (van Doorn et

al., 2010) and trust (Bowden, 2009). Future research could encapsulate these proposed

antecedents and empirically test their impact on SMEB in order to generate a more

comprehensive understanding of consumer motives for engaging with brands in the

social media forum.

The development of a more comprehensive conceptualisation of social media content

would also provide further insight into the relationship between content and SMEB.

Further examination of content gratifications beyond those tested within this thesis

(informational, entertaining, remunerative and relational) and their subsequent effect on

SMEB would enhance knowledge regarding the UGT perspective and its ability to

237 | P a g e
Chapter 7: Discussion and Conclusion
explain user motives to engage with social media content. For example, Leung (2009)

theorised that users are also motivated by uses and gratifications such as

instrumentality, psychological reassurance, fashion/status and mobility. Moreover,

Grellhesl and Punyanunt-Carter (2012) propose motives including relaxation, escape

and status seeking. Motivations to engage online may also include ‘personal identity’,

which involves individuals finding reinforcement for personal values, finding models of

behaviour, and gaining insight into one’s self (Calder et al., 2009). The extent to which

these more personal and intrinsic needs may drive online engagement behaviour

provides an interesting platform for further investigation. The categorisation of user

motives and resultant coding of social media content within this thesis encapsulated the

four main gratifications observed in recent literature stemming from the UGT

perspective. However, further research will provide a more detailed categorisation of

social media content through incorporation of additional content gratifications.

Moreover, social media gratifications sought by online users may vary depending on the

time of day, day of the week or time of the year. Within this thesis, it was identified that

social media users were more likely share and like social media content in the

afternoon. Social media users were also more likely to engage with social media content

if it was not posted on Thursdays and Sundays. Further research should investigate how

the gratifications sought by social media users may vary across time frames and hence

effect their expressions of engagement.

7.4.3 Investigation of SMEB Consequences

Future research should also investigate the consequences of SMEB. The extent to which

various types of SMEB exhibited by users results in outcomes such as future purchase

intention and behaviour, brand loyalty, word-of mouth and satisfaction would add

238 | P a g e
Chapter 7: Discussion and Conclusion
substantial value to the body of research concerning engagement. Previous engagement

scholars have theorised a number of consequences of engagement, focussing commonly

on positive outcomes such as loyalty (Bowden, 2009; Brodie et al., 2011; Hollebeek,

2011a), customer value (Vivek et al., 2012), word of mouth (Vivek et al., 2012) and

product innovation (Hoyer et al., 2010; Sawhney et al., 2005). Future research is

recommended in order to explore the consequences that may emerge as a result of

various expressions of SMEB. For example, research questions could include an

analysis of the relationship between user expressions of negatively-valenced SMEB and

consequences such as the dissemination of offline word of mouth, dissatisfaction with

the brand, product or social media brand page, and future purchase intentions.

Determination of these consequences with respect to both positively- and negatively-

valenced expressions of SMEB would add significant weight to the importance of

understanding how marketing practitioners can mitigate or neutralise negatively-

valenced SMEB, and enhance positively-valenced SMEB.

7.4.4 Incorporation of the Three Dimensional View of Customer Engagement

Another valuable area for further investigation concerns the adoption of the three

dimensional view of customer engagement. This thesis focused exclusively on the

behavioural manifestation of engagement. Future research should focus on theorising

and examining the influence of social media content on cognitive and affective

engagement. This would provide a more holistic view and comprehensive

understanding of the overall engagement attributed to social media content, and

consequently could be a better predictor of future behaviour. While this thesis provides

an in-depth investigation of the behavioural dimension of engagement, the cognitive

and affective dimensions of engagement require an equally detailed investigation. This

239 | P a g e
Chapter 7: Discussion and Conclusion
would provide a far more comprehensive insight into the overarching customer

engagement concept.

7.5 Managerial Implications


As the data used for this thesis was extracted from the Facebook profiles of Australian

wine brands, the results have use for managers in similar settings. While implications

can be drawn for managers in different environmental settings, further investigation is

required before the results are generalised.

7.5.1 High Level of Dormancy and Low Engagement Rates among Users
This thesis highlights an important issue to managers, with regards to the low average

engagement rate and corresponding high dormancy rate identified. The development of

the SMEB construct includes the important recognition and measurement of the neutral,

inactive state of engagement termed ‘dormancy’. Through the use of Facebook data,

this study implemented a formula through which to calculate the percentage of users

who exhibit dormancy. This process is explained in Chapter 5, Section 5.5.3. The

results of the study demonstrated a very high level of dormancy among social media

users. The mean score for dormancy was 90%. This result is consistent with the

concerns of previous scholars such as Nelson-Field and Taylor (2012) who have

suggested that many brands fail to stimulate high levels of engagement within social

media platforms.

This finding indicates that social media managers have not been successful in

strategically engineering their content in order to stimulate high levels of engagement

amongst their users. The mean score for organic reach was 19.5%, which represents the

percentage of the total number of users that the post reaches, when there is no paid

240 | P a g e
Chapter 7: Discussion and Conclusion
sponsorship or promotion of the post. These figures are important for managers to

understand, as they can provide benchmarking figures through which future research

and engagement rates can be compared.

To illustrate, for a sample brand of 5000 fans, the average organic reach of 19.5%

would result in a social media post reaching 975 fans. This thesis shows that on average,

10% of users who see a post are expected to engage with the post. In this example, that

would be just 97 users, out of a total of 5000 fans. The data within this thesis also shows

that the average number of times that a post would be commented on is 2 times. This

means that on average, less than 1% of users who see the Facebook content will make a

comment on the content. The results showed that posts receive on average 22 ‘likes’

(3% of reached users). On average posts are shared 2 times (less than 1% of reached

users), within a maximum of 105 shares achieved.

Managers should take these figures in to consideration when assessing their engagement

rates. In order to enhance the reach of content and resultant engagement, managers

should recognise the need to invest marketing budgets into digital campaigns. Within

Facebook, this can include utilising options within Facebook such as paying to boost

posts in order to reach a greater portion of the audience, promoting a specific call to

action, or promoting the page through paid Facebook advertising campaigns.

7.5.2 Enhancing Engagement through Strategic Content Design

For managers and designers of social media content, this thesis provides important

implications regarding the strategic design and delivery of social media content. The

findings show that informational, entertaining, remunerative and relational social media

content have a relationship with the SMEB of users.

241 | P a g e
Chapter 7: Discussion and Conclusion
Social media content selection

The results show that the provision of informational content within a post can

significantly increase the odds of users engaging in positively-valenced engagement

behaviours such as creating, contributing (shares and likes) and consuming. By placing

informational content within a social media post, consuming behaviour is most likely to

occur. Managers and designers of social media content are advised that if informational

content is delivered, the amount of information is limited to seven or fewer elements in

order to inhibit users’ experiencing information overload. If managers seek to increase

the number of likes on a post, they are advised to provide specifically five elements of

information within the post, which would increase the odds of users liking the post by

over 22 times. When the amount of informational content within a post reaches eight

elements, users are over seven times more likely to detach from the content.

Entertaining content was also found to significantly predict the occurrence of user’s

positively-valenced SMEBs of creating and contributing. There was no evidence to

suggest that if managers provide entertaining social media content, users consuming

behaviour will change. The findings also demonstrated the importance of the

consideration of how much entertaining content should be delivered to users. The

amount of entertaining content provided within a post should be carefully considered

with respect to the type of SMEB desired. No more than four elements of entertaining

content should be included when managers wish to facilitate the occurrence of creating

behaviour. The level of entertaining content within a post can also assist managers to

mitigate the occurrence of inactive engagement behaviour in the form of dormant

behaviour. Providing one or two elements of entertaining content within a post

significantly decreases the likelihood that users will remain dormant.

242 | P a g e
Chapter 7: Discussion and Conclusion
The provision of relational content had some of the strongest effects on predicting

positively-valenced SMEB. It is advised that in order to increase the likelihood of users

consuming content, relational content should be included. This effect is conditional on

the level of content, and hence managers should carefully engineer relational content to

avoid information overload which diminishes the positive effects. No more than two

elements of relational content should be included in a post in order to facilitate

consuming behaviour. If managers wish to increase the number of times users comment

on a post, it is advised that they include five elements of relational content. If managers

wish to increase the likelihood that a post will be shared through the use of relational

content, the content should be limited to a maximum of three relational elements.

The final type of social media content was remunerative content, which includes the

provision details about sales, promotions, prices and exclusive deals to social media

users. The findings show a positive relationship between providing this type of content

and active engagement behaviours of creating and contributing. Users are likely to

comment, like or share a post for the ‘chance to win’. However, there is no evidence

that they are actually consuming or processing the content. Nevertheless, it is likely that

remunerative content will be shared and liked amongst users. The amount of

remunerative content provided should be carefully considered by social media content

designers. While users do respond positively to remunerative content containing deals,

discounts, prices and promotions, if this content increases from just one to two

elements, users will have a negative response and actively detach from the content.

Combining various types of social media content

Combining entertaining content with informational content was found to place

conflicting demands on users, which significantly decreased the odds of positively-


243 | P a g e
Chapter 7: Discussion and Conclusion
valenced engagement behaviour occurring. Additionally, the simultaneous presence of

informational content and entertaining content will increase the likelihood of users

remaining dormant and therefore essentially ‘scrolling’ straight past the content.

Managers are therefore advised to present informational and entertaining content

exclusively from each other, rather than attempting to provide posts that contain both

elements. Similar effects were observed regarding the simultaneous presence of

informational and relational social media content, which had a detrimental effect on

positively-valenced SMEBs, whilst the likelihood that users would either remain

dormant, or detach from the social media content increased. Although the negative

effects of combining two social media content types were minimal, when the possible

number of relationships are considered, it should be noted that there were no additional

positive effects identified in combining content types. Therefore, for managers, there is

no benefit in designing social media content which attempts to simultaneously appear to

users’ needs for information, entertainment, remuneration and relational interaction.

Selecting the appropriate level of richness

The results indicate that managers should carefully considered the type of post used

(status, photo or video), depending on the type of content being delivered to users. For

example, the findings showed that when a post has more than two elements of

information, a high level of richness (e.g. video) should be used in order to increase the

number of times the post will be shared. For informational posts with less than two

specific items of informational content, a medium level of richness (the use of a photo)

resulted in the greatest number of post shares. For entertaining, remunerative and

relational content, there was no significant effect identified regarding the type of post

used.

244 | P a g e
Chapter 7: Discussion and Conclusion
Selecting the appropriate level of content congruity

The findings show that in order to increase the number of likes on a post, informational

content should be of medium congruity. This means that informational content should

be more generally related to the product category (e.g. wine), rather than specifically

related to the brand and product. Comparatively, the results showed that when the

content is entertaining, it should be related specifically to the brand or product (high

congruity) in order to increase the number of comments on a post (creating behaviour).

Posting entertaining content that has nothing to do with the product or brand is therefore

not a recommended strategy for wine brands.

Developing the community size

The results indicated that community size (measured by the number of likes on the

brand page) had a significant impact on the occurrence of SMEBs. Medium and large

community sizes (over 1,500 ‘fans’) demonstrated a clear advantage in the number of

likes received on entertaining posts. Additionally, for small community sizes, increasing

the amount of entertaining content within a post decreased the number of shares made

on a post. However, for medium and large community sizes, increasing the community

size to 1,500 fans or more significantly increased the number of times the post was

shared. Small community sizes were also penalised regarding dormant behaviour. While

increasing levels of entertaining content within a post decreased the dormancy of users

for medium and large community sizes, when the same posts were made in small

community sizes, users increased in their level of dormancy. These findings show a

clear justification for increasing the number of fans on social media pages in order to

enhance the community size. It is advised that wine brand endeavour to increase their

fan base to over 1,500 users in order to elicit more favourable responses to social media

245 | P a g e
Chapter 7: Discussion and Conclusion
content.

Social media content scheduling

A final important consideration for managers to consider when designing and

determining social media strategy relates to the importance of post scheduling. The

findings showed that users are significantly less likely to comment on posts when they

are made on Thursdays and Sundays. It is advised that managers should avoid posting

content on these days if they are seeking to increase the number of comments made on

the post. The likelihood of social media posts being shared also significantly decreased

on Mondays, Thursdays and Sundays.

It is therefore advised that social media content is posted on Tuesdays, Wednesdays or

Saturdays. If managers seek to increase the number of likes on the social media content,

Tuesdays and Sundays should be avoided, as the results indicated a significant decrease

in this behaviour for these two days. In line with these suggestions, the results indicated

that users are more dormant in their SMEB on Thursdays and Sundays.

The time of the day in which the post was made is an important consideration for

managers. The findings showed that posts are significantly less likely to be shared and

liked if the post is made before midday. It is therefore advised that managers schedule

the delivery of their content for the afternoon if they wish to increase the number of

likes and shares received on a post. This finding is consistent with previous speculation

regarding engagement times, with social media found to engage least during the

morning and early afternoon, with increased interaction toward the evening, reaching a

steady, high level during the night (Golder et al., 2007).

246 | P a g e
Chapter 7: Discussion and Conclusion
Managers are therefore advised to carefully consider the scheduling of the delivery of

social media content, and take advantage of the ‘Schedule Post’ function provided

within Facebook. It is recommended that posts are not made on Thursdays or Sundays

as users are least active in their SMEBs on these days. It is also recommended that

brands schedule their content for after midday. Users appear to be less active in their

SMEB in the morning.

7.6 Concluding Thoughts


As a result of this thesis, the knowledge of how brands can strategically facilitate

engagement behaviour in the social media forum has been extended. Greater insight into

the nature of SMEB has been achieved, through the conceptualisation and measurement

of positively (creating, contributing and consuming) and negatively-valenced

(detaching) engagement behaviours, in addition to a neutral state of engagement

(dormancy).

Incorporating the research areas of customer engagement, social media marketing and

UGT has allowed a more strategic and empirical investigation of engagement. The

findings from this thesis have provided a framework for understanding the nature of

engagement behaviour in the online forum. The thesis provides support for UGT (Katz

& Foulkes, 1962; Ko et al., 2005) as an appropriate theoretical lens through which

users’ responses to social media content can be further understood.

The influence of social media content on SMEB is a central contribution of this thesis,

justifying how various forms of social media content can be strategically designed to

influence the occurrence of creating, contributing, consuming, dormancy and detaching

behaviour. The change in social media engagement behaviour as a result of

247 | P a g e
Chapter 7: Discussion and Conclusion
informational, entertaining, relational and remunerative social media content

empirically demonstrates that users are active and selective recipients of content, and

freely choose the content that they wish to engage with, as supported by UGT.

While a majority of engagement literature and corresponding customer engagement

behaviour literature has focussed on the development and conceptualisation of the

concept, this thesis extends the ideas through empirical quantitative enquiry. It explores

the under-researched role of specific marketing activities as an antecedent to

engagement behaviour, in addition to moderators of this relationship. Continual

theoretical and empirical development of the antecedents and consequences of SMEB is

essential for the development of engagement research in the digital age.

248 | P a g e
Appendices

Appendices

Appendix A: Email to Participating Wine Brands

Hello______,

My name is Rebecca Dolan. I am currently completing my Ph.D. in Marketing at the University


of Adelaide. I am now in the second year of this project and am seeking expressions of interest
from wineries that may like to participate in the project.

The study explores the use of Social Media (in particular, Facebook) by Australian wine brands
and how various communication efforts influence customer engagement behaviour. The project
has a number of valuable outputs to the industry, including determining exactly how social
media communication can be used effectively to build and strengthen positive brand
relationships with new and existing customers. The study is fully funded and supported by a
Category 1 GWRDC Research Grant, which we were awarded in 2012.

At this stage of the project, we are looking for wineries that would be willing to share a portion
of their “Facebook Insights” data with us for analysis. All detail of specific brands and
corresponding data will be kept entirely confidential in research outputs and publications.

Participating wineries will receive a full report of the research findings and contributions of the
Ph.D. following the completion of the analysis. This report will include data from all phases of
the study and unique insights into the value of social media use as a marketing communications
tool in the wine industry.

If this is something you would be interested in, please feel free to get in touch with me.
Additionally, if you know of any other wine brands that may also be interested in the project,
please feel free to pass this email on.

Analysis of the data will not commence until later in the year. However we are hoping to
finalise our list of collaborating wineries in the coming months.

Thank you for your consideration and I look forward to hearing from you soon,

Kind Regards.

249 | P a g e
Appendices

Appendix B: NVivo10 Word Frequency Report

Word Count Similar Words


wines 250 #wine, #wines, 'wine, wine', wine#slide, wines
today 249 today
cellars 231 cellar, cellaring, cellars
great 223 great
tastings 211 taste, 'taste', tasted, tastes, tasting, tastings
vineyards 193 #vineyard, vineyard, vineyard', vineyards, vineyards'
shiraz 184 #shiraz, shiraz, shiraz'
weekend 180 weekend, weekends
McLaren 148 McLaren, 'McLaren
estate 145 estate
vintages 145 #vintage, @vintage, vintage, vintages
winemaker 141 winemaker, winemaker', winemakers, winemakers', winemaking
events 135 event, events
Barossa 132 #Barossa, Barossa, Barossa'
visit 129 visit, visited, visiting, visits
photos 119 photo, photos
enjoys 119 #enjoy, enjoy, enjoyable, enjoyed, enjoying, enjoyment, enjoys
check 118 check, checked, checking
morning 114 morning
thanks 113 thank, thankful, thanking, thanks
Adelaide 113 #Adelaide, @Adelaide, Adelaide, 'Adelaide, Adelaide’s
restaurants 108 #restaurant, restaurant, restaurants
bottling 108 bottle, bottle', bottled, bottles, bottling
winery 107 #winery, wineries, winery
beauty 105 beauties, beautiful, beautifully, beauty
Friday 105 #Friday, Friday
festival 104 festival, festive, festivities
looks 103 looked, looking, looks
Australia’s 102 #Australia, Australia, Australia’s
Sunday 100 Sunday, Sundays
tickets 98 ticket, ticketing, tickets
friends 95 friend, friendly, friends
little 94 little
release 90 release, release', released, releases, releasing
voyager 90 voyage’, voyager
nights 86 night, nights
happy 85 happy
first 79 first
amazing 78 #amazing, amazed, amazing, 'amazing'
vines 78 #vines, vines
cabernet 77 cabernet, cabernet’, cabernets
chapel 76 chapel
Grenache 76 #Grenache, Grenache, Grenaches
purchased 75 #purchase, purchase, purchased, purchases, purchasing
Oliver 74 olive, Oliver, Oliver’s, olives
dinner 73 dinner, dinners
glass 73 glass, glass', glasses

250 | P a g e
Appendices
share 73 share, shared, shares, sharing
wirra 73 wirra
Pindarie 71 Pindarie
Australian 71 Australian, Australians
turkey 71 turkey, 'turkey', turkeys
available 70 avail, availability, available
blanc 69 blanc, blancs
starts 69 start, started, starting, starts
coombeyarraval 68 #coombeyarravalley, @coombeyarravalley, coombeyarravalley
ley
including 67 include, included, includes, including
Christmas 65 #Christmas, Christmas
online 65 online
gardens 64 #garden, #gardens, garden, gardening, gardens, gardens'
Henschke 64 #Henschke, @Henschke, Henschke
river 64 river
weather 64 #weather, weather
#Yarra valley 63 #Yarra valley
Margaret 63 Margaret
sauvignon 63 sauvignon
excited 62 excited, excitement, exciting, excitingly
grape 62 grape, grapes, grapes'
#melbaestate 58 #melbaestate
offer 58 offer, offered, offering, offers
block 57 block, blocked, blocks
think 57 think, thinking
music 56 music, 'music, musical
celebrity 55 celebrate, celebrated, celebrating, celebration, celebrations, celebrity
loving 55 loved, lovely, loves, loving
perfectly 55 perfect, 'perfect, perfectly
whiting 55 white, whites, whites', whiting
awesome 54 awesome
specials 54 special, specials
everyone 53 everyone
chardonnay 53 #chardonnay, chardonnay, chardonnays
still 53 still, 'still
another 52 another
party 52 partie, parties, party
summer 52 summer
coming 51 comes, coming
covered 51 cover, covered, covers
review 51 review, reviewed, reviewers, reviews
Taranga 50 taranga, 'taranga
tomorrow 50 tomorrow
delicious 50 #delicious, delicious, deliciously
lunch 49 lunch, lunching
around 48 around
valley 48 valley, valleys
course 47 course, courses
Easter 47 #Easter, Easter
gourmet 47 gourmet, 'gourmet
matched 47 match, matched, matches, matching
251 | P a g e
Appendices

tonight 47 tonight, tonight’s


awards 46 award, awarded, awards
favourite 45 favourite, favourites
month 45 month, monthly, months
Riesling 45 Riesling, Riesling', Rieslings
winter 45 #winter, winter
merlot 45 #merlot, merlot
d'arenberg 44 d'arenberg
forward 44 forward
going 44 going
Sydney 44 Sydney
selection 44 selected, selecting, selection, selections
south 44 south, south'
turkeyflat 44 #turkeyflat, turkeyflat
experience 43 experience, experience', experience’, experiences, experiment,
please 43 please, pleased
website 43 website
drinks 42 drink, drinking, drinks
family 42 #family, families, family
Corrina 42 Corrina
producers 42 #produce, produce, produced, producers, produces, producing
@voyagerestate 41 @voyagerestate, voyagerestate
fruits 41 fruit, fruitful, fruits
Monday 41 Monday, Mondayitis
Saturday 41 Saturday
years 41 yearly, years
order 40 order, ordered, ordering, orders
quick 40 quick, quickly
blend 40 blend, blended, blending, blends
competition 40 #competition, competition
Coombe 40 #Coombe, Coombe
bookings 39 booked, booking, bookings
#Victoria 39 #Victoria, Victoria
d'arry 39 d'arry
Melbourne 39 #Melbourne, Melbourne
region 39 region, regional, regionality, regions
early 38 early
latest 38 latest
opening 38 opened, opening, opens
prizes 38 #prizes, prize, prized, prizes
winner 38 winner, winners, winners'
fermenters 37 ferment, fermentation, fermented, fermenter, fermenters, fermenting,
ferments
feature 37 feature, featured, features, featuring
James 37 James, James'
season 37 season, seasonal, seasons
something 37 something, something’
lucky 36 lucky
bring 35 bring, bringing, brings
afternoon 35 afternoon
every 35 every

252 | P a g e
Appendices

place 35 place, placed, places


collections 35 collected, collecting, collection, collections, collective, collects
fabulous 34 fabulous
Fleurieu 34 Fleurieu, 'fleurieu
local 34 local, locally, locals
market 34 market, marketing, markets
people 34 people
sparkling 34 #sparkling, sparkling
verandah 34 verandah
manager 34 manage, managed, manager, managing
single 33 single
world 33 world
spring 33 #spring, spring, springs
barrels 32 barrel, barreling, barrels, barrels'
chestier 32 chestier
closes 32 close, closed, closes, closing
Lenswood 32 #Lenswood, Lenswood
making 32 makes, making
melba 32 melba, melbas
picking 32 picked, picking, picking'
semillon 32 semillon
taking 32 takes, taking
watch 32 watch, watching
flavour 31 flavour, flavoured, flavours
follow 31 follow, followed, followers, following
wonderful 31 wonder, wondered, wonderful, wondering
cheers 31 #cheers, cheer, cheerful, cheers, 'cheers'
harvest 31 harvest, harvested, harvesting
congratulations 30 congratulations
lemon 30 lemon, lemons
ready 30 ready
receive 30 receive, received, receives, receiving
works 30 worked, working, works
Facebook 29 Facebook
juice 29 juice
press 29 press, pressed, pressing, pressings
yesterday 29 yesterday
announced 29 announce, announced, announcement, announcements, announcing
better 29 better, 'better
chance 29 chance, chances
international 28 intern, internal, international, 'international
person 28 person, personal, personalities, personality, personally
serving 28 serve, served, serving
fantastic 28 fantastic
Instagram 28 Instagram
launch 28 launch, launched, launches, launching
magazine 28 magazine
really 28 really
artist 27 artist, artistic, artists
getting 27 getting
Halliday 27 Halliday

253 | P a g e
Appendices

Appendix C: Word Search Formulas for Post Content Coding


Code Formula
Informational Content
Brand name =IF(ISNUMBER(SEARCH("DFW",$D3)),"1",IF(ISNUMBER(SEARCH("turkey flat",$D3)),"1","0"))
Product image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Winery image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Price =IF(ISNUMBER(SEARCH("$",$D3)),"1",IF(ISNUMBER(SEARCH("PRICE",$D3)),"1",IF(ISNUMBER(SEARCH("dollar",$D3)),"1","0")))
Venue image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Review/award image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Tasting, samples, testing =IF(ISNUMBER(SEARCH("tasting",$D3)),"1",IF(ISNUMBER(SEARCH("taste",$D3)),"1",IF(ISNUMBER(SEARCH("tried",$D3)),"1",IF(ISNU
MBER(SEARCH("samples",$D3)),"1",IF(ISNUMBER(SEARCH("try",$D3)),"1",IF(ISNUMBER(SEARCH("try",$D3)),"1",IF(ISNUMBER(SE
ARCH("trying",$D3)),"1","0")))))))
Variety =IF(ISNUMBER(SEARCH("Chardonnay",$D4)),"1",IF(ISNUMBER(SEARCH("Pinot
Grigio",$D4)),"1",IF(ISNUMBER(SEARCH("Riesling",$D4)),"1",IF(ISNUMBER(SEARCH("Sauvignon
Blanc",$D4)),"1",IF(ISNUMBER(SEARCH("Viogner",$D4)),"1",IF(ISNUMBER(SEARCH("Chenin
Blanc",$D4)),"1",IF(ISNUMBER(SEARCH("Gewurtztraminer",$D4)),"1",IF(ISNUMBER(SEARCH("Semillon",$D4)),"1",IF(ISNUMBER(SEA
RCH("Verdehlo",$D4)),"1",IF(ISNUMBER(SEARCH("Cabernet Sauvignon",$D4)),"1",IF(ISNUMBER(SEARCH("Pinot
Noir",$D4)),"1",IF(ISNUMBER(SEARCH("Tempranillo",$D4)),"1",IF(ISNUMBER(SEARCH("Carmanere",$D4)),"1",IF(ISNUMBER(SEARCH
("Durif",$D4)),"1",IF(ISNUMBER(SEARCH("Grenache",$D4)),"1",IF(ISNUMBER(SEARCH("Shiraz",$D4)),"1",IF(ISNUMBER(SEARCH("Sa
ngiovese",$D4)),"1",IF(ISNUMBER(SEARCH("Zinfandel",$D4)),"1",IF(ISNUMBER(SEARCH("Mouvedre",$D4)),"1",IF(ISNUMBER(SEARC
H("Mataro",$D4)),"1",IF(ISNUMBER(SEARCH("Syrah",$D4)),"1",IF(ISNUMBER(SEARCH("savignin",$D4)),"1",IF(ISNUMBER(SEARCH("
Traminer",$D4)),"1",IF(ISNUMBER(SEARCH("Colombard",$D4)),"1",IF(ISNUMBER(SEARCH("muscat gordo
blanco",$D4)),"1",IF(ISNUMBER(SEARCH("muscat a petits grains
blanc",$D4)),"1",IF(ISNUMBER(SEARCH("malbec",$D4)),"1",IF(ISNUMBER(SEARCH("nebbiolo",$D4)),"1",IF(ISNUMBER(SEARCH("ruby
cabernet",$D4)),"1",IF(ISNUMBER(SEARCH("petit
254 | P a g e
Appendices
verdot",$D4)),"1",IF(ISNUMBER(SEARCH("dolcetto",$D4)),"1",IF(ISNUMBER(SEARCH("durif",$D4)),"1",IF(ISNUMBER(SEARCH("barber
a",$D4)),"1",IF(ISNUMBER(SEARCH("cabernet franc",$D4)),"1",IF(ISNUMBER(SEARCH("muscat a petits grains
rouge",$D4)),"1",IF(ISNUMBER(SEARCH("merlot",$D4)),"1",IF(ISNUMBER(SEARCH("pinot D4)),"1
gris",$",IF(ISNUMBER(SEARCH("pinot
grigio",$D4)),"1",IF(ISNUMBER(SEARCH("roussane",$D4)),"1",IF(ISNUMBER(SEARCH("sultana",$D4)),"1",IF(ISNUMBER(SEARCH("treb
biano",$D4)),"1",IF(ISNUMBER(SEARCH("arneis",$D4)),"1",IF(ISNUMBER(SEARCH("crouchen",$D4)),"1",IF(ISNUMBER(SEARCH("mars
anne",$D4)),"1",IF(ISNUMBER(SEARCH("tarrango",$D4)),"1",IF(ISNUMBER(SEARCH("Touriga",$D4)),"1","0"))))))))))))))))))))))))))))))))))))
))))))))))
Product detail =IF(ISNUMBER(SEARCH("range",$D3)),"1",IF(ISNUMBER(SEARCH("product",$D3)),"1",IF(ISNUMBER(SEARCH("wine",$D3)),"1","0")))
Region =IF(ISNUMBER(SEARCH("Barossa",$D3)),"1",IF(ISNUMBER(SEARCH("Clare",$D3)),"1",IF(ISNUMBER(SEARCH("Eden",$D3)),"1",IF(IS
NUMBER(SEARCH("Currency",$D3)),"1",IF(ISNUMBER(SEARCH("Kangaroo Island",$D3)),"1",IF(ISNUMBER(SEARCH("Langhorne
Creek",$D3)),"1",IF(ISNUMBER(SEARCH("McLaren Vale",$D3)),"1",IF(ISNUMBER(SEARCH("Southern
Fleurieu",$D3)),"1",IF(ISNUMBER(SEARCH("Coonawarra",$D3)),"1",IF(ISNUMBER(SEARCH("Mount
Benson",$D3)),"1",IF(ISNUMBER(SEARCH("Padthaway",$D3)),"1",IF(ISNUMBER(SEARCH("Wrattonbully",$D3)),"1",IF(ISNUMBER(SEA
RCH("Robe",$D3)),"1",IF(ISNUMBER(SEARCH("Bordertown",$D3)),"1",IF(ISNUMBER(SEARCH("Riverland",$D3)),"1",IF(ISNUMBER(SE
ARCH("Adelaide Hills",$D3)),"1",IF(ISNUMBER(SEARCH("Lenswood",$D3)),"1",IF(ISNUMBER(SEARCH("Piccadilly
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Adelaide Plains",$D3)),"1",IF(ISNUMBER(SEARCH("Clare
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("North West",$D3)),"1",IF(ISNUMBER(SEARCH("Piccadilly
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Adelaide Plains",$D3)),"1",IF(ISNUMBER(SEARCH("Clare
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("North West",$D3)),"1",IF(ISNUMBER(SEARCH("Tamar
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Pipers River",$D3)),"1",IF(ISNUMBER(SEARCH("East
Coast",$D3)),"1",IF(ISNUMBER(SEARCH("Coal River",$D3)),"1",IF(ISNUMBER(SEARCH("Derwent
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Southern",$D3)),"1",IF(ISNUMBER(SEARCH("Bendigo",$D3)),"1",IF(ISNUMBER(SEARCH("G
oulburn Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Nagambie
Lakes",$D3)),"1",IF(ISNUMBER(SEARCH("Heathcote",$D3)),"1",IF(ISNUMBER(SEARCH("Strathbogie
Ranges",$D3)),"1",IF(ISNUMBER(SEARCH("Upper
Goulburn",$D3)),"1",IF(ISNUMBER(SEARCH("Gippsland",$D3)),"1",IF(ISNUMBER(SEARCH("Alpine
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Beechworth",$D3)),"1",IF(ISNUMBER(SEARCH("Glenrowan",$D3)),"1",IF(ISNUMBER(SEARC
H("Rutherglen",$D3)),"1",IF(ISNUMBER(SEARCH("Murray Darling",$D3)),"1",IF(ISNUMBER(SEARCH("Swan
Hill",$D3)),"1",IF(ISNUMBER(SEARCH("Geelong",$D3)),"1",IF(ISNUMBER(SEARCH("Macedon
Ranges",$D3)),"1",IF(ISNUMBER(SEARCH("Mornington
Peninsula",$D3)),"1",IF(ISNUMBER(SEARCH("Sunbury",$D3)),"1",IF(ISNUMBER(SEARCH("Yarra
255 | P a g e
Appendices
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Grampians",$D3)),"1",IF(ISNUMBER(SEARCH("Henty",$D3)),"1",IF(ISNUMBER(SEARCH("P
yrenees",$D3)),"1",IF(ISNUMBER(SEARCH("Peel",$D3)),"1",IF(ISNUMBER(SEARCH("Perth
Hills",$D3)),"1",IF(ISNUMBER(SEARCH("Swan Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Blackwood
Valley",$D3)),"1",IF(ISNUMBER(SEARCH("Geographe",$D3)),"1",IF(ISNUMBER(SEARCH("Great
Southern",$D3)),"1",IF(ISNUMBER(SEARCH("Albany",$D3)),"1",IF(ISNUMBER(SEARCH("Denmark",$D3)),"1",IF(ISNUMBER(SEARCH("
Frankland River",$D3)),"1","0")))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))
Winemaking and processing =IF(ISNUMBER(SEARCH("winemaking",$D3)),"1",IF(ISNUMBER(SEARCH("ferment",$D3)),"1",IF(ISNUMBER(SEARCH("crop",$D3)),"1",
IF(ISNUMBER(SEARCH("pick",$D3)),"1",IF(ISNUMBER(SEARCH("harvest",$D3)),"1",IF(ISNUMBER(SEARCH("crush",$D3)),"1",IF(ISNU
MBER(SEARCH("bottle",$D3)),"1",IF(ISNUMBER(SEARCH("press",$D3)),"1",IF(ISNUMBER(SEARCH("rack",$D3)),"1",IF(ISNUMBER(SE
ARCH("barrel",$D3)),"1",IF(ISNUMBER(SEARCH("blend",$D3)),"1",IF(ISNUMBER(SEARCH("vintage",$D3)),"1",IF(ISNUMBER(SEARCH
("verasion",$D3)),"1","0")))))))))))))
Vineyard/Location (?) =IF(ISNUMBER(SEARCH("vineyard",$D3)),"1",IF(ISNUMBER(SEARCH("vines",$D3)),"1",IF(ISNUMBER(SEARCH("hours",$D3)),"1",IF(I
SNUMBER(SEARCH("winery",$D3)),"1","0"))))
Hours =IF(ISNUMBER(SEARCH("open",$D3)),"1",IF(ISNUMBER(SEARCH("closed",$D3)),"1",IF(ISNUMBER(SEARCH("hours",$D3)),"1",IF(ISN
UMBER(SEARCH("opening",$D3)),"1",IF(ISNUMBER(SEARCH("times",$D3)),"1",IF(ISNUMBER(SEARCH("shut",$D3)),"1","0"))))))
Year =IF(ISNUMBER(SEARCH("19",$D3)),"1",IF(ISNUMBER(SEARCH("20",$D3)),"1","0"))
Contact details =IF(ISNUMBER(SEARCH("phone",$D3)),"1",IF(ISNUMBER(SEARCH("email",$D3)),"1",IF(ISNUMBER(SEARCH("contact",$D3)),"1",IF(IS
NUMBER(SEARCH("address",$D3)),"1",IF(ISNUMBER(SEARCH("location",$D3)),"1",IF(ISNUMBER(SEARCH("website",$D3)),"1","0"))))))
Brand facts/news =IF(ISNUMBER(SEARCH("did you
know",$D3)),"1",IF(ISNUMBER(SEARCH("fact",$D3)),"1",IF(ISNUMBER(SEARCH("news",$D3)),"1",IF(ISNUMBER(SEARCH("update",$D
3)),"1","0"))))
Service/Facility =IF(ISNUMBER(SEARCH("service",$D3)),"1",IF(ISNUMBER(SEARCH("facility",$D3)),"1",IF(ISNUMBER(SEARCH("facilities",$D3)),"1"
,IF(ISNUMBER(SEARCH("venue",$D3)),"1",IF(ISNUMBER(SEARCH("event",$D3)),"1",IF(ISNUMBER(SEARCH("function",$D3)),"1",IF(
ISNUMBER(SEARCH("occasion",$D3)),"1",IF(ISNUMBER(SEARCH("wedding",$D3)),"1",IF(ISNUMBER(SEARCH("party",$D3)),"1",IF(I
SNUMBER(SEARCH("celebration",$D3)),"1","0"))))))))))
Wine show, reviews and =IF(ISNUMBER(SEARCH("wine
awards show",$D3)),"1",IF(ISNUMBER(SEARCH("win",$D3)),"1",IF(ISNUMBER(SEARCH("won",$D3)),"1",IF(ISNUMBER(SEARCH("award",$D3
)),"1",IF(ISNUMBER(SEARCH("awarded",$D3)),"1",IF(ISNUMBER(SEARCH("recieved",$D3)),"1",IF(ISNUMBER(SEARCH("achieved",$D3
)),"1",IF(ISNUMBER(SEARCH("successful",$D3)),"1",IF(ISNUMBER(SEARCH("medal",$D3)),"1",IF(ISNUMBER(SEARCH("trophy",$D3)),"
1",IF(ISNUMBER(SEARCH("result",$D3)),"1",IF(ISNUMBER(SEARCH("points",$D3)),"1",IF(ISNUMBER(SEARCH("review",$D3)),"1",IF(I
SNUMBER(SEARCH("presented",$D3)),"1",IF(ISNUMBER(SEARCH("presenting",$D3)),"1",IF(ISNUMBER(SEARCH("gold",$D3)),"1",IF(IS

256 | P a g e
Appendices
NUMBER(SEARCH("silver",$D3)),"1",IF(ISNUMBER(SEARCH("bronze",$D3)),"1",IF(ISNUMBER(SEARCH("star",$D3)),"1",IF(ISNUMBE
R(SEARCH("judge",$D3)),"1",IF(ISNUMBER(SEARCH("named",$D3)),"1",IF(ISNUMBER(SEARCH("listed",$D3)),"1",IF(ISNUMBER(SEA
RCH("finalist",$D3)),"1",IF(ISNUMBER(SEARCH("achievement",$D3)),"1",IF(ISNUMBER(SEARCH("success",$D3)),"1",IF(ISNUMBER(SE
ARCH("rating",$D3)),"1",IF(ISNUMBER(SEARCH("wineoftheyear",$D3)),"1","0")))))))))))))))))))))))))))
Wine description =IF(ISNUMBER(SEARCH("red",$D3)),"1",IF(ISNUMBER(SEARCH("white",$D3)),"1",IF(ISNUMBER(SEARCH("fruit",$D3)),"1",IF(ISNUM
BER(SEARCH("tannin",$D3)),"1",IF(ISNUMBER(SEARCH("oak",$D3)),"1",IF(ISNUMBER(SEARCH("fresh",$D3)),"1",IF(ISNUMBER(SEA
RCH("clean",$D3)),"1",IF(ISNUMBER(SEARCH("crisp",$D3)),"1",IF(ISNUMBER(SEARCH("elegant",$D3)),"1",IF(ISNUMBER(SEARCH("s
oft",$D3)),"1",IF(ISNUMBER(SEARCH("smooth",$D3)),"1",IF(ISNUMBER(SEARCH("bold",$D3)),"1",IF(ISNUMBER(SEARCH("chocolate",
$D3)),"1",IF(ISNUMBER(SEARCH("rich",$D3)),"1",IF(ISNUMBER(SEARCH("full
bodied",$D3)),"1",IF(ISNUMBER(SEARCH("yum",$D3)),"1",IF(ISNUMBER(SEARCH("tasty",$D3)),"1",IF(ISNUMBER(SEARCH("delicious"
,$D3)),"1",IF(ISNUMBER(SEARCH("spice",$D3)),"1",IF(ISNUMBER(SEARCH("zest",$D3)),"1",IF(ISNUMBER(SEARCH("acid",$D3)),"1",I
F(ISNUMBER(SEARCH("aroma",$D3)),"1",IF(ISNUMBER(SEARCH("dense",$D3)),"1",IF(ISNUMBER(SEARCH("palate",$D3)),"1",IF(ISNU
MBER(SEARCH("flavour",$D3)),"1",IF(ISNUMBER(SEARCH("fragrance",$D3)),"1",IF(ISNUMBER(SEARCH("balanced",$D3)),"1",IF(ISNU
MBER(SEARCH("caramel",$D3)),"1",IF(ISNUMBER(SEARCH("complementing",$D3)),"1",IF(ISNUMBER(SEARCH("citrus",$D3)),"1",IF(IS
NUMBER(SEARCH("chalky",$D3)),"1",IF(ISNUMBER(SEARCH("characters",$D3)),"1",IF(ISNUMBER(SEARCH("notes",$D3)),"1",IF(ISNU
MBER(SEARCH("raisin",$D3)),"1",IF(ISNUMBER(SEARCH("mocha",$D3)),"1",IF(ISNUMBER(SEARCH("tannic",$D3)),"1",IF(ISNUMBER
(SEARCH("toffee",$D3)),"1",IF(ISNUMBER(SEARCH("vibrant",$D3)),"1",IF(ISNUMBER(SEARCH("colour",$D3)),"1","0")))))))))))))))))))))))
))))))))))))))))
Entertaining Content
Emoticon =IF(ISNUMBER(SEARCH(":-)",$D4)),"1","0")
Humour =IF(ISNUMBER(SEARCH("fun",$D4)),"1",IF(ISNUMBER(SEARCH("funny",$D4)),"1",IF(ISNUMBER(SEARCH("banter",$D4)),"1",IF(ISNU
MBER(SEARCH("joke",$D4)),"1",IF(ISNUMBER(SEARCH("gag",$D4)),"1",IF(ISNUMBER(SEARCH("happy",$D4)),"1",IF(ISNUMBER(SE
ARCH("joking",$D4)),"1",IF(ISNUMBER(SEARCH("kidding",$D4)),"1",IF(ISNUMBER(SEARCH("april
fools",$D4)),"1",IF(ISNUMBER(SEARCH("hilarious",$D4)),"1",IF(ISNUMBER(SEARCH("cool",$D4)),"1",IF(ISNUMBER(SEARCH("whimsi
cal",$D4)),"1",IF(ISNUMBER(SEARCH("exciting",$D4)),"1",IF(ISNUMBER(SEARCH("haha",$D4)),"1",IF(ISNUMBER(SEARCH("hehe",$D
4)),"1",IF(ISNUMBER(SEARCH("entertain",$D4)),"1",IF(ISNUMBER(SEARCH("laugh",$D4)),"1",IF(ISNUMBER(SEARCH("giggle",$D4)),"
1","0"))))))))))))))))))
Vineyard Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Meme Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
257 | P a g e
Appendices
Animal Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Slang =IF(ISNUMBER(SEARCH("lol",$D4)),"1",IF(ISNUMBER(SEARCH("omg",$D4)),"1",IF(ISNUMBER(SEARCH("jk",$D4)),"1",IF(ISNUMBE
R(SEARCH("wtf",$D4)),"1",IF(ISNUMBER(SEARCH("L8R",$D4)),"1",IF(ISNUMBER(SEARCH("plz",$D4)),"1",IF(ISNUMBER(SEARCH("t
tyl",$D4)),"1",IF(ISNUMBER(SEARCH("cheers",$D4)),"1",IF(ISNUMBER(SEARCH("guys",$D4)),"1",IF(ISNUMBER(SEARCH("wow",$D4)
),"1",IF(ISNUMBER(SEARCH("arvo",$D4)),"1",IF(ISNUMBER(SEARCH("aussie",$D4)),"1","0"))))))))))))
Remunerative Content
Deal/Offer =IF(ISNUMBER(SEARCH("special",$D4)),"1",IF(ISNUMBER(SEARCH("discount",$D4)),"1",IF(ISNUMBER(SEARCH("exclusive",$D4)),"1"
,IF(ISNUMBER(SEARCH("deal",$D4)),"1",IF(ISNUMBER(SEARCH("sale",$D4)),"1",IF(ISNUMBER(SEARCH("promotion",$D4)),"1",IF(ISN
UMBER(SEARCH("clearance",$D4)),"1",IF(ISNUMBER(SEARCH("bargain",$D4)),"1",IF(ISNUMBER(SEARCH("on
sale",$D4)),"1",IF(ISNUMBER(SEARCH("marked down",$D4)),"1",IF(ISNUMBER(SEARCH("low price",$D4)),"1","0")))))))))))
Purchase instructions =IF(ISNUMBER(SEARCH("buy",$D4)),"1",IF(ISNUMBER(SEARCH("order",$D4)),"1",IF(ISNUMBER(SEARCH("purchase",$D4)),"1",IF(IS
NUMBER(SEARCH("order form",$D4)),"1",IF(ISNUMBER(SEARCH("store",$D4)),"1",IF(ISNUMBER(SEARCH("shop",$D4)),"1","0"))))))
Competition =IF(ISNUMBER(SEARCH("win",$D4)),"1",IF(ISNUMBER(SEARCH("reward",$D4)),"1",IF(ISNUMBER(SEARCH("free",$D4)),"1",IF(ISNU
MBER(SEARCH("prize",$D4)),"1","0"))))
Relational Content
Question =IF(ISNUMBER(SEARCH("~?",$D4)),"1",IF(ISNUMBER(SEARCH("question",$D4)),"1","0"))
Quiz/game =IF(ISNUMBER(SEARCH("quiz",$D4)),"1",IF(ISNUMBER(SEARCH("game",$D4)),"1",IF(ISNUMBER(SEARCH("test",$D4)),"1",IF(ISNUM
BER(SEARCH("guess",$D4)),"1",IF(ISNUMBER(SEARCH("challenge",$D4)),"1","0")))))
Holiday, event, day =IF(ISNUMBER(SEARCH("birthday",$D4)),"1",IF(ISNUMBER(SEARCH("christmas",$D4)),"1",IF(ISNUMBER(SEARCH("easter",$D4)),"1",I
F(ISNUMBER(SEARCH("boxing day",$D4)),"1",IF(ISNUMBER(SEARCH("new year",$D4)),"1",IF(ISNUMBER(SEARCH("Australia
day",$D4)),"1",IF(ISNUMBER(SEARCH("good friday",$D4)),"1",IF(ISNUMBER(SEARCH("anzac
day",$D4)),"1",IF(ISNUMBER(SEARCH("queen's birthday",$D4)),"1",IF(ISNUMBER(SEARCH("labour
day",$D4)),"1",IF(ISNUMBER(SEARCH("holiday",$D4)),"1",IF(ISNUMBER(SEARCH("public
holiday",$D4)),"1",IF(ISNUMBER(SEARCH("Melbourne cup",$D4)),"1",IF(ISNUMBER(SEARCH("April
fool",$D4)),"1",IF(ISNUMBER(SEARCH("father's day",$D4)),"1",IF(ISNUMBER(SEARCH("mother's
day",$D4)),"1",IF(ISNUMBER(SEARCH("monday",$D4)),"1",IF(ISNUMBER(SEARCH("tuesday",$D4)),"1",IF(ISNUMBER(SEARCH("wedne
sday",$D4)),"1",IF(ISNUMBER(SEARCH("thursday",$D4)),"1",IF(ISNUMBER(SEARCH("friday",$D4)),"1",IF(ISNUMBER(SEARCH("saturd
ay",$D4)),"1",IF(ISNUMBER(SEARCH("sunday",$D4)),"1","0")))))))))))))))))))))))

258 | P a g e
Appendices
Emotion =IF(ISNUMBER(SEARCH("delighted",$D5)),"1",IF(ISNUMBER(SEARCH("ebullient",$D5)),"1",IF(ISNUMBER(SEARCH("ecstatic",$D5)),"1
",IF(ISNUMBER(SEARCH("elated",$D5)),"1",IF(ISNUMBER(SEARCH("energetic",$D5)),"1",IF(ISNUMBER(SEARCH("enthusiastic",$D5)),"
Happy 1",IF(ISNUMBER(SEARCH("euphoric",$D5)),"1",IF(ISNUMBER(SEARCH("excited",$D5)),"1",IF(ISNUMBER(SEARCH("exhilarated",$D5)),
Caring "1",IF(ISNUMBER(SEARCH("overjoyed",$D5)),"1",IF(ISNUMBER(SEARCH("thrilled",$D5)),"1",IF(ISNUMBER(SEARCH("tickled
Depression pink",$D5)),"1",IF(ISNUMBER(SEARCH("turned
Inadequateness on",$D5)),"1",IF(ISNUMBER(SEARCH("vibrant",$D5)),"1",IF(ISNUMBER(SEARCH("zippy",$D5)),"1",IF(ISNUMBER(SEARCH("aglow",$D
Fear 5)),"1",IF(ISNUMBER(SEARCH("bouyant",$D5)),"1",IF(ISNUMBER(SEARCH("cheerful",$D5)),"1",IF(ISNUMBER(SEARCH("elevated",$D5
Confusion )),"1",IF(ISNUMBER(SEARCH("gleeful",$D5)),"1",IF(ISNUMBER(SEARCH("happy",$D5)),"1",IF(ISNUMBER(SEARCH("in high
Hurt spirits",$D5)),"1",IF(ISNUMBER(SEARCH("jovial",$D5)),"1",IF(ISNUMBER(SEARCH("light-
Anger hearted",$D5)),"1",IF(ISNUMBER(SEARCH("lively",$D5)),"1",IF(ISNUMBER(SEARCH("merry",$D5)),"1",IF(ISNUMBER(SEARCH("riding
Loneliness high",$D5)),"1",IF(ISNUMBER(SEARCH("sparling",$D5)),"1",IF(ISNUMBER(SEARCH("up",$D5)),"1","0")))))))))))))))))))))))))))))
Remorse =IF(ISNUMBER(SEARCH("adoring",$D5)),"1",IF(ISNUMBER(SEARCH("ardent",$D5)),"1",IF(ISNUMBER(SEARCH("cherishing",$D5)),"1",
IF(ISNUMBER(SEARCH("compassionate",$D5)),"1",IF(ISNUMBER(SEARCH("crazy
about",$D5)),"1",IF(ISNUMBER(SEARCH("devoted",$D5)),"1",IF(ISNUMBER(SEARCH("doting",$D5)),"1",IF(ISNUMBER(SEARCH("ferve
nt",$D5)),"1",IF(ISNUMBER(SEARCH("idolizing",$D5)),"1",IF(ISNUMBER(SEARCH("infatuated",$D5)),"1",IF(ISNUMBER(SEARCH("passi
onate",$D5)),"1",IF(ISNUMBER(SEARCH("wild
about",$D5)),"1",IF(ISNUMBER(SEARCH("worshipful",$D5)),"1",IF(ISNUMBER(SEARCH("zealous",$D5)),"1",IF(ISNUMBER(SEARCH("a
dmiring",$D5)),"1",IF(ISNUMBER(SEARCH("affectionate",$D5)),"1",IF(ISNUMBER(SEARCH("attached",$D5)),"1",IF(ISNUMBER(SEARCH
("fond",$D5)),"1",IF(ISNUMBER(SEARCH("huggy",$D5)),"1",IF(ISNUMBER(SEARCH("kind",$D5)),"1",IF(ISNUMBER(SEARCH("kind
hearted",$D5)),"1",IF(ISNUMBER(SEARCH("loving",$D5)),"1",IF(ISNUMBER(SEARCH("partial",$D5)),"1",IF(ISNUMBER(SEARCH("soft
on",$D5)),"1",IF(ISNUMBER(SEARCH("sympathetic",$D5)),"1",IF(ISNUMBER(SEARCH("tender",$D5)),"1",IF(ISNUMBER(SEARCH("trusti
ng",$D5)),"1",IF(ISNUMBER(SEARCH("warm
hearted",$D5)),"1",IF(ISNUMBER(SEARCH("appreciative",$D5)),"1",IF(ISNUMBER(SEARCH("attentive",$D5)),"1",IF(ISNUMBER(SEARC
H("considerate",$D5)),"1",IF(ISNUMBER(SEARCH("friendly",$D5)),"1",IF(ISNUMBER(SEARCH("interested
in",$D5)),"1",IF(ISNUMBER(SEARCH("kind",$D5)),"1",IF(ISNUMBER(SEARCH("respective",$D5)),"1",IF(ISNUMBER(SEARCH("thoughtf
ul",$D5)),"1",IF(ISNUMBER(SEARCH("tolerant",$D5)),"1",IF(ISNUMBER(SEARCH("warm
toward",$D5)),"1",IF(ISNUMBER(SEARCH("yielding",$D5)),"1","0")))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("alienated",$D5)),"1",IF(ISNUMBER(SEARCH("barren",$D5)),"1",IF(ISNUMBER(SEARCH("beaten",$D5)),"1",IF
(ISNUMBER(SEARCH("bleak",$D5)),"1",IF(ISNUMBER(SEARCH("bleeding",$D5)),"1",IF(ISNUMBER(SEARCH("dejected",$D5)),"1",IF(IS
NUMBER(SEARCH("depressed",$D5)),"1",IF(ISNUMBER(SEARCH("desolate",$D5)),"1",IF(ISNUMBER(SEARCH("despondent",$D5)),"1",I
F(ISNUMBER(SEARCH("dismal",$D5)),"1",IF(ISNUMBER(SEARCH("empty",$D5)),"1",IF(ISNUMBER(SEARCH("gloomy",$D5)),"1",IF(IS
NUMBER(SEARCH("grieve",$D5)),"1",IF(ISNUMBER(SEARCH("grim",$D5)),"1",IF(ISNUMBER(SEARCH("hopeless",$D5)),"1",IF(ISNUM

259 | P a g e
Appendices
BER(SEARCH("in
despair",$D5)),"1",IF(ISNUMBER(SEARCH("woeful",$D5)),"1",IF(ISNUMBER(SEARCH("worried",$D5)),"1",IF(ISNUMBER(SEARCH("awf
ul",$D5)),"1",IF(ISNUMBER(SEARCH("blue",$D5)),"1",IF(ISNUMBER(SEARCH("crestfall",$D5)),"1",IF(ISNUMBER(SEARCH("demoralize
d",$D5)),"1",IF(ISNUMBER(SEARCH("devalued",$D5)),"1",IF(ISNUMBER(SEARCH("discouraged",$D5)),"1",IF(ISNUMBER(SEARCH("dis
pirited",$D5)),"1",IF(ISNUMBER(SEARCH("distressed",$D5)),"1",IF(ISNUMBER(SEARCH("downcast",$D5)),"1",IF(ISNUMBER(SEARCH("
downhearted",$D5)),"1",IF(ISNUMBER(SEARCH("fed
up",$D5)),"1",IF(ISNUMBER(SEARCH("lost",$D5)),"1",IF(ISNUMBER(SEARCH("melancholy",$D5)),"1",IF(ISNUMBER(SEARCH("miserab
le",$D5)),"1",IF(ISNUMBER(SEARCH("regretful",$D5)),"1",IF(ISNUMBER(SEARCH("rotten",$D5)),"1",IF(ISNUMBER(SEARCH("sorrowful
",$D5)),"1",IF(ISNUMBER(SEARCH("tearful",$D5)),"1",IF(ISNUMBER(SEARCH("upset",$D5)),"1",IF(ISNUMBER(SEARCH("weepy",$D5))
,"1",IF(ISNUMBER(SEARCH("disappointed",$D5)),"1",IF(ISNUMBER(SEARCH("funk",$D5)),"1",IF(ISNUMBER(SEARCH("glum",$D5)),"1
",IF(ISNUMBER(SEARCH("moody",$D5)),"1",IF(ISNUMBER(SEARCH("morose",$D5)),"1",IF(ISNUMBER(SEARCH("sombre",$D5)),"1",IF
(ISNUMBER(SEARCH("subdued",$D5)),"1",IF(ISNUMBER(SEARCH("uncomfortable",$D5)),"1",IF(ISNUMBER(SEARCH("unhappy",$D5)),
"1","0")))))))))))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("blemished",$D5)),"1",IF(ISNUMBER(SEARCH("blotched",$D5)),"1",IF(ISNUMBER(SEARCH("broken",$D5)),"1
",IF(ISNUMBER(SEARCH("crippled",$D5)),"1",IF(ISNUMBER(SEARCH("damaged",$D5)),"1",IF(ISNUMBER(SEARCH("false",$D5)),"1",IF
(ISNUMBER(SEARCH("feeble",$D5)),"1",IF(ISNUMBER(SEARCH("finished",$D5)),"1",IF(ISNUMBER(SEARCH("flawed",$D5)),"1",IF(ISN
UMBER(SEARCH("helpless",$D5)),"1",IF(ISNUMBER(SEARCH("impotent",$D5)),"1",IF(ISNUMBER(SEARCH("inferior",$D5)),"1",IF(ISN
UMBER(SEARCH("invalid",$D5)),"1",IF(ISNUMBER(SEARCH("powerless",$D5)),"1",IF(ISNUMBER(SEARCH("useless",$D5)),"1",IF(ISNU
MBER(SEARCH("washed
up",$D5)),"1",IF(ISNUMBER(SEARCH("whipped",$D5)),"1",IF(ISNUMBER(SEARCH("worthless",$D5)),"1",IF(ISNUMBER(SEARCH("defea
ted",$D5)),"1",IF(ISNUMBER(SEARCH("deficient",$D5)),"1",IF(ISNUMBER(SEARCH("dopey",$D5)),"1",IF(ISNUMBER(SEARCH("feeble",
$D5)),"1",IF(ISNUMBER(SEARCH("helpless",$D5)),"1",IF(ISNUMBER(SEARCH("impaired",$D5)),"1",IF(ISNUMBER(SEARCH("imperfect"
,$D5)),"1",IF(ISNUMBER(SEARCH("incapable",$D5)),"1",IF(ISNUMBER(SEARCH("incompetent",$D5)),"1",IF(ISNUMBER(SEARCH("inco
mplete",$D5)),"1",IF(ISNUMBER(SEARCH("ineffective",$D5)),"1",IF(ISNUMBER(SEARCH("inept",$D5)),"1",IF(ISNUMBER(SEARCH("insi
gnificant",$D5)),"1",IF(ISNUMBER(SEARCH("meagre",$D5)),"1",IF(ISNUMBER(SEARCH("puny",$D5)),"1",IF(ISNUMBER(SEARCH("tenu
ous",$D5)),"1",IF(ISNUMBER(SEARCH("wishful",$D5)),"1",IF(ISNUMBER(SEARCH("lame",$D5)),"1",IF(ISNUMBER(SEARCH("substanda
rd",$D5)),"1",IF(ISNUMBER(SEARCH("unimportant",$D5)),"1","0"))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("alarmed",$D5)),"1",IF(ISNUMBER(SEARCH("appalled",$D5)),"1",IF(ISNUMBER(SEARCH("desperate",$D5)),"1
",IF(ISNUMBER(SEARCH("distressed",$D5)),"1",IF(ISNUMBER(SEARCH("frightened",$D5)),"1",IF(ISNUMBER(SEARCH("horrified",$D5))
,"1",IF(ISNUMBER(SEARCH("intimidated",$D5)),"1",IF(ISNUMBER(SEARCH("panicky",$D5)),"1",IF(ISNUMBER(SEARCH("paralysed",$D
5)),"1",IF(ISNUMBER(SEARCH("petrified",$D5)),"1",IF(ISNUMBER(SEARCH("shocked",$D5)),"1",IF(ISNUMBER(SEARCH("terrified",$D5
)),"1",IF(ISNUMBER(SEARCH("terror

260 | P a g e
Appendices
stricken",$D5)),"1",IF(ISNUMBER(SEARCH("wrecked",$D5)),"1",IF(ISNUMBER(SEARCH("afraid",$D5)),"1",IF(ISNUMBER(SEARCH("app
rehensive",$D5)),"1",IF(ISNUMBER(SEARCH("apprehensive",$D5)),"1",IF(ISNUMBER(SEARCH("awkward",$D5)),"1",IF(ISNUMBER(SEA
RCH("defensive",$D5)),"1",IF(ISNUMBER(SEARCH("fearful",$D5)),"1",IF(ISNUMBER(SEARCH("fidgety",$D5)),"1",IF(ISNUMBER(SEAR
CH("fretful",$D5)),"1",IF(ISNUMBER(SEARCH("jumpy",$D5)),"1",IF(ISNUMBER(SEARCH("nervous",$D5)),"1",IF(ISNUMBER(SEARCH("
scared",$D5)),"1",IF(ISNUMBER(SEARCH("shy",$D5)),"1",IF(ISNUMBER(SEARCH("skittish",$D5)),"1",IF(ISNUMBER(SEARCH("spineless
",$D5)),"1",IF(ISNUMBER(SEARCH("taut",$D5)),"1",IF(ISNUMBER(SEARCH("threatened",$D5)),"1",IF(ISNUMBER(SEARCH("troubled",$
D5)),"1",IF(ISNUMBER(SEARCH("wired",$D5)),"1",IF(ISNUMBER(SEARCH("anxious",$D5)),"1",IF(ISNUMBER(SEARCH("careful",$D5)),
"1",IF(ISNUMBER(SEARCH("cautious",$D5)),"1",IF(ISNUMBER(SEARCH("disquieted",$D5)),"1",IF(ISNUMBER(SEARCH("goose
bump",$D5)),"1",IF(ISNUMBER(SEARCH("shay",$D5)),"1",IF(ISNUMBER(SEARCH("tense",$D5)),"1",IF(ISNUMBER(SEARCH("timid",$D
5)),"1",IF(ISNUMBER(SEARCH("uneasy",$D5)),"1",IF(ISNUMBER(SEARCH("unsure",$D5)),"1",IF(ISNUMBER(SEARCH("watchful",$D5)),
"1",IF(ISNUMBER(SEARCH("worried",$D5)),"1","0"))))))))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("baffled",$D19)),"1",IF(ISNUMBER(SEARCH("befuddled",$D19)),"1",IF(ISNUMBER(SEARCH("chaotic",$D19)),
"1",IF(ISNUMBER(SEARCH("confounded",$D19)),"1",IF(ISNUMBER(SEARCH("confused",$D19)),"1",IF(ISNUMBER(SEARCH("dizzy",$D
19)),"1",IF(ISNUMBER(SEARCH("flustered",$D19)),"1",IF(ISNUMBER(SEARCH("rattled",$D19)),"1",IF(ISNUMBER(SEARCH("reeling",$D
19)),"1",IF(ISNUMBER(SEARCH("shocked",$D19)),"1",IF(ISNUMBER(SEARCH("shook
up",$D19)),"1",IF(ISNUMBER(SEARCH("speechless",$D19)),"1",IF(ISNUMBER(SEARCH("startled",$D19)),"1",IF(ISNUMBER(SEARCH("st
umped",$D19)),"1",IF(ISNUMBER(SEARCH("stunned",$D19)),"1",IF(ISNUMBER(SEARCH("taken-
aback",$D19)),"1",IF(ISNUMBER(SEARCH("thrown",$D19)),"1",IF(ISNUMBER(SEARCH("thunderstruck",$D19)),"1",IF(ISNUMBER(SEAR
CH("adrift",$D19)),"1",IF(ISNUMBER(SEARCH("ambivalent",$D19)),"1",IF(ISNUMBER(SEARCH("bewildered",$D19)),"1",IF(ISNUMBER(
SEARCH("bewildered",$D19)),"1",IF(ISNUMBER(SEARCH("puzzled",$D19)),"1",IF(ISNUMBER(SEARCH("blurred",$D19)),"1",IF(ISNUMB
ER(SEARCH("disconcerted",$D19)),"1",IF(ISNUMBER(SEARCH("disordered",$D19)),"1",IF(ISNUMBER(SEARCH("disorganised",$D19)),"1"
,IF(ISNUMBER(SEARCH("disquieted",$D19)),"1",IF(ISNUMBER(SEARCH("disturbed",$D19)),"1",IF(ISNUMBER(SEARCH("foggy",$D19)),
"1",IF(ISNUMBER(SEARCH("frustrated",$D19)),"1",IF(ISNUMBER(SEARCH("misled",$D19)),"1",IF(ISNUMBER(SEARCH("mistaken",$D1
9)),"1",IF(ISNUMBER(SEARCH("misunderstood",$D19)),"1",IF(ISNUMBER(SEARCH("mixed
up",$D19)),"1",IF(ISNUMBER(SEARCH("perplexed",$D19)),"1",IF(ISNUMBER(SEARCH("troubled",$D19)),"1",IF(ISNUMBER(SEARCH("d
istracted",$D19)),"1",IF(ISNUMBER(SEARCH("uncertain",$D19)),"1",IF(ISNUMBER(SEARCH("uncomfortable",$D19)),"1",IF(ISNUMBER(S
EARCH("undecided",$D19)),"1",IF(ISNUMBER(SEARCH("unsettled",$D19)),"1",IF(ISNUMBER(SEARCH("unsure",$D19)),"1","0")))))))))))))
))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("abused",$D19)),"1",IF(ISNUMBER(SEARCH("aching",$D19)),"1",IF(ISNUMBER(SEARCH("anguished",$D19)),"
1",IF(ISNUMBER(SEARCH("crushed",$D19)),"1",IF(ISNUMBER(SEARCH("degraded",$D19)),"1",IF(ISNUMBER(SEARCH("destroyed",$D1
9)),"1",IF(ISNUMBER(SEARCH("devastated",$D19)),"1",IF(ISNUMBER(SEARCH("discarded",$D19)),"1",IF(ISNUMBER(SEARCH("disgrac
ed",$D19)),"1",IF(ISNUMBER(SEARCH("forsaken",$D19)),"1",IF(ISNUMBER(SEARCH("humiliated",$D19)),"1",IF(ISNUMBER(SEARCH("

261 | P a g e
Appendices
mocked",$D19)),"1",IF(ISNUMBER(SEARCH("punished",$D19)),"1",IF(ISNUMBER(SEARCH("rejected",$D19)),"1",IF(ISNUMBER(SEARC
H("ridiculed",$D19)),"1",IF(ISNUMBER(SEARCH("ruined",$D19)),"1",IF(ISNUMBER(SEARCH("scorned",$D19)),"1",IF(ISNUMBER(SEAR
CH("stabbed",$D19)),"1",IF(ISNUMBER(SEARCH("tortured",$D19)),"1",IF(ISNUMBER(SEARCH("annoyed",$D19)),"1",IF(ISNUMBER(SEA
RCH("belittled",$D19)),"1",IF(ISNUMBER(SEARCH("cheapened",$D19)),"1",IF(ISNUMBER(SEARCH("criticized",$D19)),"1",IF(ISNUMBE
R(SEARCH("damaged",$D19)),"1",IF(ISNUMBER(SEARCH("depreciated",$D19)),"1",IF(ISNUMBER(SEARCH("devalued",$D19)),"1",IF(ISN
UMBER(SEARCH("discredited",$D19)),"1",IF(ISNUMBER(SEARCH("distressed",$D19)),"1",IF(ISNUMBER(SEARCH("impaired",$D19)),"1"
,IF(ISNUMBER(SEARCH("injured",$D19)),"1",IF(ISNUMBER(SEARCH("maligned",$D19)),"1",IF(ISNUMBER(SEARCH("marred",$D19)),"1
",IF(ISNUMBER(SEARCH("miffed",$D19)),"1",IF(ISNUMBER(SEARCH("mistreated",$D19)),"1",IF(ISNUMBER(SEARCH("resentful",$D19)
),"1",IF(ISNUMBER(SEARCH("troubled",$D19)),"1",IF(ISNUMBER(SEARCH("wounded",$D19)),"1",IF(ISNUMBER(SEARCH("let
down",$D19)),"1",IF(ISNUMBER(SEARCH("minimized",$D19)),"1",IF(ISNUMBER(SEARCH("minimised",$D19)),"1",IF(ISNUMBER(SEAR
CH("neglected",$D19)),"1",IF(ISNUMBER(SEARCH("put away",$D19)),"1",IF(ISNUMBER(SEARCH("put
down",$D19)),"1",IF(ISNUMBER(SEARCH("rueful",$D19)),"1",IF(ISNUMBER(SEARCH("tender",$D19)),"1",IF(ISNUMBER(SEARCH("unh
appy",$D19)),"1","0"))))))))))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("affronted",$D19)),"1",IF(ISNUMBER(SEARCH("belligerent",$D19)),"1",IF(ISNUMBER(SEARCH("bitter",$D19))
,"1",IF(ISNUMBER(SEARCH("burned
up",$D19)),"1",IF(ISNUMBER(SEARCH("enraged",$D19)),"1",IF(ISNUMBER(SEARCH("fuming",$D19)),"1",IF(ISNUMBER(SEARCH("furi
ous",$D19)),"1",IF(ISNUMBER(SEARCH("heated",$D19)),"1",IF(ISNUMBER(SEARCH("incensed",$D19)),"1",IF(ISNUMBER(SEARCH("inf
uriated",$D19)),"1",IF(ISNUMBER(SEARCH("intense",$D19)),"1",IF(ISNUMBER(SEARCH("outraged",$D19)),"1",IF(ISNUMBER(SEARCH(
"provoked",$D19)),"1",IF(ISNUMBER(SEARCH("seething",$D19)),"1",IF(ISNUMBER(SEARCH("storming",$D19)),"1",IF(ISNUMBER(SEAR
CH("truculent",$D19)),"1",IF(ISNUMBER(SEARCH("vengeful",$D19)),"1",IF(ISNUMBER(SEARCH("vindictive",$D19)),"1",IF(ISNUMBER(
SEARCH("wild",$D19)),"1",IF(ISNUMBER(SEARCH("aggravated",$D19)),"1",IF(ISNUMBER(SEARCH("annoyed",$D19)),"1",IF(ISNUMBE
R(SEARCH("antagonistic",$D19)),"1",IF(ISNUMBER(SEARCH("crabby",$D19)),"1",IF(ISNUMBER(SEARCH("cranky",$D19)),"1",IF(ISNU
MBER(SEARCH("exasperated",$D19)),"1",IF(ISNUMBER(SEARCH("fuming",$D19)),"1",IF(ISNUMBER(SEARCH("grouchy",$D19)),"1",IF(I
SNUMBER(SEARCH("hostile",$D19)),"1",IF(ISNUMBER(SEARCH("ill-
tempered",$D19)),"1",IF(ISNUMBER(SEARCH("indignant",$D19)),"1",IF(ISNUMBER(SEARCH("irate",$D19)),"1",IF(ISNUMBER(SEARCH(
"irritated",$D19)),"1",IF(ISNUMBER(SEARCH("offended",$D19)),"1",IF(ISNUMBER(SEARCH("ratty",$D19)),"1",IF(ISNUMBER(SEARCH("
resentful",$D19)),"1",IF(ISNUMBER(SEARCH("sore",$D19)),"1",IF(ISNUMBER(SEARCH("spiteful",$D19)),"1",IF(ISNUMBER(SEARCH("te
sty",$D19)),"1",IF(ISNUMBER(SEARCH("ticked
off",$D19)),"1",IF(ISNUMBER(SEARCH("bugged",$D19)),"1",IF(ISNUMBER(SEARCH("chargrined",$D19)),"1",IF(ISNUMBER(SEARCH("d
ismayed",$D19)),"1",IF(ISNUMBER(SEARCH("galled",$D19)),"1",IF(ISNUMBER(SEARCH("rueful",$D19)),"1",IF(ISNUMBER(SEARCH("g
rim",$D19)),"1",IF(ISNUMBER(SEARCH("impatient",$D19)),"1",IF(ISNUMBER(SEARCH("irked",$D19)),"1",IF(ISNUMBER(SEARCH("petu
lant",$D19)),"1",IF(ISNUMBER(SEARCH("resentful",$D19)),"1",IF(ISNUMBER(SEARCH("sullen",$D19)),"1",IF(ISNUMBER(SEARCH("upti

262 | P a g e
Appendices
ght",$D19)),"1","0")))))))))))))))))))))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("abandoned",$D19)),"1",IF(ISNUMBER(SEARCH("black",$D19)),"1",IF(ISNUMBER(SEARCH("cutoff",$D19)),"1
",IF(ISNUMBER(SEARCH("deserted",$D19)),"1",IF(ISNUMBER(SEARCH("destroyed",$D19)),"1",IF(ISNUMBER(SEARCH("empty",$D19)),
"1",IF(ISNUMBER(SEARCH("forsaken",$D19)),"1",IF(ISNUMBER(SEARCH("isolated",$D19)),"1",IF(ISNUMBER(SEARCH("marooned",$D
19)),"1",IF(ISNUMBER(SEARCH("neglected",$D19)),"1",IF(ISNUMBER(SEARCH("ostracized",$D19)),"1",IF(ISNUMBER(SEARCH("outcast
",$D19)),"1",IF(ISNUMBER(SEARCH("rejected",$D19)),"1",IF(ISNUMBER(SEARCH("shunned",$D19)),"1",IF(ISNUMBER(SEARCH("aliena
ted",$D19)),"1",IF(ISNUMBER(SEARCH("alone",$D19)),"1",IF(ISNUMBER(SEARCH("cheerless",$D19)),"1",IF(ISNUMBER(SEARCH("com
panionless",$D19)),"1",IF(ISNUMBER(SEARCH("dejected",$D19)),"1",IF(ISNUMBER(SEARCH("despondent",$D19)),"1",IF(ISNUMBER(SE
ARCH("estranged",$D19)),"1",IF(ISNUMBER(SEARCH("excluded",$D19)),"1",IF(ISNUMBER(SEARCH("left
out",$D19)),"1",IF(ISNUMBER(SEARCH("lonely",$D19)),"1",IF(ISNUMBER(SEARCH("oppressed",$D19)),"1",IF(ISNUMBER(SEARCH("un
cherished",$D19)),"1",IF(ISNUMBER(SEARCH("detached",$D19)),"1",IF(ISNUMBER(SEARCH("discouraged",$D19)),"1",IF(ISNUMBER(SE
ARCH("distant",$D19)),"1",IF(ISNUMBER(SEARCH("insulated",$D19)),"1",IF(ISNUMBER(SEARCH("melancholy",$D19)),"1",IF(ISNUMBE
R(SEARCH("remote",$D19)),"1",IF(ISNUMBER(SEARCH("withdrawn",$D19)),"1","0")))))))))))))))))))))))))))))))))
=IF(ISNUMBER(SEARCH("abashed",$D19)),"1",IF(ISNUMBER(SEARCH("debased",$D19)),"1",IF(ISNUMBER(SEARCH("degraded",$D19))
,"1",IF(ISNUMBER(SEARCH("delnquent",$D19)),"1",IF(ISNUMBER(SEARCH("depraved",$D19)),"1",IF(ISNUMBER(SEARCH("disgraced",
$D19)),"1",IF(ISNUMBER(SEARCH("evil",$D19)),"1",IF(ISNUMBER(SEARCH("exposed",$D19)),"1",IF(ISNUMBER(SEARCH("humiliated"
,$D19)),"1",IF(ISNUMBER(SEARCH("judged",$D19)),"1",IF(ISNUMBER(SEARCH("mortified",$D19)),"1",IF(ISNUMBER(SEARCH("shame
d",$D19)),"1",IF(ISNUMBER(SEARCH("sinful",$D19)),"1",IF(ISNUMBER(SEARCH("wicked",$D19)),"1",IF(ISNUMBER(SEARCH("wrong",
$D19)),"1",IF(ISNUMBER(SEARCH("ashamed",$D19)),"1",IF(ISNUMBER(SEARCH("contrite",$D19)),"1",IF(ISNUMBER(SEARCH("culpabl
e",$D19)),"1",IF(ISNUMBER(SEARCH("demeaned",$D19)),"1",IF(ISNUMBER(SEARCH("downhearted",$D19)),"1",IF(ISNUMBER(SEARCH
("flustered",$D19)),"1",IF(ISNUMBER(SEARCH("guilty",$D19)),"1",IF(ISNUMBER(SEARCH("penitent",$D19)),"1",IF(ISNUMBER(SEARC
H("regretful",$D19)),"1",IF(ISNUMBER(SEARCH("remorseful",$D19)),"1",IF(ISNUMBER(SEARCH("repentant",$D19)),"1",IF(ISNUMBER(S
EARCH("shamefaced",$D19)),"1",IF(ISNUMBER(SEARCH("sorrowful",$D19)),"1",IF(ISNUMBER(SEARCH("sorry",$D19)),"1",IF(ISNUMB
ER(SEARCH("blushing",$D19)),"1",IF(ISNUMBER(SEARCH("chargrined",$D19)),"1",IF(ISNUMBER(SEARCH("chastened",$D19)),"1",IF(IS
NUMBER(SEARCH("crestfallen",$D19)),"1",IF(ISNUMBER(SEARCH("embarrassed",$D19)),"1",IF(ISNUMBER(SEARCH("hesitant",$D19)),
"1",IF(ISNUMBER(SEARCH("humble",$D19)),"1",IF(ISNUMBER(SEARCH("meek",$D19)),"1",IF(ISNUMBER(SEARCH("regretful",$D19)),
"1",IF(ISNUMBER(SEARCH("reluctant",$D19)),"1",IF(ISNUMBER(SEARCH("sheepish",$D19)),"1","0"))))))))))))))))))))))))))))))))))))))))
Ask_action =IF(ISNUMBER(SEARCH("comment if",$D4)),"1",IF(ISNUMBER(SEARCH("like if",$D4)),"1",IF(ISNUMBER(SEARCH("share
if",$D4)),"1","0")))
Family Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Customer Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert

263 | P a g e
Appendices
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Winemaker Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
image ID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Employee Image* =IF(ISNUMBER(SEARCH("[insertimageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[150,5]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert
imageID#]",$H3)),"1",IF(ISNUMBER(SEARCH("[insert image ID#]",$H3)),"1","0"))))
Friends and fans =IF(ISNUMBER(SEARCH("friends",$D4)),"1",IF(ISNUMBER(SEARCH("fans",$D4)),"1",IF(ISNUMBER(SEARCH("customers",$D4)),"1",IF(
ISNUMBER(SEARCH("supporters",$D4)),"1","0"))))
Family =IF(ISNUMBER(SEARCH("brother",$D4)),"1",IF(ISNUMBER(SEARCH("sister",$D4)),"1",IF(ISNUMBER(SEARCH("daughter",$D4)),"1",IF(
ISNUMBER(SEARCH("cousin",$D4)),"1",IF(ISNUMBER(SEARCH("grandfather",$D4)),"1",IF(ISNUMBER(SEARCH("grandpa",$D4)),"1",IF(
ISNUMBER(SEARCH("pop",$D4)),"1",IF(ISNUMBER(SEARCH("pa",$D4)),"1",IF(ISNUMBER(SEARCH("nan",$D4)),"1",IF(ISNUMBER(S
EARCH("grandmother",$D4)),"1",IF(ISNUMBER(SEARCH("grandma",$D4)),"1",IF(ISNUMBER(SEARCH("mum",$D4)),"1",IF(ISNUMBER(
SEARCH("mother",$D4)),"1",IF(ISNUMBER(SEARCH("generation",$D4)),"1","0")))))))))))))))))

264 | P a g e
Appendices

Appendix D: Kappa Coefficient Calculation


Image Node Source Source Size Kappa
1 animal Picture [290,5] 280x410 pixels 0.06
2 animal Picture [401,5] 720x720 pixels 0.62
3 animal Picture [406,5] 540x720 pixels 0.77
4 celebrity Picture [6,5] 640x640 pixels 0.69
5 competition Picture [352,5] 410x410 pixels 0.96
6 competition Picture [360,5] 410x410 pixels 0.78
7 competition Picture [377,5] 410x410 pixels 0.90
8 competition Picture [384,5] 410x410 pixels 0.69
9 competition Picture [392,5] 410x410 pixels 0.93
10 customer Picture [20,5] 640x480 pixels 0.58
11 customer Picture [29,5] 640x640 pixels 0.51
12 customer Picture [293,5] 640x640 pixels 0.66
13 customer Picture [338,5] 640x640 pixels 0.53
14 customer Picture [407,5] 720x480 pixels 0.48
15 customer Picture [424,5] 720x480 pixels 0.46
16 customer Picture [425,5] 720x480 pixels 0.62
17 customer Picture [465,5] 640x640 pixels 0.87
18 customer Picture [6,5] 640x640 pixels 0.36
19 customer Picture [66,5] 640x640 pixels 0.48
20 customer Picture [82,5] 640x640 pixels 0.79
21 customer Picture [92,5] 720x480 pixels 0.81
22 employee Picture [173,5] 640x640 pixels 0.60
23 employee Picture [192,5] 640x640 pixels 0.47
24 employee Picture [2,5] 640x640 pixels 0.82
25 employee Picture [369,5] 720x480 pixels 0.73
26 employee Picture [438,5] 720x720 pixels 0.71
27 employee Picture [439,5] 720x480 pixels 0.83
28 employee Picture [6,5] 640x640 pixels 0.85
29 event Picture [1,5] 460x680 pixels 0.71
30 event Picture [160,5] 500x720 pixels 0.96
31 event Picture [24,5] 510x720 pixels 0.73
32 event Picture [27,5] 720x500 pixels 0.83
33 event Picture [289,5] 520x720 pixels 0.91
34 event Picture [52,5] 720x420 pixels 0.76
35 food Picture [107,5] 640x640 pixels 0.88
36 food Picture [114,5] 640x640 pixels 0.88
37 food Picture [173,5] 640x640 pixels 0.54
38 food Picture [188,5] 640x640 pixels 0.90
39 food Picture [205,5] 640x640 pixels 0.85
40 food Picture [230,5] 640x640 pixels 0.59
41 food Picture [294,5] 640x640 pixels 0.68
42 food Picture [296,5] 640x640 pixels 0.86
43 food Picture [460,5] 640x640 pixels 0.85
44 food Picture [53,5] 640x640 pixels 0.78
45 food Picture [60,5] 640x640 pixels 0.92
46 food Picture [69,5] 640x640 pixels 0.96
47 food Picture [81,5] 640x640 pixels 0.92
48 food Picture [89,5] 640x640 pixels 0.45
49 food Picture [400,5] 720x560 pixels 0.93
50 food Picture [21,5] 640x640 pixels 0.88
51 food Picture [437,5] 720x720 pixels 0.96
52 fun/interesting fact Picture [381,5] 720x470 pixels 0.83
53 fun/interesting fact Picture [451,5] 480x640 pixels 0.96

265 | P a g e
Appendices
54 fun/interesting fact Picture [464,5] 600x240 pixels 0.98
55 fun/interesting fact Picture [292,5] 640x640 pixels 0.72
56 fun/interesting fact Picture [316,5] 640x640 pixels 0.98
57 fun/interesting fact Picture [32,5] 640x640 pixels 0.82
58 fun/interesting fact Picture [334,5] 640x640 pixels 0.78
59 fun/interesting fact Picture [348,5] 640x640 pixels 0.83
60 fun/interesting fact Picture [381,5] 720x470 pixels 0.96
61 fun/interesting fact Picture [405,5] 220x320 pixels 0.93
62 fun/interesting fact Picture [414,5] 390x430 pixels 0.93
63 fun/interesting fact Picture [435,5] 480x640 pixels 0.70
64 meme_cartoon Picture [341,5] 640x640 pixels 0.90
65 meme_cartoon Picture [447,5] 270x190 pixels 0.87
66 product Picture [326,5] 720x720 pixels 0.86
67 product Picture [152,5] 640x640 pixels 0.87
68 product Picture [29,5] 640x640 pixels 0.69
69 product Picture [295,5] 640x640 pixels 0.84
70 product Picture [301,5] 640x640 pixels 0.74
71 product Picture [313,5] 640x640 pixels 0.81
72 product Picture [319,5] 640x640 pixels 0.90
73 product Picture [359,5] 640x640 pixels 0.93
74 product Picture [431,5] 520x520 pixels 0.80
75 product Picture [438,5] 720x720 pixels 0.80
76 product Picture [453,5] 410x310 pixels 0.83
77 product Picture [460,5] 640x640 pixels 0.84
78 product Picture [82,5] 640x640 pixels 0.74
79 review_award Picture [192,5] 640x640 pixels 0.93
80 review_award Picture [241,5] 640x640 pixels 0.73
81 review_award Picture [298,5] 300x120 pixels 0.89
82 review_award Picture [342,5] 540x640 pixels 0.86
83 scenic Picture [471,5] 720x380 pixels 0.89
84 vineyard Picture [190,5] 640x640 pixels 0.94
85 vineyard Picture [25,5] 640x480 pixels 0.87
86 vineyard Picture [293,5] 640x640 pixels 0.71
87 vineyard Picture [338,5] 640x640 pixels 0.84
88 vineyard Picture [448,5] 500x650 pixels 0.83
89 vineyard Picture [82,5] 640x640 pixels 0.66
90 winery Picture [12,5] 720x370 pixels 0.91
91 winery Picture [126,5] 640x640 pixels 0.78
92 winery Picture [136,5] 640x640 pixels 0.86
93 winery Picture [173,5] 640x640 pixels 0.48
94 winery Picture [179,5] 640x640 pixels 0.88
95 winery Picture [211,5] 560x720 pixels 0.78
96 winery Picture [212,5] 720x480 pixels 0.80
97 winery Picture [213,5] 720x450 pixels 0.89
98 winery Picture [228,5] 640x640 pixels -0.12
99 winery Picture [274,5] 640x640 pixels 0.91
100 winery Picture [29,5] 640x640 pixels 0.85
Mean 0.77
Median 0.83

266 | P a g e
Appendices

Appendix E: Binary Logistic Regression Results (Interactions)

Appendix E.1. Logistic regression results showing effect of content interactions on


CREATING SMEB

SMEB Interaction b Sig Exp(B)


Creating Entertainment (1) by Information (1) -.771 .009 .463
behaviour Entertainment (1) by Information (2) -.484 .121 .617
Entertainment (1) by Information (3) -.537 .125 .585
Entertainment (1) by Information (4) -1.194 .002 .303
Entertainment (1) by Information (5) -1.427 .004 .240
Entertainment (1) by Information (6) -.851 .250 .427
Entertainment (1) by Information (7) -1.592 .102 .204
Entertainment (1) by Information (8) -3.288 .044 .037
Entertainment (1) by Information (9) 20.514 1.000 811204446.677
Entertainment (1) by Information (10) -21.779 1.000 .000
Entertainment (1) by Information (11) -21.249 1.000 .000
Entertainment (2) by Information (1) -.096 .821 .908
Entertainment (2) by Information (2) .043 .922 1.044
Entertainment (2) by Information (3) -.122 .802 .885
Entertainment (2) by Information (4 -.814 .140 .443
Entertainment (2) by Information (5) -.623 .318 .536
Entertainment (2) by Information (6) -.622 .516 .537
Entertainment (2) by Information (7) -1.060 .382 .346
Entertainment (2) by Information (9) -21.792 1.000 .000
Entertainment (3) by Information (1) -1.270 .189 .281
Entertainment (3) by Information (2) -.657 .559 .518
Entertainment (3) by Information (3) -1.184 .285 .306
Entertainment (3) by Information (4) -1.687 .125 .185
Entertainment (3) by Information (5) -1.126 .448 .324
Entertainment (3) by Information (6) -23.612 1.000 .000
Entertainment (3) by Information (7) 18.312 1.000 89715073.346
Entertainment (3) by Information (9) 19.464 1.000 283965469.173
Entertainment (4) by Information (1) 41.585 .999 4058346530.000
Entertainment (4) by Information (2) 20.496 1.000 796640413.821
Entertainment (4) by Information (3) 21.414 1.000 1995287943.194
Entertainment (4) by Information (4) 40.876 .999 0890183040.000
Entertainment (4) by Information (5) 19.467 1.000 284768480.040
Entertainment (5) by Information (2) 21.447 .999 2061924092.082
Entertainment (5) by Information (6) 20.720 1.000 997229921.602
Remunerative (1) by Information (1) -.056 .913 .945
Remunerative (2) by Information (1) 20.921 .999 1218518809.291
Remunerative (3) by Information (1) -21.534 1.000 .000
Remunerative (1) by Information (2) .123 .807 1.131
Remunerative (2) by Information (2) 22.820 .999 8139521364.930
Remunerative (3) by Information (2) -21.236 1.000 .000
Remunerative (1) by Information (3) .103 .853 1.108
Remunerative (2) by Information (3) 20.409 .999 730052762.602
Remunerative (1) by Information (4) .326 .573 1.385
Remunerative (2) by Information (4) 21.222 .999 1646651799.617

267 | P a g e
Appendices
Remunerative (1) by Information (5) .370 .540 1.448
Remunerative (2) by Information (5) 21.794 .999 2917578857.558
Remunerative (3) by Information (5) 21.086 1.000 1437922889.289
Remunerative (1) by Information (6) .537 .586 1.711
Remunerative (2) by Information (6) 20.372 .999 703551923.446
Remunerative (1) by Information (7) -1.071 .297 .343
Remunerative (2) by Information (7) 21.089 .999 1441711487.591
Remunerative (1) by Information (8) -22.729 .999 .000
Remunerative (2) by Information (8) 40.502 .999 3925934530.000
Remunerative (4) by Information (8) -22.327 1.000 .000
Remunerative (1) by Information (9) -21.380 .999 .000
Remunerative (2) by Information (9) -21.504 1.000 .000
Remunerative (1) by Information (10) -42.431 .999 .000
Relational (1) by Information (1) -.015 .960 .985
Relational (2) by Information (1) -.087 .820 .917
Relational (3) by Information (1) -.227 .734 .797
Relational (4) by Information (1) .265 .824 1.303
Relational (1) by Information (1) -20.983 1.000 .000
Relational (1) by Information (2) .259 .435 1.295
Relational (2) by Information (2) -.104 .796 .902
Relational (3) by Information (2) -.733 .283 .480
Relational (4) by Information (2) -.539 .642 .583
Relational (5) by Information (2) -20.248 1.000 .000
Relational (1) by Information (3) -.018 .962 .982
Relational (2) by Information (3) -.339 .454 .712
Relational (3) by Information (3) -1.368 .062 .255
Relational (4) by Information (3) -.382 .755 .682
Relational (5) by Information (3) -21.890 1.000 .000
Relational (6) by Information (3) -.907 .471 .404
Relational (1) by Information (4) -.044 .918 .957
Relational (2) by Information (4) -.128 .794 .880
Relational (3) by Information (4) -.748 .339 .473
Relational (4) by Information (4) -.806 .531 .447
Relational (5) by Information (4) -22.788 1.000 .000
Relational (1) by Information (5) -.103 .872 .902
Relational (2) by Information (5) -1.048 .103 .351
Relational (3) by Information (5) -1.969 .135 .140
Relational (4) by Information (5) -.311 .827 .733
Relational (5) by Information (5) -23.543 1.000 .000
Relational (7) by Information (5) 20.747 1.000 1024173452.320
Relational (1) by Information (6) 1.655 .189 5.231
Relational (2) by Information (6) .153 .913 1.166
Relational (3) by Information (6) 21.124 .999 1492898486.423
Relational (4) by Information (6) 22.000 .999 3583349865.029
Relational (5) by Information (6) .835 1.000 2.306
Relational (6) by Information (6) 1.001 .585 2.721
Relational (1) by Information (7) -.261 .823 .770
Relational (2) by Information (7) .134 .917 1.143
Relational (3) by Information (7) -1.019 .439 .361
Relational (4) by Information (7) 20.493 .999 794037046.922
Relational (5) by Information (7) -.740 1.000 .477
Relational (1) by Information (8) 1.036 .405 2.818
268 | P a g e
Appendices
Relational (2) by Information (8) -.616 .626 .540
Relational (3) by Information (8) 20.821 1.000 1102575070.980
Relational (4) by Information (8) 21.668 1.000 2572305642.047
Relational (6) by Information (8) -20.444 1.000 .000
Relational (1) by Information (9) 21.522 1.000 2223434339.096
Relational (2) by Information (9) -.289 .843 .749
Relational (3) by Information (9) -21.651 1.000 .000
Relational (4) by Information (9) 21.205 1.000 1618533567.831
Relational (2) by Information (10) -42.981 .999 .000
Relational (3) by Information (10) -43.363 .999 .000
Entertainment(1) By Relational(1) .220 .357 1.247
Entertainment(1) By Relational(2) -.347 .198 .707
Entertainment(1) By Relational(3) .457 .262 1.579
Entertainment(1) By Relational(4) -1.121 .064 .326
Entertainment(1) By Relational(5) -1.431 .281 .239
Entertainment(1) By Relational(6) -.612 .491 .542
Entertainment(2) By Relational(1) -.217 .529 .805
Entertainment(2) By Relational(2) -.935 .012 .392
Entertainment(2) By Relational(3) .041 .934 1.042
Entertainment(2) By Relational(4) -.904 .258 .405
Entertainment(2) By Relational(5) -1.260 .378 .284
Entertainment(2) By Relational(7) 21.110 1.000 1472026196.930
Entertainment(3) By Relational(1) 1.863 .013 6.441
Entertainment(3) By Relational(2) .572 .446 1.773
Entertainment(3) By Relational(3) 1.050 .309 2.858
Entertainment(3) By Relational(4) 21.106 .999 1466610850.838
Entertainment(3) By Relational(5) -1.335 .477 .263
Entertainment(4) By Relational(1) 20.978 .999 1289909361.940
Entertainment(4) By Relational(2) .152 .919 1.165
Entertainment(4) By Relational(3) 1.448 .301 4.255
Entertainment(5) By Relational(1) 21.557 1.000 2300994335.269
Entertainment(5) By Relational(2) 20.902 1.000 1195502793.450
Entertainment(5) By Relational(3) 21.651 1.000 2528984142.325
Entertainment(1) By Remunerative(1) -.605 .057 .546
Entertainment(1) By Remunerative(2) .190 .773 1.209
Entertainment(1) By Remunerative(3) 42.440 .999 5032539100.000
Entertainment(2) By Remunerative(1) -.271 .487 .763
Entertainment(2) By Remunerative(2) .634 .476 1.884
Entertainment(3) By Remunerative(1) 20.396 .999 720660656.928
Entertainment(3) By Remunerative(2) 20.648 1.000 927585820.260
Entertainment(4) By Remunerative(1) 19.903 .999 440504729.752
Entertainment(5) By Remunerative(1) -.419 1.000 .658
Relational(1) By Remunerative(1) .072 .849 1.075
Relational(1) By Remunerative(2) .007 .991 1.007
Relational(2) By Remunerative(1) -.206 .618 .814
Relational(2) By Remunerative(2) -.062 .929 .940
Relational(2) By Remunerative(3) -21.585 .999 .000
Relational(2) By Remunerative(4) -20.935 1.000 .000
Relational(3) By Remunerative(1) -.289 .552 .749
Relational(3) By Remunerative(2) .252 .806 1.287
Relational(3) By Remunerative(3) 20.916 1.000 1212183296.915
Relational(4) By Remunerative(1) 1.518 .190 4.561
269 | P a g e
Appendices
Relational(4) By Remunerative(2) -.884 .516 .413
Relational(5) By Remunerative(1) 20.832 .999 1115051199.143
Relational(6) By Remunerative(1) -20.510 1.000 .000
Relational(7) By Remunerative(2) 21.499 1.000 2173190758.023

270 | P a g e
Appendices
Appendix E.2. Logistic regression results showing effect of content interactions and
CONTRIBUTING (SHARES) SMEB

SMEB Interaction b Sig Exp(B)


Contributing Entertainment (1) by Information (1) -.430 .328 .650
behaviour Entertainment (1) by Information (2) .006 .989 1.006
(shares) Entertainment (1) by Information (3) -.213 .746 .808
Entertainment (1) by Information (4) -0.90 .021 .401
Entertainment (1) by Information (5) 17.049 .998 25365590.234
Entertainment (1) by Information (6) -.268 1.000 .765
Entertainment (1) by Information (7) -.651 1.000 .521
Entertainment (2) by Information (1) -.990 .289 .372
Entertainment (2) by Information (2) -.190 .858 .827
Entertainment (2) by Information (3) -.614 .640 .541
Entertainment (2) by Information (4 15.743 .998 6871292.316
Entertainment (2) by Information (5) 15.651 .998 6268298.901
Entertainment (2) by Information (6) -1.200 1.000 .301
Entertainment (2) by Information (7) -1.615 1.000 .199
Entertainment (3) by Information (1) -20.837 .999 .000
Entertainment (3) by Information (2) -1.042 1.000 .353
Entertainment (3) by Information (3) -1.422 1.000 .241
Entertainment (3) by Information (4) -2.893 1.000 .055
Entertainment (3) by Information (5) -3.062 1.000 .047
Entertainment (3) by Information (6) -19.988 1.000 .000
Entertainment (3) by Information (7) -20.407 1.000 .000
Entertainment (4) by Information (1) -.922 1.000 .398
Entertainment (4) by Information (2) -20.677 1.000 .000
Entertainment (4) by Information (3) -.721 1.000 .486
Entertainment (4) by Information (4) -1.117 1.000 .327
Entertainment (4) by Information (5) -1.609 1.000 .200
Entertainment (5) by Information (2) 19.851 .999 417948960.164
Entertainment (5) by Information (6) .283 1.000 1.327
Information(1) By Remunerative(1) .517 .339 1.677
Information(1) By Remunerative(2) 1.320 .353 3.742
Information(1) By Remunerative(3) -20.963 1.000 .000
Information(2) By Remunerative(1) .809 .128 2.247
Information(2) By Remunerative(2) .988 .494 2.685
Information(2) By Remunerative(3) -21.171 1.000 .000
Information(3) By Remunerative(1) .666 .253 1.947
Information(3) By Remunerative(2) .259 .847 1.296
Information(4) By Remunerative(1) .019 .974 1.020
Information(4) By Remunerative(2) .868 .540 2.383
Information(5) By Remunerative(1) .534 .394 1.705
Information(5) By Remunerative(2) .907 .546 2.476
Information(5) By Remunerative(3) -21.471 1.000 .000
Information(6) By Remunerative(1) 1.125 .258 3.081
Information(6) By Remunerative(2) 21.358 .999 1886907904.100
Information(7) By Remunerative(1) .171 .873 1.186
Information(7) By Remunerative(2) 20.662 .999 940632528.112
Information(8) By Remunerative(1) -22.270 .999 .000
Information(8) By Remunerative(2) -22.630 1.000 .000

271 | P a g e
Appendices
Information(8) By Remunerative(4) 19.882 1.000 430963907.721
Information(9) By Remunerative(1) .173 .933 1.188
Information(9) By Remunerative(2) -21.390 1.000 .000
Information(10) By Remunerative(1) .127 1.000 1.136
Information(1) By Relational(1) .090 .777 1.094
Information(1) By Relational(2) -.716 .04 .488
Information(1) By Relational(3) -.312 .637 .732
Information(1) By Relational(4) 21.425 .999 2016353213.212
Information(1) By Relational(5) -.496 1.000 .609
Information(2) By Relational(1) .462 .179 1.588
Information(2) By Relational(2) -.478 .243 .620
Information(2) By Relational(3) .277 .685 1.320
Information(2) By Relational(4) 20.076 .999 523733854.554
Information(2) By Relational(5) 20.413 1.000 733385357.463
Information(3) By Relational(1) .321 .412 1.379
Information(3) By Relational(2) -.262 .568 .770
Information(3) By Relational(3) .658 .375 1.931
Information(3) By Relational(4) 20.539 .999 831318578.002
Information(3) By Relational(5) 42.109 .999 1513532160.000
Information(3) By Relational(6) -21.009 .999 .000
Information(4) By Relational(1) .020 .965 1.020
Information(4) By Relational(2) -.552 .267 .576
Information(4) By Relational(3) .577 .483 1.780
Information(4) By Relational(4) 20.470 .999 776182902.512
Information(4) By Relational(5) 19.174 1.000 212347462.754
Information(5) By Relational(1) -.294 .644 .745
Information(5) By Relational(2) -.869 .180 .419
Information(5) By Relational(3) -.322 .733 .724
Information(5) By Relational(4) 20.480 .999 784207042.220
Information(5) By Relational(5) 18.607 1.000 120441271.026
Information(5) By Relational(7) 20.559 1.000 848242512.026
Information(6) By Relational(1) .650 .563 1.915
Information(6) By Relational(2) -1.604 .232 .201
Information(6) By Relational(3) .123 .947 1.130
Information(6) By Relational(4) 21.528 .999 2235148173.166
Information(6) By Relational(5) 20.647 1.000 926456750.517
Information(6) By Relational(6) 21.148 .999 1528696840.379
Information(7) By Relational(1) -21.405 .999 .000
Information(7) By Relational(2) -20.033 .999 .000
Information(7) By Relational(3) -20.440 .999 .000
Information(7) By Relational(4) -.748 1.000 .473
Information(7) By Relational(5) -1.513 1.000 .220
Information(8) By Relational(1) .154 .902 1.166
Information(8) By Relational(2) 1.039 .409 2.826
Information(8) By Relational(3) 21.946 1.000 3398012563.561
Information(8) By Relational(4) 42.261 .999 1221987070.000
Information(8) By Relational(6) -19.812 1.000 .000
Information(9) By Relational(1) -20.529 1.000 .000
Information(9) By Relational(2) .075 .959 1.078
Information(9) By Relational(3) -20.434 1.000 .000
Information(9) By Relational(4) 42.545 .999 8330573300.000
Information(10) By Relational(2) -.739 1.000 .478
272 | P a g e
Appendices
Information(10) By Relational(3) -.035 1.000 .966
Entertainment(1) By Relational(1) .262 .276 1.300
Entertainment(1) By Relational(2) .126 .643 1.134
Entertainment(1) By Relational(3) .405 .322 1.500
Entertainment(1) By Relational(4) .114 .847 1.121
Entertainment(1) By Relational(5) -.501 .670 .606
Entertainment(1) By Relational(6) -.394 .660 .675
Entertainment(2) By Relational(1) .307 .369 1.360
Entertainment(2) By Relational(2) -.028 .939 .972
Entertainment(2) By Relational(3) 1.470 .935 4.349
Entertainment(2) By Relational(4) .086 .912 1.090
Entertainment(2) By Relational(5) .127 .917 1.135
Entertainment(2) By Relational(7) 21.512 1.000 2201126659.893
Entertainment(3) By Relational(1) .372 .581 1.450
Entertainment(3) By Relational(2) 1.661 .067 5.265
Entertainment(3) By Relational(3) 1.472 .229 4.357
Entertainment(3) By Relational(4) 20.926 .999 1224148851.646
Entertainment(3) By Relational(5) -21.695 .999 .000
Entertainment(4) By Relational(1) .390 .816 1.478
Entertainment(4) By Relational(2) -.263 .863 .769
Entertainment(4) By Relational(3) 2.362 .099 10.613
Entertainment(5) By Relational(1) 21.421 1.000 2009808964.911
Entertainment(5) By Relational(2) 21.535 1.000 2250990891.306
Entertainment(5) By Relational(3) 21.772 1.000 2853507358.773
Entertainment(1) By Remunerative(1) .149 .641 1.160
Entertainment(1) By Remunerative(2) -.770 .239 .463
Entertainment(1) By Remunerative(3) -.035 1.000 .966
Entertainment(2) By Remunerative(1) .653 .110 1.921
Entertainment(2) By Remunerative(2) -.043 .961 .957
Entertainment(3) By Remunerative(1) -.631 .622 .532
Entertainment(3) By Remunerative(2) 19.214 1.000 221140991.037
Entertainment(4) By Remunerative(1) -.687 .655 .503
Entertainment(5) By Remunerative(1) -.553 1.000 .575
Relational(1) By Remunerative(1) -.104 .785 .901
Relational(1) By Remunerative(2) .561 .446 1.753
Relational(2) By Remunerative(1) -.155 .708 .856
Relational(2) By Remunerative(2) -.248 .724 .780
Relational(2) By Remunerative(3) -21.206 .999 .000
Relational(2) By Remunerative(4) 21.560 1.000 2309593774.072
Relational(3) By Remunerative(1) -.627 .197 .534
Relational(3) By Remunerative(2) .885 .472 2.424
Relational(3) By Remunerative(3) -21.328 1.000 .000
Relational(4) By Remunerative(1) 1.313 .257 3.719
Relational(4) By Remunerative(2) -1.170 .392 .310
Relational(5) By Remunerative(1) -.264 .809 .768
Relational(6) By Remunerative(1) -20.590 1.000 .000
Relational(7) By Remunerative(2) 20.952 1.000 1256481593.414

273 | P a g e
Appendices
Appendix E.3. Logistic regression results showing effect of content interactions on
CONTRIBUTING (LIKES) SMEB

SMEB Interaction b Sig Exp(B)


Contributing Entertainment (1) by Information (1) -.430 .328 .650
behaviour Entertainment (1) by Information (2) .006 .989 1.006
(likes) Entertainment (1) by Information (3) -.213 .746 .808
Entertainment (1) by Information (4) -1.044 .287 .352
Entertainment (1) by Information (5) 17.049 .998 25365590.234
Entertainment (2) by Information (1) -.990 .289 .372
Entertainment (2) by Information (2) -.190 .858 .827
Entertainment (2) by Information (3) -.614 .640 .541
Entertainment (2) by Information (4 15.743 .998 6871292.316
Entertainment (2) by Information (5) 15.651 .998 6268298.901
Entertainment (3) by Information (1) -20.837 .999 .000
Entertainment (3) by Information (2) -1.042 1.000 .353
Entertainment (3) by Information (3) -1.422 1.000 .241
Entertainment (3) by Information (4) -2.893 1.000 .055
Entertainment (3) by Information (5) -3.062 1.000 .047
Entertainment (4) by Information (1) -.922 1.000 .398
Entertainment (4) by Information (2) -20.677 1.000 .000
Entertainment (4) by Information (3) -.721 1.000 .486
Entertainment (4) by Information (4) -1.117 1.000 .327
Entertainment (4) by Information (5) -1.609 1.000 .200
Entertainment (5) by Information (2) 19.851 .999 417948960.164
Information(1) By Remunerative(1) .517 .339 1.677
Information(1) By Remunerative(2) 1.320 .353 3.742
Information(1) By Remunerative(3) -20.963 1.000 .000
Information(2) By Remunerative(1) .809 .128 2.247
Information(2) By Remunerative(2) .988 .494 2.685
Information(2) By Remunerative(3) -21.171 1.000 .000
Information(3) By Remunerative(1) .666 .253 1.947
Information(3) By Remunerative(2) .259 .847 1.296
Information(4) By Remunerative(1) .019 .974 1.020
Information(4) By Remunerative(2) .868 .540 2.383
Information(5) By Remunerative(1) .534 .394 1.705
Information(5) By Remunerative(2) .907 .546 2.476
Information(5) By Remunerative(3) -21.471 1.000 .000
Information(6) By Remunerative(1) 1.125 .258 3.081
Information(6) By Remunerative(2) 21.358 .999 1886907904.100
Information(7) By Remunerative(1) .171 .873 1.186
Information(7) By Remunerative(2) 20.662 .999 940632528.112
Information(8) By Remunerative(1) -22.270 .999 .000
Information(8) By Remunerative(2) -22.630 1.000 .000
Information(8) By Remunerative(4) 19.882 1.000 430963907.721
Information(9) By Remunerative(1) .173 .933 1.188
Information(9) By Remunerative(2) -21.390 1.000 .000
Information(10) By Remunerative(1) .127 1.000 1.136
Information(1) By Relational(1) .090 .777 1.094
Information(1) By Relational(2) -.716 .065 .488
Information(1) By Relational(3) -.312 .637 .732

274 | P a g e
Appendices
Information(1) By Relational(4) 21.425 .999 2016353213.212
Information(1) By Relational(5) -.496 1.000 .609
Information(2) By Relational(1) .462 .179 1.588
Information(2) By Relational(2) -.478 .243 .620
Information(2) By Relational(3) .277 .685 1.320
Information(2) By Relational(4) 20.076 .999 523733854.554
Information(2) By Relational(5) 20.413 1.000 733385357.463
Information(3) By Relational(1) .321 .412 1.379
Information(3) By Relational(2) -.262 .568 .770
Information(3) By Relational(3) .658 .375 1.931
Information(3) By Relational(4) 20.539 .999 831318578.002
Information(3) By Relational(5) 42.109 .999 1939932160.000
Information(3) By Relational(6) -21.009 .999 .000
Information(4) By Relational(1) .020 .965 1.020
Information(4) By Relational(2) -.552 .267 .576
Information(4) By Relational(3) .577 .483 1.780
Information(4) By Relational(4) 20.470 .999 776182902.512
Information(4) By Relational(5) 19.174 1.000 212347462.754
Information(5) By Relational(1) -.294 .644 .745
Information(5) By Relational(2) -.869 .180 .419
Information(5) By Relational(3) -.322 .733 .724
Information(5) By Relational(4) 20.480 .999 784207042.220
Information(5) By Relational(5) 18.607 1.000 120441271.026
Information(5) By Relational(7) 20.559 1.000 848242512.026
Information(6) By Relational(1) .650 .563 1.915
Information(6) By Relational(2) -1.604 .232 .201
Information(6) By Relational(3) .123 .947 1.130
Information(6) By Relational(4) 21.528 .999 2235148173.166
Information(6) By Relational(5) 20.647 1.000 926456750.517
Information(6) By Relational(6) 21.148 .999 1528696840.379
Information(7) By Relational(1) -21.405 .999 .000
Information(7) By Relational(2) -20.033 .999 .000
Information(7) By Relational(3) -20.440 .999 .000
Information(7) By Relational(4) -.748 1.000 .473
Information(7) By Relational(5) -1.513 1.000 .220
Information(8) By Relational(1) .154 .902 1.166
Information(8) By Relational(2) 1.039 .409 2.826
Information(8) By Relational(3) 21.946 1.000 3398012563.561
Information(8) By Relational(4) 42.261 .999 2258087070.000
Information(8) By Relational(6) -19.812 1.000 .000
Information(9) By Relational(1) -20.529 1.000 .000
Information(9) By Relational(2) .075 .959 1.078
Information(9) By Relational(3) -20.434 1.000 .000
Information(9) By Relational(4) 42.545 .999 2998480300.000
Information(10) By Relational(2) -.739 1.000 .478
Information(10) By Relational(3) -.035 1.000 .966
Entertainment(1) By Relational(1) .510 .218 1.665
Entertainment(1) By Relational(2) -.567 .245 .567
Entertainment(1) By Relational(3) .091 .913 1.096
Entertainment(1) By Relational(4) -.485 .742 .616
Entertainment(1) By Relational(5) -.309 1.000 .734
Entertainment(1) By Relational(6) 19.040 .999 185687176.039
275 | P a g e
Appendices
Entertainment(2) By Relational(1) -.068 .934 .935
Entertainment(2) By Relational(2) .070 .953 1.073
Entertainment(2) By Relational(3) 17.426 .998 36973093.888
Entertainment(2) By Relational(4) -2.622 .098 .073
Entertainment(2) By Relational(5) -.969 1.000 .380
Entertainment(2) By Relational(7) 19.408 1.000 268367406.485
Entertainment(3) By Relational(1) .692 .577 1.997
Entertainment(3) By Relational(2) 18.159 .999 77013588.058
Entertainment(3) By Relational(3) 18.468 .999 104875326.101
Entertainment(3) By Relational(4) 17.928 .999 61088986.07
Entertainment(3) By Relational(5) -.805 1.000 .447
Entertainment(4) By Relational(1) 18.972 .999 173576131.543
Entertainment(4) By Relational(2) 18.783 .999 143726600.720
Entertainment(4) By Relational(3) 18.822 .999 149443166.801
Entertainment(5) By Relational(1) 19.038 1.000 185314080.521
Entertainment(5) By Relational(2) 18.741 1.000 137750631.092
Entertainment(5) By Relational(3) 19.664 1.000 346762256.487
Entertainment(1) By Remunerative(1) -.909 .125 .403
Entertainment(1) By Remunerative(2) 17.108 .999 26910700.667
Entertainment(1) By Remunerative(3) 1.330 1.000 3.783
Entertainment(2) By Remunerative(1) 17.298 .998 32539977.736
Entertainment(2) By Remunerative(2) 16.630 .999 16682424.378
Entertainment(3) By Remunerative(1) 17.585 .999 43336539.807
Entertainment(3) By Remunerative(2) 17.053 1.000 25468628.371
Entertainment(4) By Remunerative(1) 17.464 1.000 38421500.162
Entertainment(5) By Remunerative(1) -1.522 1.000 .218
Relational(1) By Remunerative(1) 1.548 .041 4.700
Relational(1) By Remunerative(2) -18.590 .998 .000
Relational(2) By Remunerative(1) .340 .659 1.405
Relational(2) By Remunerative(2) -.425 1.000 .654
Relational(2) By Remunerative(3) 17.493 1.000 39529477.377
Relational(2) By Remunerative(4) 19.317 1.000 244953008.929
Relational(3) By Remunerative(1) -.065 .947 .937
Relational(3) By Remunerative(2) -.692 1.000 .500
Relational(3) By Remunerative(3) 18.680 1.000 129598517.55
Relational(4) By Remunerative(1) 18.635 .999 123867843.175
Relational(4) By Remunerative(2) -.220 1.000 .803
Relational(5) By Remunerative(1) .350 1.000 1.419
Relational(6) By Remunerative(1) 1.130 1.000 3.095
Relational(7) By Remunerative(2) 1.496 1.000 4.464

276 | P a g e
Appendices
Appendix E.4. Logistic regression results showing effect of content interactions on
CONSUMING SMEB.

SMEB Interaction b Sig Exp(B)


Consuming Entertainment (1) by Information (1) -.195 .849 .823
behaviour Entertainment (1) by Information (2) .397 .719 1.488
Entertainment (1) by Information (3) .356 .800 1.428
Entertainment (1) by Information (4) -16.646 .997 .000
Entertainment (1) by Information (5) .244 1.000 1.277
Entertainment (1) by Information (6) .342 1.000 1.408
Entertainment (1) by Information (7) .268 1.000 1.308
Entertainment (1) by Information (8) .263 1.000 1.301
Entertainment (1) by Information (9) -.026 1.000 .975
Entertainment (1) by Information (10) -.333 1.000 .717
Entertainment (1) by Information (11) 17.448 1.000 37815558.801
Entertainment (2) by Information (1) -1.201 .401 .301
Entertainment (2) by Information (2) -1.125 .435 .325
Entertainment (2) by Information (3) 16.569 .998 15703909.770
Entertainment (2) by Information (4 -.308 1.000 .735
Entertainment (2) by Information (5) -.195 1.000 .823
Entertainment (2) by Information (6) .038 1.000 1.039
Entertainment (2) by Information (7) -.522 1.000 .593
Entertainment (2) by Information (9) -.868 1.000 .420
Entertainment (3) by Information (1) -18.938 .999 .000
Entertainment (3) by Information (2) -.414 1.000 .661
Entertainment (3) by Information (3) -.409 1.000 .664
Entertainment (3) by Information (4) -17.341 .999 .000
Entertainment (3) by Information (5) -17.651 .999 .000
Entertainment (3) by Information (6) -35.038 .999 .000
Entertainment (3) by Information (7) -17.553 1.000 .000
Entertainment (3) by Information (9) -18.102 1.000 .000
Entertainment (4) by Information (1) -.568 1.000 .567
Entertainment (4) by Information (2) -1.664 1.000 .189
Entertainment (4) by Information (3) -.579 1.000 .560
Entertainment (4) by Information (4) -16.845 1.000 .000
Entertainment (4) by Information (5) -16.970 1.000 .000
Entertainment (5) by Information (2) 17.274 .999 31766637.247
Entertainment (5) by Information (6) -.232 1.000 .793
Information(1) By Remunerative(1) 18.681 .997 129674677.870
Information(1) By Remunerative(2) .190 1.000 1.209
Information(1) By Remunerative(3) 15.858 1.000 7707246.835
Information(2) By Remunerative(1) 19.049 .997 187537688.763
Information(2) By Remunerative(2) .707 1.000 2.028
Information(2) By Remunerative(3) 16.049 1.000 9332595.986
Information(3) By Remunerative(1) 18.588 .998 118191845.404
Information(3) By Remunerative(2) .283 1.000 1.328
Information(4) By Remunerative(1) 17.813 .998 54454583.815
Information(4) By Remunerative(2) -.987 1.000 .373
Information(5) By Remunerative(1) 1.848 1.000 6.347
Information(5) By Remunerative(2) -16.721 1.000 .000
Information(5) By Remunerative(3) -.699 1.000 .497

277 | P a g e
Appendices
Information(6) By Remunerative(1) 1.424 1.000 4.154
Information(6) By Remunerative(2) -17.137 1.000 .000
Information(7) By Remunerative(1) 1.366 1.000 3.920
Information(7) By Remunerative(2) -16.644 1.000 .000
Information(8) By Remunerative(1) 1.333 1.000 3.791
Information(8) By Remunerative(2) -16.325 1.000 .000
Information(8) By Remunerative(4) -1.429 1.000 .240
Information(9) By Remunerative(1) 1.887 1.000 6.603
Information(9) By Remunerative(2) -17.114 1.000 .000
Information(10) By Remunerative(1) .872 1.000 2.391
Information(1) By Relational(1) -1.223 .260 .294
Information(1) By Relational(2) -17.769 .997 .000
Information(1) By Relational(3) 17.954 .997 62719604.481
Information(1) By Relational(4) -1.161 1.000 .313
Information(1) By Relational(5) -1.878 1.000 .153
Information(2) By Relational(1) .154 .895 1.167
Information(2) By Relational(2) .185 1.000 1.203
Information(2) By Relational(3) 18.862 .998 155501741.794
Information(2) By Relational(4) -.294 1.000 .745
Information(2) By Relational(5) -.756 1.000 .469
Information(3) By Relational(1) -.284 .849 .753
Information(3) By Relational(2) -.203 1.000 .817
Information(3) By Relational(3) 18.376 .998 95606034.041
Information(3) By Relational(4) -.588 1.000 .555
Information(3) By Relational(5) -1.284 1.000 .277
Information(3) By Relational(6) 16.976 .999 23571316.746
Information(4) By Relational(1) -1.545 1.000 .213
Information(4) By Relational(2) -35.160 .996 .000
Information(4) By Relational(3) .894 1.000 2.444
Information(4) By Relational(4) -18.412 .999 .000
Information(4) By Relational(5) -19.002 1.000 .000
Information(5) By Relational(1) -1.094 1.000 .335
Information(5) By Relational(2) -17.636 .999 .000
Information(5) By Relational(3) 1.585 1.000 4.877
Information(5) By Relational(4) -18.017 .999 .000
Information(5) By Relational(5) -18.597 1.000 .000
Information(5) By Relational(7) 1.576 1.000 4.837
Information(6) By Relational(1) -.997 1.000 .369
Information(6) By Relational(2) -18.018 .999 .000
Information(6) By Relational(3) .880 1.000 2.410
Information(6) By Relational(4) -17.943 1.000 .000
Information(6) By Relational(5) -19.020 1.000 .000
Information(6) By Relational(6) -1.168 1.000 .311
Information(7) By Relational(1) -.902 1.000 .406
Information(7) By Relational(2) -18.012 .999 .000
Information(7) By Relational(3) .573 1.000 1.774
Information(7) By Relational(4) -18.515 1.000 .000
Information(7) By Relational(5) -19.133 1.000 .000
Information(8) By Relational(1) 17.302 .999 32656415.652
Information(8) By Relational(2) -.280 1.000 .756
Information(8) By Relational(3) 19.614 1.000 329817069.544
Information(8) By Relational(4) -1.010 1.000 .364
278 | P a g e
Appendices
Information(8) By Relational(6) 18.394 1.000 97403883.825
Information(9) By Relational(1) 16.034 1.000 9194045.947
Information(9) By Relational(2) -1.041 1.000 .353
Information(9) By Relational(3) 18.579 1.000 117128915.728
Information(9) By Relational(4) -1.660 1.000 .190
Information(10) By Relational(2) -19.021 1.000 .000
Information(10) By Relational(3) .122 1.000 1.130
Entertainment(1) By Relational(1) .369 .684 1.446
Entertainment(1) By Relational(2) .178 .906 1.195
Entertainment(1) By Relational(3) 17.258 .997 31278733.481
Entertainment(1) By Relational(4) .251 1.000 1.286
Entertainment(1) By Relational(5) .522 1.000 1.686
Entertainment(1) By Relational(6) 17.148 .999 28011175.717
Entertainment(2) By Relational(1) 17.304 .997 32743551.355
Entertainment(2) By Relational(2) -.324 .836 .723
Entertainment(2) By Relational(3) 17.839 .998 55876913.024
Entertainment(2) By Relational(4) .556 1.000 1.743
Entertainment(2) By Relational(5) .490 1.000 1.632
Entertainment(2) By Relational(7) 19.600 1.000 325365298.073
Entertainment(3) By Relational(1) -19.740 .998 .000
Entertainment(3) By Relational(2) -2.144 1.000 .117
Entertainment(3) By Relational(3) -.481 1.000 .618
Entertainment(3) By Relational(4) -18.535 1.000 .000
Entertainment(3) By Relational(5) -18.350 1.000 .000
Entertainment(4) By Relational(1) .297 1.000 1.346
Entertainment(4) By Relational(2) -1.131 1.000 .323
Entertainment(4) By Relational(3) .477 1.000 1.610
Entertainment(5) By Relational(1) 16.040 1.000 9252213.626
Entertainment(5) By Relational(2) -.182 1.000 .833
Entertainment(5) By Relational(3) 18.407 1.000 98653664.723
Relational(1) By Remunerative(1) .737 .081 2.090
Relational(1) By Remunerative(2) 18.692 .996 131202032.679
Relational(1) By Remunerative(3) .103 1.000 1.109
Relational(2) By Remunerative(1) 2.447 1.000 11.551
Relational(2) By Remunerative(2) 19.781 1.000 389645508.881
Relational(2) By Remunerative(3) 1.407 1.000 4.084
Relational(2) By Remunerative(4) -.122 1.000 .885
Relational(3) By Remunerative(1) 1.292 .215 3.640
Relational(3) By Remunerative(2) 18.547 .998 113429576.896
Relational(3) By Remunerative(3) .123 1.000 1.131
Relational(3) By Remunerative(4) -.952 1.000 .386
Relational(4) By Remunerative(1) 17.597 .997 43892980.443
Relational(4) By Remunerative(2) 19.037 .999 185196242.220
Relational(4) By Remunerative(3) .687 1.000 1.987
Relational(5) By Remunerative(1) 17.952 .998 62598449.432
Relational(5) By Remunerative(2) 18.636 .999 124024110.012
Relational(6) By Remunerative(1) 16.592 .999 16064730.394
Relational(6) By Remunerative(2) 18.897 1.000 161086463.445
Relational(7) By Remunerative(3) 1.361 1.000 3.900
Entertainment(1) by Remunerative(1) 0.26 0.84 1.30
Entertainment(1) by Remunerative (2) 0.00 1.00 1.00
Entertainment(1) by Remunerative (3) 0.45 1.00 1.57
279 | P a g e
Appendices
Entertainment(2) by Remunerative (1) 17.38 1.00 35491539.81
Entertainment(2) by Remunerative (2) 0.45 1.00 1.57
Entertainment(3) by Remunerative (1) 16.43 1.00 13667800.08
Entertainment(3) by Remunerative (2) -0.19 1.00 0.83
Entertainment(4) by Remunerative (1) 0.39 1.00 1.48
Entertainment(5) by Remunerative (1) -1.04 1.00 0.35
Entertainment(1) by Remunerative (1) 0.26 0.84 1.30

280 | P a g e
Appendices
Appendix E.5. Logistic regression results showing effect of content interactions on
DORMANT SMEB

SMEB Interaction b Sig Exp(B)


Dormancy Entertainment(1) By Information(1) .635 .139 1.888
Entertainment(1) By Information(2) .608 .061 1.837
Entertainment(1) By Information(3) 1.131 .002 3.098
Entertainment(1) By Information(4) .320 .426 1.377
Entertainment(1) By Information(5) 1.269 .011 3.558
Entertainment(1) By Information(6) 1.667 .025 5.298
Entertainment(1) By Information(7) .877 .315 2.405
Entertainment(1) By Information(8) 3.190 .052 24.291
Entertainment(1) By Information(9) -41.180 .999 .000
Entertainment(1) By Information(10) .453 1.000 1.573
Entertainment(1) By Information(11) 20.674 1.000 951453849.936
Entertainment(2) By Information(1) .240 .580 1.271
Entertainment(2) By Information(2) .672 .140 1.957
Entertainment(2) By Information(3) .245 .620 1.278
Entertainment(2) By Information(4) .555 .327 1.742
Entertainment(2) By Information(5) .634 .302 1.886
Entertainment(2) By Information(6) .692 .465 1.997
Entertainment(2) By Information(7) 2.046 .071 7.740
Entertainment(2) By Information(9) .572 1.000 1.771
Entertainment(3) By Information(1) .140 .878 1.150
Entertainment(3) By Information(2) .666 .539 1.946
Entertainment(3) By Information(3) -.493 .651 .611
Entertainment(3) By Information(4) .909 .403 2.482
Entertainment(3) By Information(5) .720 .589 2.054
Entertainment(3) By Information(6) -19.304 1.000 .000
Entertainment(3) By Information(7) 21.733 1.000 274578963.420
Entertainment(3) By Information(9) -41.856 .999 .000
Entertainment(4) By Information(1) -19.733 1.000 .000
Entertainment(4) By Information(2) -20.238 1.000 .000
Entertainment(4) By Information(3) -21.591 1.000 .000
Entertainment(4) By Information(4) .543 1.000 1.721
Entertainment(4) By Information(5) -20.161 1.000 .000
Entertainment(5) By Information(2) -21.800 .999 .000
Entertainment(5) By Information(6) -20.612 1.000 .000
Information(1) By Remunerative(1) -43.614 .999 .000
Information (1) By Remunerative(2) -43.610 .999 .000
Information (1) By Remunerative(3) -43.698 .999 .000
Information (2) By Remunerative(1) -43.030 .999 .000
Information (2) By Remunerative(2) -42.792 .999 .000
Information (2) By Remunerative(3) -42.637 .999 .000
Information (3) By Remunerative(1) -1.392 .307 .248
Information (3) By Remunerative(2) -.554 .703 .575
Information (4) By Remunerative(1) -1.075 .442 .341
Information (4) By Remunerative(2) -.589 .692 .555
Information (5) By Remunerative(1) -2.349 .123 .096
Information (5) By Remunerative(2) -2.195 .168 .111
Information (6) By Remunerative(1) -22.053 .999 .000

281 | P a g e
Appendices
Information (6) By Remunerative(2) -22.274 .999 .000
Information (7) By Remunerative(1) -2.536 .149 .079
Information (7) By Remunerative(2) -2.910 .137 .054
Information (8) By Remunerative(1) 18.159 1.000 76956381.111
Information (8) By Remunerative(2) 40.966 .999 618198200.000
Information (9) By Remunerative(1) -43.429 .999 .000
Information (9) By Remunerative(2) -.616 1.000 .540
Information (10) By Remunerative(1) .304 1.000 1.356
Information(1) By Relational(1) 0.53 0.10 1.70
Information (1) By Relational(2) 0.83 0.03 2.30
Information (1) By Relational(3) 0.13 0.84 1.13
Information (1) By Relational(4) 1.60 0.23 4.95
Information (1) By Relational(5) 21.05 1.00 1391153869.26
Information (2) By Relational(1) 0.52 0.14 1.68
Information (2) By Relational(2) 0.63 0.13 1.88
Information (2) By Relational(3) 0.78 0.25 2.19
Information (2) By Relational(4) 2.08 0.11 8.04
Information (2) By Relational(5) 21.81 1.00 2953092372.09
Information (3) By Relational(1) 0.57 0.15 1.77
Information (3) By Relational(2) 1.00 0.03 2.72
Information (3) By Relational(3) 0.17 0.82 1.18
Information (3) By Relational(4) 2.70 0.05 14.89
Information (3) By Relational(5) 21.73 1.00 2740145635.41
Information (3) By Relational(6) 21.21 1.00 1628348321.47
Information (4) By Relational(1) 0.64 0.15 1.89
Information (4) By Relational(2) 1.31 0.01 3.71
Information (4) By Relational(3) 0.93 0.23 2.53
Information(4) By Relational(4) 3.69 0.02 40.16
Information (4) By Relational(5) 21.19 1.00 1590886840.48
Information (5) By Relational(1) 0.26 0.68 1.30
Information (5) By Relational(2) 0.97 0.13 2.64
Information (5) By Relational(3) 1.50 0.11 4.49
Information (5) By Relational(4) 1.89 0.21 .64
Information (5) By Relational(5) 23.09 1.00 1069859393.09
Information (5) By Relational(7) 20.99 1.00 1306619258.13
Information (6) By Relational(1) 1.91 0.14 6.72
Information (6) By Relational(2) 2.96 0.05 19.32
Information (6) By Relational(3) -19.83 1.00 0.00
Information (6) By Relational(4) 4.18 0.04 65.48
Information (6) By Relational(5) 1.52 1.00 4.56
Information (6) By Relational(6) 22.11 1.00 4011916200.56
Information (7) By Relational(1) 1.61 0.23 5.02
Information (7) By Relational(2) 2.39 0.09 10.90
Information (7) By Relational(3) 1.98 0.17 7.25
Information (7) By Relational(4) 24.71 1.00 5369777464.42
Information (7) By Relational(5) 23.98 1.00 25876368695.5
Information (8) By Relational(1) -1.20 0.34 0.30
Information (8) By Relational(2) 0.26 0.84 1.29
Information (8) By Relational(3) -21.45 1.00 0.00
Information (8) By Relational(4) -19.53 1.00 0.00
Information (8) By Relational(6) 20.24 1.00 614350319.72
Information (9) By Relational(1) -21.43 1.00 0.00
282 | P a g e
Appendices
Information (9) By Relational(2) 21.15 1.00 1535331361.32
Information (9) By Relational(3) 20.41 1.00 729417160.02
Information (9) By Relational(4) -20.31 1.00 0.00
Information (10) By Relational(2) 0.12 1.00 1.13
Information (10) By Relational(3) -0.23 1.00 0.80
Information (1) By Relational(1) 0.53 0.10 1.70
Entertainment(1) By Remunerative(1) .947 .004 2.578
Entertainment(1) By Remunerative(2) 1.026 .122 2.789
Entertainment(1) By Remunerative(3) -40.766 .999 .000
Entertainment(2) By Remunerative(1) .380 .329 1.462
Entertainment(2) By Remunerative(2) .725 .370 2.064
Entertainment(3) By Remunerative(1) 1.661 .202 5.265
Entertainment(3) By Remunerative(2) -19.838 1.000 .000
Entertainment(4) By Remunerative(1) -21.401 .999 .000
Entertainment(5) By Remunerative(1) 1.219 1.000 3.385
Entertainment(1) By Relational(1) .003 .989 1.003
Entertainment(1) By Relational(2) .461 .099 1.586
Entertainment(1) By Relational(3) -.150 .720 .860
Entertainment(1) By Relational(4) .759 .221 2.137
Entertainment(1) By Relational(5) 1.956 .142 7.069
Entertainment(1) By Relational(6) 21.013 .999 133570045.892
Entertainment(2) By Relational(1) .422 .228 1.525
Entertainment(2) By Relational(2) .293 .440 1.340
Entertainment(2) By Relational(3) .069 .890 1.072
Entertainment(2) By Relational(4) 1.355 .106 3.877
Entertainment(2) By Relational(5) 2.530 .069 12.558
Entertainment(2) By Relational(7) 20.932 1.000 123223006.004
Entertainment(3) By Relational(1) -.111 .870 .895
Entertainment(3) By Relational(2) .834 .265 2.303
Entertainment(3) By Relational(3) .068 .946 1.071
Entertainment(3) By Relational(4) -.351 .823 .704
Entertainment(3) By Relational(5) 23.600 .999 176481626.517
Entertainment(4) By Relational(1) -22.922 .999 .000
Entertainment(4) By Relational(2) -.754 .652 .471
Entertainment(4) By Relational(3) -1.562 .271 .210
Entertainment(5) By Relational(1) -21.225 1.000 .000
Entertainment(5) By Relational(2) -21.176 1.000 .000
Entertainment(5) By Relational(3) -22.121 1.000 .000
Relational(1) By Remunerative(1) -0.42 0.28 0.66
Relational(1) By Remunerative (2) -0.80 0.25 0.45
Relational(2) By Remunerative (1) 0.29 0.50 1.33
Relational(2) By Remunerative (2) -0.39 0.59 0.68
Relational(2) By Remunerative (3) 20.64 1.00 921396436.90
Relational(2) By Remunerative (4) 20.31 1.00 658513069.42
Relational(3) By Remunerative (1) 1.22 0.02 3.39
Relational(3) By Remunerative (2) 0.16 0.87 1.17
Relational(3) By Remunerative(3) -20.32 1.00 0.00
Relational(4) By Remunerative(1) -1.33 0.16 0.26
Relational(4) By Remunerative(2) 20.39 1.00 715974819.87
Relational(5) By Remunerative(1) -0.20 0.85 0.82
Relational(6) By Remunerative(1) 0.00 1.00 1.00
Relational(7) By Remunerative(2) 20.60 1.00 883121832.62
283 | P a g e
Appendices
Relational(1) By Remunerative(1) -0.42 0.28 0.66
Relational(1) By Remunerative(2) -0.80 0.25 0.45
Relational(2) By Remunerative(1) 0.29 0.50 1.33
Relational(2) By Remunerative(2) -0.39 0.59 0.68
Relational(2) By Remunerative(3) 20.64 1.00 921396436.90
Relational(2) By Remunerative(4) 20.31 1.00 658513069.42

284 | P a g e
Appendices
Appendix E.6. Logistic regression results showing effect of content interactions on
DETACHING SMEB

SMEB Interaction b Sig Exp(B)


Detaching Entertainment(1) By Information(1) -0.19 0.67 1.18
behaviour Entertainment(1) By Information (2) 0.40 0.27 1.55
Entertainment(1) By Information (3) 0.36 0.34 0.62
Entertainment(1) By Information (4) -16.65 0.03 0.32
Entertainment(1) By Information (5) 0.24 0.58 0.68
Entertainment(1) By Information (6) 0.34 0.88 1.13
Entertainment(1) By Information (7) 0.27 0.40 2.87
Entertainment(1) By Information (8) 0.26 0.56 0.42
Entertainment(1) By Information (9) -0.03 1.00 0.00
Entertainment(1) By Information (10) -0.33 1.00 0.58
Entertainment(1) By Information (11) 17.45 1.00 0.00
Entertainment(2) By Information (1) -1.20 0.96 1.03
Entertainment(2) By Information (2) -1.12 0.75 1.20
Entertainment(2) By Information (3) 16.57 0.63 1.38
Entertainment(2) By Information (4) -0.31 0.91 0.92
Entertainment(2) By Information (5) -0.19 0.41 1.90
Entertainment(2) By Information (6) 0.04 0.86 1.20
Entertainment(2) By Information (7) -0.52 1.00 0.00
Entertainment(2) By Information (9) -0.87 1.00 0.00
Entertainment(3) By Information (1) -18.94 1.00 430874509.25
Entertainment(3) By Information (2) -0.41 1.00 1010470885.90
Entertainment(3) By Information (3) -0.41 1.00 636204591.52
Entertainment(3) By Information (4) -17.34 1.00 383526207.89
Entertainment(3) By Information (5) -17.65 1.00 465321488.78
Entertainment(3) By Information (6) -35.04 1.00 0.19
Entertainment(3) By Information (7) -17.55 1.00 24210891213.00
Entertainment(3) By Information (9) -18.10 1.00 0.15
Entertainment(4) By Information (1) -0.57 1.00 1159103964.70
Entertainment(4) By Information (2) -1.66 1.00 167187670.24
Entertainment(4) By Information (3) -0.58 1.00 437574272.77
Entertainment(4) By Information (4) -16.85 1.00 0.54
Entertainment(4) By Information (5) -16.97 1.00 0.53
Entertainment(5) By Information (2) 17.27 0.09 11.81
Entertainment(5) By Information(6) -0.23 1.00 0.00
Information(1) By Remunerative(1) 0.70 0.41 2.01
Information (1) By Remunerative (2) 0.23 0.87 1.26
Information (1) By Remunerative (3) -20.12 1.00 0.00
Information (2) By Remunerative (1) 0.86 0.31 2.36
Information (2) By Remunerative (2) 0.09 0.95 1.09
Information (2) By Remunerative (3) -19.91 1.00 0.00
Information (3) By Remunerative (1) 0.26 0.79 1.30
Information (3) By Remunerative (2) -0.98 0.51 0.37
Information (4) By Remunerative (1) 1.00 0.26 2.73
Information (4) By Remunerative (2) 0.15 0.92 1.16
Information (5) By Remunerative (1) 0.61 0.53 1.85
Information (5) By Remunerative (2) 0.77 0.61 2.16
Information (5) By Remunerative (3) -20.01 1.00 0.00

285 | P a g e
Appendices
Information (6) By Remunerative (1) 1.60 0.15 4.94
Information (6) By Remunerative (2) -0.44 0.82 0.64
Information (7) By Remunerative (1) 1.94 0.14 6.94
Information (7) By Remunerative (2) -0.21 0.91 0.81
Information (8) By Remunerative (1) -42.29 1.00 0.00
Information (8) By Remunerative (2) -42.80 1.00 0.00
Information (8) By Remunerative (4) -43.25 1.00 0.00
Information (9) By Remunerative (1) 21.56 1.00 2319200299.08
Information (9) By Remunerative (2) -1.29 1.00 0.28
Information (10) By Remunerative (1) -0.39 1.00 0.68
Information(1) By Relational(1) 0.20 0.62 1.22
Information (1) By Relational(2) 0.32 0.53 1.37
Information (1) By Relational(3) 19.51 1.00 297482290.16
Information (1) By Relational(4) -0.88 0.54 0.41
Information (1) By Relational(5) -1.24 1.00 0.29
Information (2) By Relational(1) 0.25 0.56 1.29
Information (2) By Relational(2) 0.45 0.39 1.57
Information (2) By Relational(3) 19.81 1.00 400946813.04
Information (2) By Relational(4) -0.19 0.89 0.83
Information (2) By Relational(5) 20.04 1.00 506784491.92
Information (3) By Relational(1) -0.87 0.10 0.42
Information (3) By Relational(2) -0.41 0.52 0.67
Information (3) By Relational(3) 19.86 1.00 421433758.15
Information (3) By Relational(4) -1.49 0.36 0.23
Information (3) By Relational(5) -0.67 1.00 0.51
Information (3) By Relational(6) -19.68 1.00 0.00
Information (4) By Relational(1) -0.13 0.81 0.88
Information (4) By Relational(2) -0.17 0.79 0.85
Information (4) By Relational(3) 20.02 1.00 492635869.22
Information (4) By Relational(4) -0.65 0.67 0.52
Information (4) By Relational(5) 19.96 1.00 468382612.75
Information (5) By Relational(1) -0.66 0.35 0.51
Information (5) By Relational(2) -1.00 0.21 0.37
Information (5) By Relational(3) -0.66 1.00 0.52
Information (5) By Relational(4) -21.19 1.00 0.00
Information (5) By Relational(5) 18.72 1.00 135125429.81
Information (5) By Relational(7) 23.02 1.00 9906681704.14
Information (6) By Relational(1) 0.67 0.60 1.95
Information (6) By Relational(2) 0.05 0.97 1.05
Information (6) By Relational(3) -0.53 1.00 0.59
Information (6) By Relational(4) -20.48 1.00 0.00
Information (6) By Relational(5) 40.85 1.00 55081085068.00
Information (6) By Relational(6) 0.84 0.65 2.31
Information (7) By Relational(1) -20.90 1.00 0.00
Information (7) By Relational(2) 0.51 0.70 1.66
Information (7) By Relational(3) -1.45 1.00 0.23
Information (7) By Relational(4) -21.89 1.00 0.00
Information (7) By Relational(5) -2.16 1.00 0.11
Information (8) By Relational(1) 1.44 0.28 4.23
Information (8) By Relational(2) 2.13 0.10 8.45
Information (8) By Relational(3) 43.06 1.00 50306368794.00
Information (8) By Relational(4) 21.65 1.00 2533386209.98
286 | P a g e
Appendices
Information (8) By Relational(6) -19.87 1.00 0.00
Information (9) By Relational(1) -18.88 1.00 0.00
Information (9) By Relational(2) 1.66 0.26 5.26
Information (9) By Relational(3) -0.32 1.00 0.72
Information (9) By Relational(4) -20.54 1.00 0.00
Information (10) By Relational(2) -1.12 1.00 0.33
Information (10) by Relational(3) 18.58 1.00 117323980.33
Entertainment(1) By Remunerative(1) -.585 0.19 .557
Entertainment(1) By Remunerative (2) -.292 0.69 .747
Entertainment(1) By Remunerative (3) -.650 1.00 .522
Entertainment(2) By Remunerative (1) .441 0.35 1.554
Entertainment(2) By Remunerative (2) .119 0.89 1.126
Entertainment(3) By Remunerative (1) -20.776 1.00 .000
Entertainment(3) By Remunerative (2) 20.541 1.00 833745640.352
Entertainment(4) By Remunerative (1) 1.824 0.27 6.198
Entertainment(5) By Remunerative (1) -22.843 1.00 .000
Entertainment(1) By Remunerative (1) -.585 0.19 .557
Entertainment(1) By Remunerative (2) -.292 0.69 .747
Entertainment(1) By Remunerative (3) -.650 1.00 .522
Entertainment(2) By Remunerative (1) .441 0.35 1.554
Entertainment(2) By Remunerative (2) .119 0.89 1.126
Entertainment(3) By Remunerative (1) -20.776 1.00 .000
Entertainment(3) By Remunerative (2) 20.541 1.00 833745640.352
Entertainment(4) By Remunerative (1) 1.824 0.27 6.198
Entertainment(5) By Remunerative (1) -22.843 1.00 .000
Relational(1) By Remunerative(1) -.337 0.48 .714
Relational(1) By Remunerative (2) -.064 0.93 .938
Relational(2) By Remunerative (1) -.198 0.70 .820
Relational(2) By Remunerative (2) -.703 0.37 .495
Relational(2) By Remunerative (3) -19.962 1.00 .000
Relational(2) By Remunerative (4) -20.267 1.00 .000
Relational(3) By Remunerative (1) -.975 0.16 .377
Relational(3) By Remunerative (2) -1.459 0.25 .232
Relational(3) By Remunerative (3) -20.368 1.00 .000
Relational(4) By Remunerative (1) .136 0.91 1.146
Relational(4) By Remunerative(2) -.311 0.83 .733
Relational(5) By Remunerative(1) -.885 0.50 .413
Relational(6) By Remunerative(1) -20.263 1.00 .000
Relational(7) By Remunerative(2) 22.450 1.00 5621063601.119
Relational(1) By Remunerative (1) -.337 0.48 .714
Relational(1) By Remunerative (2) -.064 0.93 .938
Relational(2) By Remunerative (1) -.198 0.70 .820
Relational(2) By Remunerative (2) -.703 0.37 .495
Relational(2) By Remunerative (3) -19.962 1.00 .000
Relational(2) By Remunerative (4) -20.267 1.00 .000
Relational(3) By Remunerative (1) -.975 0.16 .377
Relational(3) By Remunerative (2) -1.459 0.25 .232
Relational(3) By Remunerative (3) -20.368 1.00 .000
Relational(4) By Remunerative (1) .136 0.91 1.146
Relational(4) By Remunerative (2) -.311 0.83 .733
Relational(5) By Remunerative (1) -.885 0.50 .413
Relational(6) By Remunerative (1) -20.263 1.00 .000
287 | P a g e
Appendices
Relational(7) By Remunerative (2) 22.450 1.00 5621063601.119
Entertainment(1) By Relational(1) .092 0.76 1.096
Entertainment(1) By Relational(2) .158 0.65 1.171
Entertainment(1) By Relational(3) -.830 0.13 .436
Entertainment(1) By Relational(4) -1.497 0.10 .224
Entertainment(1) By Relational(5) -.856 0.50 .425
Entertainment(1) By Relational(6) -.357 0.75 .700
Entertainment(2) By Relational(1) -.441 0.34 .643
Entertainment(2) By Relational(2) .653 0.16 1.922
Entertainment(2) By Relational(3) -.263 0.67 .769
Entertainment(2) By Relational(4) -.192 0.84 .825
Entertainment(2) By Relational(5) -.663 0.61 .515
Entertainment(2) By Relational(7) 23.653 1.00 1872355431.805
Entertainment(3) By Relational(1) 1.065 0.19 2.902
Entertainment(3) By Relational(2) 1.737 0.04 5.679
Entertainment(3) By Relational(3) -.433 0.75 .648
Entertainment(3) By Relational(4) -20.333 1.00 .000
Entertainment(3) By Relational(5) -20.318 1.00 .000
Entertainment(4) By Relational(1) 2.430 0.20 11.361
Entertainment(4) By Relational(2) -18.941 1.00 .000
Entertainment(4) By Relational(3) 1.210 0.41 3.353
Entertainment(5) By Relational(1) -20.179 1.00 .000
Entertainment(5) By Relational(2) -18.446 1.00 .000
Entertainment(5) By Relational(3) 23.658 1.00 1881386105.436

288 | P a g e
References

References
Abdul-Ghani, E., Hyde, K. F., & Marshall, R. (2011). Emic and etic interpretations of
engagement with a consumer-to-consumer online auction site. Journal of
Business Research, 64(10), 1060-1066
Achterberg, W., Pot, A. M., Kerkstra, A., Ooms, M., Muller, M., & Ribbe, M. (2003).
The effect of depression on social engagement in newly admitted Dutch nursing
home residents. The Gerontologist, 43(2), 213-218
Aichner, T., & Jacob, F. (2015). Measuring the degree of corporate social media use.
International Journal of Market Research, 57(2), 257-275
Aksoy, L., van Riel, A., Kandampully, J., Wirtz, J., den Ambtman, A., Bloemer, J., . . .
Gurhan Canli, Z. (2013). Managing brands and customer engagement in online
brand communities. Journal of Service Management, 24(3), 223-244
Algesheimer, R., Dholakia, U. M., & Herrmann, A. (2005). The Social Influence of
Brand Community: Evidence from European Car Clubs. Journal of Marketing,
69(3), 19-34
Alonso, A. D., Bressan, A., O'Shea, M., & Krajsic, V. (2013). Website and Social
Media Usage: Developments of Wine Tourison, Hospitality and the Wine
Sector. Tourism Planning and Development, 25(3), 229-248
Altschwager, T., Conduit, J., & Goodman, S. (2013). Branded Marketing Events:
Facilitating Customer Brand Engagement. International Conference of the
Academy of Wine Business Research, Ontario, Canada.
Anderson, W. T., Challagalla, G. N., & McFarland, R. G. (1999). Anatomy of
exchange. Journal of Marketing Theory and Practice, 8-19
Andriof, J., & Waddock, S. (2002). Unfolding stakeholder engagement Unfolding
stakeholder thinking: Theory, responsibility and engagement (Vol. 17, pp. 17-
42): Greenleaf Publishing in association with GSE Research.
Appelbaum, A. (2001). The Constant Consumer. . Retrieved from
http://gmj.gallup.com/content/745/Constant-Customer.aspx.
Ashforth, B. E., & Humphrey, R. H. (1993). Emotional labor in service roles: The
influence of identity. Academy of management review, 18(1), 88-115
Ashley, C., Noble, S. M., Donthu, N., & Lemon, K. N. (2011). Why customers won't
relate: Obstacles to relationship marketing engagement. Journal of Business
Research, 64(7), 749-756
Baek, K., Holton, A., Harp, D., & Yaschur, C. (2011). The links that bind: Uncovering
novel motivations for linking on Facebook. Computers in Human Behavior,
27(6), 2243-2248
Barber, N., Dodd, T., & Ghiselli, R. (2008). Capturing the Younger Wine Consumer.
Journal of Wine Research, 19(2), 123-141
Baron, P., & Corbin, L. (2012). Student engagement: rhetoric and reality. Higher
Education Research & Development, 31(6), 759-772
Bauer, R. A., & Greyser, S. A. (1968). Advertising in America, the consumer view.
Boston: Division of Research, Graduate School of Business, Harvard University.
Berelson, B. (1952). Content analysis in communication research. New York, US: Free
Press

289 | P a g e
References
Bergami, M., & Bagozzi, R. P. (2000). Self‐categorization, affective commitment and
group self‐esteem as distinct aspects of social identity in the organization.
British Journal of Social Psychology, 39(4), 555-577
Bergen, M. (2014). Ad Age Survey: How advertisers are spending on Facebook, Twitter
and YouTube. Retrieved from http://adage.com/article/digital/ad-age-reader-
survey-twitter-facebook-youtube/293923/
Bhattacharya, C. B., & Sen, S. (2003). Consumer-company identification: a framework
for understanding consumers’ relationships with companies. Journal of
marketing, 67(2), 76-88
Bijmolt, T. H., Leeflang, P. S., Block, F., Eisenbeiss, M., Hardie, B. G., Lemmens, A.,
& Saffert, P. (2010). Analytics for customer engagement. Journal of Service
Research, 13(3), 341-356
Binning, K. R., Unzueta, M. M., Huo, Y. J., & Molina, L. E. (2009). The Interpretation
of Multiracial Status and Its Relation to Social Engagement and Psychological
Well‐Being. Journal of Social Issues, 65(1), 35-49
Black, H. G., & Kelley, S. W. (2009). A storytelling perspective on online customer
reviews reporting service failure and recovery. Journal of Travel & Tourism
Marketing, 26(2), 169-179
Blackshaw, P., & Nazzaro, M. (2006). Word of mouth in the age of the web-fortified
consumer. Consumer-generated media (CGM), 101
Bless, H. (2000). The interplay of affect and cognition: The mediating role of general
knowledge structures. In J. P. Forgas (Ed.), Feeling and thinking: the role of
affect in social cognition. Studies in emotion and social interaction, second
series. (pp. 421). New York, NY, US: Cambridge University Press.
Bolton, R., & Saxena-Iyer, S. (2009). Interactive services: a framework, synthesis and
research directions. Journal of Interactive Marketing, 23(1), 91-104
Bowden, J. L.-H. (2009). The process of customer engagement: a conceptual
framework. The Journal of Marketing Theory and Practice, 17(1), 63-74
Bowden, J. L. H., Gabbott, M., & Naumann, K. (2014). Service relationships and the
customer disengagement – engagement conundrum. Journal of Marketing
Management, 1-33. doi: 10.1080/0267257X.2014.983143
Boyd, D. M., & Ellison, N. B. (2008). Social Network Sites: Definition, History, and
Scholarship. Journal of Computer-Mediated Communication, 13(1), 210-230
Breslin, K. (2013). Presentation on Constellation Digital Marketing in 2012. San
Francisco, CA.
Brodie, R. J., Hollebeek, L., Jurić, B., & Ilić, A. (2011). Customer engagement
conceptual domain, fundamental propositions, and implications for research.
Journal of Service Research, 14(3), 252-271
Brodie, R. J., Ilic, A., Juric, B., & Hollebeek, L. (2013). Consumer engagement in a
virtual brand community: An exploratory analysis. Journal of Business
Research, 66(1), 105-114
Brooks, M. (1989). Instant rapport: Warner Books New York.
Bruwer, J., & Alant, K. (2009). The hedonic nature of wine tourism consumption: an
experiential view. International Journal of Wine Business Research, 21(3), 235-
257
Bryson, C., & Hand, L. (2007). The role of engagement in inspiring teaching and
learning. Innovations in education and teaching international, 44(4), 349-362

290 | P a g e
References
Bulearca, M., & Bulearca, S. (2010). Twitter: a Viable Marketing Tool for SMEs?
Global Business & Management Research, 2(4), 296-309
Calder, B. J., Malthouse, E. C., & Schaedel, U. (2009). An experimental study of the
relationship between online engagement and advertising effectiveness. Journal
of Interactive Marketing (Mergent, Inc.), 23(4), 321-331
Campanelli, M. (2007). Engagement is the Next Phase in Marketing Communications:
Experian Summit. Retrieved May, 10, 2012
Carletta, J. (1996). Assessing agreement on classification tasks: the kappa statistic.
Computational linguistics, 22(2), 249-254
Carney, T. F. (1972). Content Analysis: A Technique For Systematic Inference From
Communications Author: . University of Mantioba: University of Mantioba
Press.
Chan, K. W., & Li, S. Y. (2010). Understanding consumer-to-consumer interactions in
virtual communities: The salience of reciprocity. Journal of Business Research,
63(9), 1033-1040
Chen, G. M. (2011). Tweet this: A uses and gratifications perspective on how active
Twitter use gratifies a need to connect with others. Computers in Human
Behavior, 27(2), 755-762
Chen, Q., Clifford, S. J., & Wells, W. D. (2002). Attitude toward the site II: new
information. Journal of Advertising Research, 42(2), 33-46
Cheung, C. M., Chiu, P.-Y., & Lee, M. K. (2011). Online social networks: Why do
students use facebook? Computers in Human Behavior, 27(4), 1337-1343
Cho, C.-H. (1999). How advertising works on the WWW: Modified elaboration
likelihood model. Journal of Current Issues & Research in Advertising, 21(1),
34-50
Chu, S.-C. (2011). Viral advertising in social media: Participation in Facebook groups
and responses among college-aged users. Journal of Interactive Advertising,
12(1), 30-43
Chu, S.-C., & Kim, Y. (2011). Determinants of consumer engagement in electronic
word-of-mouth (eWOM) in social networking sites. International Journal of
Advertising, 30(1), 47-75
Chung, C., & Austria, K. (2010). Social Media Gratification and Attitude toward Social
Media Marketing Messages: A Study of the Effect of Social Media Marketing
Messages on Online Shopping Value. Proceedings of the Northeast Business &
Economics Association
Cole, M. (2006). Qualitative research: a challenging paradigm for infection control.
British Journal of Infection Control, 7(6), 25-29
Csikszentmihalyi, M. (1990). Flow: The psychology of optimal performance. NY:
Cambridge UniversityPress
Cvijikj, I. P., & Michahelles, F. (2011). A case study of the effects of moderator posts
within a facebook brand page Social Informatics (pp. 161-170): Springer.
Cvijikj, I. P., & Michahelles, F. (2013). Online engagement factors on Facebook brand
pages. Social Network Analysis and Mining, 3(4), 843-861
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media
richness and structural design. Management Science, 32(5), 554-571
Davis Mersey, R., Malthouse, E. C., & Calder, B. J. (2010). Engagement with online
media. Journal of Media Business Studies, 7(2), 39-56

291 | P a g e
References
Day, G. S. (1976). A two-dimensional concept of brand loyalty Mathematical Models in
Marketing (pp. 89-89): Springer.
Day, G. S., & Montgomery, D. B. (1999). Charting new directions for marketing. The
Journal of Marketing, 3-13
De Vries, L., Gensler, S., & Leeflang, P. S. (2012). Popularity of brand posts on brand
fan pages: an investigation of the effects of social media marketing. Journal of
Interactive Marketing, 26(2), 83-91
De Vries, N. J., & Carlson, J. (2014). Examining the drivers and brand performance
implications of customer engagement with brands in the social media
environment. Journal of Brand Management, 21(6), 495-515
Delgado-Ballester, E., Munuera-Aleman, J. L., & Yague-Guillen, M. J. (2003).
Development and validation of a brand trust scale. International Journal of
Market Research, 45(1), 35-54
Deshpandé, R., Farley, J. U., & Webster, F. E., Jr. (1993). Corporate Culture, Customer
Orientation, and Innovativeness in Japanese Firms: A Quadrad Analysis.
Journal of Marketing, 57(1), 23-37
Dholakia, U. M., Bagozzi, R. P., & Pearo, L. K. (2004). A social influence model of
consumer participation in network- and small-group-based virtual communities.
International Journal of Research in Marketing, 21(3), 241-263
Drummond, T. (2004). Vocabulary of Emotions.
Duck, S. E., & Perlman, D. E. (1985). Understanding personal relationships: An
interdisciplinary approach: Sage Publications, Inc.
Ducoffe, R. H. (1995). How consumers assess the value of advertising. Journal of
Current Issues & Research in Advertising, 17(1), 1-18
Ducoffe, R. H. (1996). Advertising value and advertising on the web. Journal of
Advertising Research, 36, 21-36
Dumas, J. E., Begle, A. M., French, B., & Pearl, A. (2010). Effects of monetary
incentives on engagement in the PACE parenting program. Journal of Clinical
Child & Adolescent Psychology, 39(3), 302-313
Eighmey, J., & McCord, L. (1998). Adding value in the information age: uses and
gratifications of sites on the world wide web. Journal of Business Research,
41(3), 187-194
Eisenbeiss, M., Blechschmidt, B., Backhaus, K., & Freund, P. A. (2012). “The (real)
world is not enough:” motivational drivers and user behavior in virtual worlds.
Journal of Interactive Marketing, 26(1), 4-20
Elpers, J. L., Wedel, M., & Pieters, R. G. (2003). Why do consumers stop viewing
television commercials? Two experiments on the influence of moment-to-
moment entertainment and information value. Journal of Marketing Research,
40(4), 437-453
Emerson, R. M. (1976). Social exchange theory. Annual review of sociology, 335-362
Engel, J. F., Kegerreis, R. J., & Blackwell, R. D. (1969). Word-of-mouth
communication by the innovator. The Journal of Marketing, 15-19
Eppler, M. J., & Mengis, J. (2004). The concept of information overload: A review of
literature from organization science, accounting, marketing, MIS, and related
disciplines. The information society, 20(5), 325-344

292 | P a g e
References
Facebook. (2014). Removal of PTAT metric. Retrieved from
https://developers.facebook.com/docs/apps/migrations/pages-api-changes-2014-
07-02
Facebook. (2015). Page Post Metrics. Retrieved from
https://www.facebook.com/help/336143376466063/
Fehrer, J., Woratschek, H., & Germelmann, C. C. (2013). Antecedents and
Consequences of Customer Engagement–A literature review. ANZMAC,
Auckland.
Fiedler, K. (2001). Toward an integrative account of affect and cognition phenomena
using the bias computer algorithm. Feeling and thinking: The role of affect in
social cognition, 223
Field, A. (2013). Discovering statistics using IBM SPSS statistics: Sage.
Foley, M. (2006). Measuring the Turn-On. AAAA/ARF Consumer Engagement
Conference, New York.
Forgas, J. P. (2001). Feeling and thinking: The role of affect in social cognition. Studies
in emotion and social interaction, second series. . New York, NY, US.:
Cambridge University Press.
Fortin, D. R., & Dholakia, R. R. (2005). Interactivity and vividness effects on social
presence and involvement with a web-based advertisement. Journal of Business
Research, 58(3), 387-396
Fricker, R. D., & Schonlau, M. (2002). Advantages and disadvantages of Internet
research surveys: Evidence from the literature. Field methods, 14(4), 347-367
Füller, J. (2006). Why consumers engage in virtual new product developments initiated
by producers. Advances in Consumer research, 33, 639
Füller, J., Bartl, M., Ernst, H., & Mühlbacher, H. (2006). Community based innovation:
How to integrate members of virtual communities into new product
development. Electronic Commerce Research, 6(1), 57-73
Gambetti, R. C., & Graffigna, G. (2010). The concept of engagement A systematic
analysis of the ongoing marketing debate. International Journal of Market
Research(52), 801-826
Gambetti, R. C., Graffigna, G., & Biraghi, S. (2012). The Grounded Theory approach to
consumer-brand engagement. International Journal of Market Research, 54(5),
659-687
Gardner, M. P., Mitchell, A. A., & Russo, J. E. (1985). Low involvement strategies for
processing advertisements. Journal of Advertising, 14(2), 4-56
Ghuneim, M. (2006, June 29). Terms of Engagement: Measuring the Active Consumer.
Retrieved from http://wiredset.com/blogs/markghuneim/2008/03/26/terms-of-
engagementmeasuring-the-active-consumer/
Gilly, M. C., Graham, J. L., Wolfinbarger, M. F., & Yale, L. J. (1998). A dyadic study
of interpersonal information search. Journal of the Academy of Marketing
Science, 26(2), 83-100
Giurgiu, L., & Barsan, G. (2008). The prosumer–core and consequence of the web 2.0
era. Journal of Social Informatics, 9, 53-59
Gjoka, M., Sirivianos, M., Markopoulou, A., & Yang, X. (2008). Poking facebook:
characterization of osn applications Proceedings of the first workshop on Online
social networks (pp. 31-36): ACM.

293 | P a g e
References
Golder, S. A., Wilkinson, D. M., & Huberman, B. A. (2007). Rhythms of social
interaction: Messaging within a massive online network Communities and
Technologies 2007 (pp. 41-66): Springer.
Goodstein, R. C. (1993). Category-based applications and extensions in advertising:
motivating more extensive ad processing. Journal of Consumer Research, 87-99
Gorry, G. A., & Westbrook, R. A. (2011). Can you hear me now? Learning from
customer stories. Business horizons, 54(6), 575-584
Greenwood, M. (2007). Stakeholder engagement: Beyond the myth of corporate
responsibility. Journal of Business Ethics, 74(4), 315-327
Greer, C. R., & Lei, D. (2012). Collaborative innovation with customers: a review of the
literature and suggestions for future research*. International Journal of
Management Reviews, 14(1), 63-84
Grellhesl, M., & Punyanunt-Carter, N. M. (2012). Using the uses and gratifications
theory to understand gratifications sought through text messaging practices of
male and female undergraduate students. Computers in Human Behavior, 28(6),
2175-2181
Grönroos, C. (2004). The relationship marketing process: communication, interaction,
dialogue, value. Journal of business & industrial marketing, 19(2), 99-113
Guest, L. (1944). A study of brand loyalty. Journal of Applied Psychology, 28(1), 16
Gummerus, J., Liljander, V., Pura, M., & Van Riel, A. (2004). Customer loyalty to
content-based web sites: the case of an online health-care service. Journal of
services Marketing, 18(3), 175-186
Gummerus, J., Liljander, V., Weman, E., & Pihlström, M. (2012). Customer
engagement in a Facebook brand community. Management Research Review,
35(9), 857-877
Gupta, P., & Harris, J. (2010). How e-WOM recommendations influence product
consideration and quality of choice: A motivation to process information
perspective. Journal of Business Research, 63(9), 1041-1049
Gustafsson, A., Johnson, M. D., & Roos, I. (2005). The effects of customer satisfaction,
relationship commitment dimensions, and triggers on customer retention.
Journal of marketing, 69(4), 210-218
Hair, J. F., Black, W. C., Babin, B. J., Anderson, R. E., & Tatham, R. L. (2006).
Multivariate data analysis (Vol. 6): Pearson Prentice Hall Upper Saddle River,
NJ.
Halkias, G., & Kokkinaki, F. (2013). Increasing advertising effectiveness through
incongruity-based tactics: The moderating role of consumer involvement.
Journal of Marketing Communications, 19(3), 182-197
Hallberg, U. E., & Schaufeli, W. B. (2006). " Same same" but different? can work
engagement be discriminated from job involvement and organizational
commitment? European Psychologist, 11(2), 119
Ham, C.-D., Lee, J., & Lee, H.-S. (2014). Understanding consumers' creating behaviour
in social media: an application of uses and gratifications and the theory of
reasoned action. International Journal of Internet Marketing and Advertising
8(4)
Hansson, L., Wrangmo, A., & Solberg Søilen, K. (2013). Optimal ways for companies
to use Facebook as a marketing channel. Journal of Information,
Communication and Ethics in Society, 11(2), 112-126

294 | P a g e
References
Hargittai, E. (2007). Whose space? Differences among users and non‐users of social
network sites. Journal of Computer‐Mediated Communication, 13(1), 276-297
Harrell, F. E. (2013). Regression modeling strategies: with applications to linear
models, logistic regression, and survival analysis: Springer Science & Business
Media.
Harris, J. (2006, 29 June). Consumer Engagement: What Does It Mean? Retrieved from
http://www.imediaconnection.com/content/9729.imc
Hastie, R. (1980). Memory for behavioral information that confirms or contradicts a
personality impression. Person Memory: The Cognittr~ Basis of Social
Perceptions, 155
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process
analysis: A regression-based approach: Guilford Press.
Hayes, A. F. (2015). Hacking PROCESS to Estimate a Simple Moderation Model with a
Three-Category Moderator.
Heath, R. (2007). How Do We Predict Advertising Attention and Engagement.
University of Bath School of Management Working Paper Series
Heckler, S. E., & Childers, T. L. (1992). The role of expectancy and relevancy in
memory for verbal and visual information: what is incongruency? Journal of
Consumer Research, 475-492
Hemetsberger, A. (2002). Fostering cooperation on the Internet: social exchange
processes in innovative virtual consumer communities. Advances in Consumer
Research, 29(1), 354-356
Hennig-Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004). Electronic
word‐of‐mouth via consumer‐opinion platforms: What motivates consumers to
articulate themselves on the Internet? Journal of Interactive Marketing, 18(1),
38-52
Hennig-Thurau, T., Malthouse, E. C., Friege, C., Gensler, S., Lobschat, L.,
Rangaswamy, A., & Skiera, B. (2010). The impact of new media on customer
relationships. Journal of Service Research, 13(3), 311-330
Herzog, H. (1944). What do we really know about daytime serial listeners. Radio
research, 1943, 3-33
Higgins, E. T. (2006). Value from hedonic experience <em>and</em> engagement.
Psychological Review, 113(3), 439-460
Hiltz, S. R., & Turoff, M. (1985). Structuring computer-mediated communication
systems to avoid information overload. Communications of the ACM, 28(7),
680-689
Hollebeek, L., D. (2011a). Demystifying customer brand engagement: Exploring the
loyalty nexus. Journal of Marketing Management, 27(7-8), 785-807
Hollebeek, L., D. (2011b). Exploring customer brand engagement: definition and
themes. Journal of Strategic Marketing, 19(7), 555-573
Hollebeek, L., D, & Chen, T. (2014). Exploring positively-versus negatively-valenced
brand engagement: a conceptual model. Journal of Product & Brand
Management, 23(1), 62-74
Hollebeek, L. D., & Brodie, R. J. (2009). Wine service marketing, value co-creation and
involvement: research issues. International Journal of Wine Business Research,
21(4), 339-353

295 | P a g e
References
Hollebeek, L. D., Glynn, M. S., & Brodie, R. J. (2014). Consumer brand engagement in
social media: Conceptualization, scale development and validation. Journal of
interactive marketing, 28(2), 149-165
Holsti, O. R. (1969). Content analysis for the social sciences and humanities: Addison-
Wesley.
Hosmer Jr, D. W., & Lemeshow, S. (2004). Applied logistic regression: John Wiley &
Sons.
Houston, M. J., Childers, T. L., & Heckler, S. E. (1987). Picture-word consistency and
the elaborative processing of advertisements. Journal of Marketing Research,
359-369
Hoyer, W. D., Chandy, R., Dorotic, M., Krafft, M., & Singh, S. S. (2010). Consumer
cocreation in new product development. Journal of Service Research, 13(3),
283-296
Hsiao, C.-C., & Chiou, J.-S. (2012). The impact of online community position on online
game continuance intention: Do game knowledge and community size matter?
Information & Management, 49(6), 292-300
Hsu, C.-L., & Lin, J. C.-C. (2008). Acceptance of blog usage: The roles of technology
acceptance, social influence and knowledge sharing motivation. Information &
Management, 45(1), 65-74
Hu, S., & Wolniak, G. C. (2010). Initial evidence on the influence of college student
engagement on early career earnings. Research in Higher Education, 51(8), 750-
766
Hughes, D. J., Rowe, M., Batey, M., & Lee, A. (2012). A tale of two sites: Twitter vs.
Facebook and the personality predictors of social media usage. Computers in
Human Behavior, 28(2), 561-569
Hum, N. J., Chamberlin, P. E., Hambright, B. L., Portwood, A. C., Schat, A. C., &
Bevan, J. L. (2011). A picture is worth a thousand words: A content analysis of
Facebook profile photographs. Computers in Human Behavior, 27(5), 1828-
1833
Huo, Y. J., Binning, K. R., & Molina, L. E. (2010). Testing an integrative model of
respect: Implications for social engagement and well-being. Personality and
Social Psychology Bulletin, 36(2), 200-212
Huotari, K., & Hamari, J. (2012). Defining gamification: a service marketing
perspective Proceeding of the 16th International Academic MindTrek
Conference (pp. 17-22): ACM.
Jaakkola, E., & Alexander, M. (2014). The role of customer engagement behavior in
value co-creation a service system perspective. Journal of Service Research,
17(3), 247-261
Jacoby, J., & Kyner, D. B. (1973). Brand loyalty vs. repeat purchasing behavior.
Journal of Marketing research, 1-9
Jahn, B., & Kunz, W. (2012). How to transform consumers into fans of your brand.
Journal of Service Management, 23(3), 344-361
Jennings, M. K., & Stoker, L. (2004). Social trust and civic engagement across time and
generations. Acta politica, 39(4), 342-379
Johnson, M. D., & Fornell, C. (1991). A framework for comparing customer satisfaction
across individuals and product categories. Journal of Economic Psychology,
12(2), 267-286

296 | P a g e
References
Joinson, A. N. (2008). Looking at, looking up or keeping up with people?: motives and
use of facebook. Proceeding of the twenty-sixth annual SIGCHI conference on
Human factors in computing systems, Florence, Italy.
Kahn, W. A. (1990). Psychological conditions of personal engagement and
disengagement at work. Academy of management journal, 33(4), 692-724
Kahu, E. R. (2013). Framing student engagement in higher education. Studies in Higher
Education, 38(5), 758-773
Kaplan, A. M., & Haenlein, M. (2010). Users of the world, unite! The challenges and
opportunities of Social Media. Business Horizons, 53(1), 59-68
Kärkkäinen, H., Jussila, J. J., & Leino, M. (2012). Learning from and with customers
with social media: A model for social customer learning. International Journal
of Management, Knowledge and Learning(1), 5-25
Karpen, I. O., Bove, L. L., & Lukas, B. A. (2011). Linking service-dominant logic and
strategic business practice: a conceptual model of a service-dominant
orientation. Journal of Service Research
Katz, E., & Foulkes, D. (1962). On the use of the mass media as 'escape': clarification of
a concept. The Public Opinion Quarterly 26(3), 277-388
Kaye, B. K., & Johnson, T. J. (2003). From here to obscurity?: Media substitution
theory and traditional media in an on‐line world. Journal of the American
Society for Information Science and Technology, 54(3), 260-273
Keener, M. S. (1999). Strengthening institutional engagement: Addressing faculty
issues to facilitate change. Journal of Higher Education Outreach and
Engagement, 4(1), 29-36
Keller, K. L. (2009). Building strong brands in a modern marketing communications
environment. Journal of Marketing Communications, 15(2/3), 139-155
King, N., & Horrocks, C. (2010). Interviews in Qualitative Research. London: SAGE
Publications Inc.
King, R. C., & Xia, W. (1997). Media appropriateness: Effects of experience on
communication media choice. Decision Sciences, 28(4), 877-910
Ko, H., Cho, C.-H., & Roberts, M. S. (2005). Internet Uses and Gratifications. Journal
of Advertising, 34(2), 57-70
Koetsier, J. (2013). Facebook: 15 million businesses, companies, and organizations now
have a Facebook page. San Francisco, CA: VentureBeat
Korgaonkar, P. K., & Wolin, L. D. (1999). A multivariate analysis of web usage.
Journal of Advertising Research, 39, 53-68
Krivak, T. (2008). Facebook 101: Ten things you need to know about Facebook.
Information Today, 25(3), 1
Ku, Y.-C., Chu, T.-H., & Tseng, C.-H. (2013). Gratifications for using CMC
technologies: A comparison among SNS, IM, and e-mail. Computers in Human
Behavior, 29(1), 226-234
Lampe, C., Ellison, N., & Steinfield, C. (2006). A Face (book) in the crowd: Social
searching vs. social browsing Proceedings of the 2006 20th anniversary
conference on Computer supported cooperative work (pp. 167-170): ACM.
Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for
categorical data. biometrics, 159-174

297 | P a g e
References
LaRose, R., Mastro, D., & Eastin, M. S. (2001). Understanding Internet usage a social-
cognitive approach to uses and gratifications. Social Science Computer Review,
19(4), 395-413
Lassar, W., Mittal, B., & Sharma, A. (1995). Measuring customer-based brand equity.
Journal of Consumer Marketing, 12(4), 11
Lau, G. T., & Ng, S. (2001). Individual and situational factors influencing negative
word‐of‐mouth behaviour. Canadian Journal of Administrative Sciences/Revue
Canadienne des Sciences de l'Administration, 18(3), 163-178
Lee, D., Hosanagar, K., & Nair, H. S. (2013). The effect of advertising content on
consumer engagement: evidence from facebook. Available at SSRN
Lee, Y. H. (2000). Manipulating ad message involvement through information
expectancy: effects on attitude evaluation and confidence. Journal of
Advertising, 29(2), 29-43
Lee, Y. H., & Mason, C. (1999). Responses to information incongruency in advertising:
The role of expectancy, relevancy, and humor. Journal of Consumer Research,
26(2), 156-169
Leigon, B. (2011). Grape/Wine Marketing with new media and return of the boomer.
Practical Winery & Vineyard Journal
Leskovec, J. (2011). Social media analytics: tracking, modeling and predicting the flow
of information through networks Proceedings of the 20th international
conference companion on World wide web (pp. 277-278): ACM.
Leung, L. (2009). User-generated content on the internet: an examination of
gratifications, civic engagement and psychological empowerment. New Media &
Society, 11(8), 1327-1347
Leung, L. (2013). Generational differences in content generation in social media: The
roles of the gratifications sought and of narcissism. Computers in Human
Behavior, 29(3), 997-1006
Leung, L., & Wei, R. (2000). More than just talk on the move: Uses and gratifications
of the cellular phone. Journalism & Mass Communication Quarterly, 77(2),
308-320
Levy, M. (2009). WEB 2.0 implications on knowledge management. Journal of
knowledge management, 13(1), 120-134
Lillevalja, J. (2010). The State of Inbound Marketing: Hubspot.
Lin, C. A. (1999). Online service adoption likelihood. Journal of Advertising Research,
39, 79-89
Lin, K.-Y., & Lu, H.-P. (2011). Why people use social networking sites: An empirical
study integrating network externalities and motivation theory. Computers in
Human Behavior, 27(3), 1152-1161
Liu, C., Rau, P.-L. P., & Gao, F. (2010). Mobile information search for location-based
information. Computers in industry, 61(4), 364-371
Lohtia, R., Donthu, N., & Hershberger, E. K. (2003). The impact of content and design
elements on banner advertising click-through rates. Journal of Advertising
Research, 43(4), 410-418
Luo, X. (2002). Uses and gratifications theory and e-consumer behaviors: a structural
equation modeling study. Journal of Interactive Advertising, 2(2), 34-41
Lusch, R. F. (2007). Marketing's evolving identity: defining our future. Journal of
Public Policy & Marketing, 26(2), 261-268

298 | P a g e
References
Lusch, R. F., Vargo, S. L., & Tanniru, M. (2010). Service, value networks and learning.
Journal of the academy of marketing science, 38(1), 19-31
Maddox, K. (1998). E-commerce becoming reality. Advertising Age, 69(43), S1-S2
Malthouse, E. C., Haenlein, M., Skiera, B., Wege, E., & Zhang, M. (2013). Managing
customer relationships in the social media era: introducing the social CRM
house. Journal of Interactive Marketing, 27(4), 270-280
Mangold, W. G., & Faulds, D. J. (2009). Social media: The new hybrid element of the
promotion mix. Business Horizons, 52(4), 357-365
Manovich, L. (2011). Trending: the promises and the challenges of big social data.
Debates in the digital humanities, 460-475
Mastermind. (2015). State of the Australian and New Zealand Wine Industry on
Facebook.
Mayfield, A. (2008). What is Social Media?” e-book from iCrossing: August.
McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D., & Barton, D. (2012). Big
data. The management revolution. Harvard Bus Rev, 90(10), 61-67
McEwen, W. (2004). Why Satisfaction Isn't Satistying Gallup Management Jounral
Online, November(11)
McGee, M. (2013). EdgeRank Is Dead: Facebook's News Feed Algorithm now has
close to 100K Weight Factors.
McGinn, T., Guyatt, G., Cook, R., Korenstein, D., & Meade, M. O. (2008). Measuring
agreement beyond chance. JAMA’s Users’ Guides to the Medical Literature: A
Manual for Evidence-Based Clinical Practice. 2nd ed. New York City: McGraw-
Hill Medical, 481-489
McGuire, W. J. (1974). Psychological Motives and Communication Gratification. In J.
Blumler & E. Katz (Eds.), The Uses of Mass Communications. Beverly Hills,
CA: Sage Publications.
McQuail, D. (1983). Mass Communication Theory. London and Beverly Hills,
California: Sage
McWilliam, G. (2012). Building stronger brands through online communities. Sloan
management review, 41(3)
Meeus, M. T., Oerlemans, L. A., & Hage, J. (2001). Patterns of interactive learning in a
high-tech region. Organization Studies, 22(1), 145-172
Mehta, A. (1999). Using Self-Concept to Assess Advertising Effectiveness. Journal of
Advertising Research, 39(1), 81-89
Menguc, B., Auh, S., & Yannopoulos, P. (2014). Customer and supplier involvement in
design: The moderating role of incremental and radical innovation capability.
Journal of Product Innovation Management, 31(2), 313-328
Meyers-Levy, J., & Tybout, A. M. (1989). Schema congruity as a basis for product
evaluation. Journal of Consumer Research, 39-54
Mollen, A., & Wilson, H. (2010). Engagement, telepresence and interactivity in online
consumer experience: Reconciling scholastic and managerial perspectives.
Journal of business research, 63(9), 919-925
Mondak, J. J., Hibbing, M. V., Canache, D., Seligson, M. A., & Anderson, M. R.
(2010). Personality and civic engagement: An integrative framework for the
study of trait effects on political behavior. American Political Science Review,
104(01), 85-110

299 | P a g e
References
Moore, B. (2012, 13 January). Facebook marketing case study: Pacific Rim Riesling
Wine. Retrieved from http://www.furlongpr.com/facebook-marketing-case-
study-pacific-rim-riesling-wine/#more-8914
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship
marketing. the journal of marketing, 20-38
MSI. (2014). 2014-2016 Research Priorities Retrieved from
http://www.msi.org/research/2014-2016-research-priorities/
MSI, M. S. I. (2010). 2010-2012 Research Priorities Retrieved from
http://www.msi.org/research/index.cfm?id¼271
Muehling, D. D., & Laczniak, R. N. (1988). Advertising's immediate and delayed
influence on brand attitudes: Considerations across message-involvement levels.
Journal of Advertising, 17(4), 23-34
Muntinga, D. G., Moorman, M., & Smit, E. G. (2011). Introducing COBRAs.
International Journal of Advertising, 30(1), 13-46
Nambisan, S., & Baron, R. A. (2007). Interactions in virtual customer environments:
Implications for product support and customer relationship management.
Journal of Interactive Marketing, 21(2), 42-62
Nelson-Field, K., & Taylor, J. (2012). Facebook fans: A fan for life? Admap: Warp.
Neuendorf, K. A. (2002). The Content Analysis Guidebook. Thousand Oaks, CA: Sage
Publications.
Nonnecke, B., & Preece, J. (1999). Shedding light on lurkers in online communities.
Ethnographic Studies in Real and Virtual Environments: Inhabited Information
Spaces and Connected Communities, Edinburgh, 123-128
Nuttney, A. (2010). The Social Networking Market Opportunity. Insights, Birmingham
O'Keefe, G. J., & Sulanowski, B. K. (1995). More than just talk: Uses, gratifications,
and the telephone. Journalism & Mass Communication Quarterly, 72(4), 922-
933
Owyang, J. (2007, 29 June). Defining Engagement. Retrieved from http://www.web-
strategist.com/blog/2007/02/01/defining-engagement/
Pagani, M., Hofacker, C. F., & Goldsmith, R. E. (2011). The influence of personality on
active and passive use of social networking sites. Psychology & Marketing,
28(5), 441-456
Papacharissi, Z., & Rubin, A. M. (2000). Predictors of Internet use. Journal of
Broadcasting & Electronic Media, 44(2), 175-196
Park, N., Jin, B., & Jin, S.-A. A. (2011). Effects of self-disclosure on relational intimacy
in Facebook. Computers in Human Behavior, 27(5), 1974-1983
Park, N., Kee, K. F., & Valenzuela, S. (2009). Being immersed in social networking
environment: Facebook groups, uses and gratifications, and social outcomes.
CyberPsychology & Behavior, 12(6), 729-733
Patterson, P., Yu, T., & De Ruyter, K. (2006). Understanding customer engagement in
services Advancing theory, maintaining relevance, proceedings of ANZMAC
2006 conference, Brisbane (pp. 4-6).
Pelling, E. L., & White, K. M. (2009). The Theory of Planned Behavior Applied to
Young People's Use of Social Networking Web Sites. CyberPsychology &
Behavior, 12(6), 755-759
Peppers, D., & Rogers, M. (2005). Return on customer: creating maximum value from
your scarcest resource (Vol. VIII). New York: Currency Doubleday.

300 | P a g e
References
Peters, K., Chen, Y., Kaplan, A. M., Ognibeni, B., & Pauwels, K. (2013). Social Media
Metrics—A Framework and Guidelines for Managing Social Media. Journal of
Interactive Marketing, 27(4), 281-298
Peterson, E. T. (2007, June 29). How To Measure Visitor Engagement, Redux.
Retrieved from http://blog.webanalyticsdemystified.com/weblog/2007/10/how-
to-measurevisitorengagement-redux.html
Phillips, B. J., & McQuarrie, E. F. (2010). Narrative and persuasion in fashion
advertising. Journal of Consumer Research, 37(3), 368-392
Plé, L., & Cáceres, R. C. (2010). Not always co-creation: introducing interactional co-
destruction of value in service-dominant logic. Journal of Services Marketing,
24(6), 430-437
Preece, J., Nonnecke, B., & Andrews, D. (2004). The top five reasons for lurking:
improving community experiences for everyone. Computers in human behavior,
20(2), 201-223
QSR International. (2015). Run a Coding Comparison query. Retrieved from
http://help-
nv10.qsrinternational.com/desktop/procedures/run_a_coding_comparison_query
.htm
Quan-Haase, A., & Young, A. L. (2010). Uses and gratifications of social media: A
comparison of Facebook and instant messaging. Bulletin of Science, Technology
& Society, 30(5), 350-361
Raacke, J., & Bonds-Raacke, J. (2008). MySpace and Facebook: Applying the uses and
gratifications theory to exploring friend-networking sites. CyberPsychology &
Behavior, 11(2), 169-174
Rafaeli, S. (1984). The electronic bulletin board: A computer-driven mass medium.
Social Science Computer Review, 2(3), 123-136
Raney, A., Janicke, S., & Tamborini, R. (2013). How we enjoy and why we seek out
morally complex characters in media entertainment. Media and the moral mind,
152-169
Raney, A. A., Arpan, L. M., Pashupati, K., & Brill, D. A. (2003). At the movies, on the
web: an investigation of the effects of entertaining and interactive web content
on site and brand evaluations. Journal of Interactive Marketing, 17(4), 38-53
Rayburn, J. D. (1996). Uses and gratifications. An integrated approach to
communication theory and research, 97-119
Reed, A. (2002). Social identity as a useful perspective for self‐concept–based
consumer research. Psychology & Marketing, 19(3), 235-266
Ribiere, V. M., & Tuggle, F. D. (2010). Fostering innovation with KM 2.0. Vine, 40(1),
90-101
Riegner, C. (2007). Word of mouth on the web: the impact of web 2.0 on consumer
purchase decisions. Journal of Advertising Research, 47(4), 436-447
Riffe, D., Lacy, S., & Fico, F. (2014). Analyzing media messages: Using quantitative
content analysis in research: Routledge.
Rosenkrans, G. (2009). The creativeness and effectiveness of online interactive rich
media advertising. Journal of Interactive Advertising, 9(2), 18-31
Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust1. Journal of
personality, 35(4), 651-665

301 | P a g e
References
Rubin, A. M. (1983). Television uses and gratifications: The interactions of viewing
patterns and motivations. Journal of Broadcasting & Electronic Media, 27(1),
37-51
Rubin, A. M. (2002). The uses-and-gratifications perspective of media effects. In B.
Jennings (Ed.), Media effects: Advances in theory and research (2 ed., pp. 525-
548). Mahwah, NJ, US: Lawrence Erlbaum Associates Publishers.
Ruggiero, T. E. (2000). Uses and gratifications theory in the 21st century. Mass
communication & society, 3(1), 3-37
Rutz, O. J., & Bucklin, R. E. (2011). From generic to branded: A model of spillover in
paid search advertising. Journal of Marketing Research, 48(1), 87-102
Ryan, D. (2014). Understanding digital marketing: marketing strategies for engaging
the digital generation: Kogan Page Publishers.
Ryan, T., & Xenos, S. (2011). Who uses Facebook? An investigation into the
relationship between the Big Five, shyness, narcissism, loneliness, and Facebook
usage. Computers in Human Behavior, 27(5), 1658-1664
Saks, A. M. (2006). Antecedents and consequences of employee engagement. Journal
of managerial psychology, 21(7), 600-619
Sashi, C. (2012). Customer engagement, buyer-seller relationships, and social media.
Management Decision, 50(2), 253-272
Sawhney, M., Verona, G., & Prandelli, E. (2005). Collaborating to create: The Internet
as a platform for customer engagement in product innovation. Journal of
Interactive Marketing, 19(4), 4-17
Scarpi, D. (2010). Does Size Matter? An Examination of Small and Large Web-Based
Brand Communities. Journal of Interactive Marketing, 24(1), 14-21. doi:
10.1016/j.intmar.2009.10.002
Schaufeli, W. B., Salanova, M., González-Romá, V., & Bakker, A. B. (2002). The
measurement of engagement and burnout: A two sample confirmatory factor
analytic approach. Journal of Happiness studies, 3(1), 71-92
Schmidt, C. T., & Scholl, R. W. (2004). Motivation: affirming behavior. Retrieved from
http://www.uri.edu/research/lrc/scholl/webnotes/Motivation_Affirming.htm
Schneckenberg, D. (2009). Web 2.0 and the empowerment of the knowledge worker.
Journal of knowledge management, 13(6), 509-520
Scholer, A. A., & Higgins, E. T. (2009). Exploring the complexities of value creation:
The role of engagement strength. Journal of Consumer Psychology (Elsevier
Science), 19(2), 137-143. doi: 10.1016/j.jcps.2009.02.007
Severin, W. J., & Tankard, J. W. (1997). Uses of mass media. Communication theories:
Origins, methods, and uses in the mass media
Shamdasani, P. N., Stanaland, A. J., & Tan, J. (2001). Location, location, location:
Insights for advertising placement on the web. Journal of Advertising Research
Shang, R.-A., Chen, Y.-C., & Liao, H.-J. (2006). The value of participation in virtual
consumer communities on brand loyalty. Internet research, 16(4), 398-418
Shaw, R. S., Chen, C. C., Harris, A. L., & Huang, H.-J. (2009). The impact of
information richness on information security awareness training effectiveness.
Computers & Education, 52(1), 92-100
Sheridan, T. B., & Ferrell, W. R. (1974). Man-machine systems; Information, control,
and decision models of human performance: The MIT Press.

302 | P a g e
References
Shevlin, R. (2007). ‘The Value of Customer Engagement Accessed April (Vol. 20, pp.
2010).
Shirky, C. (2009). How social media can make history Speech at the TED Conference
(Vol. 19).
Simmel, G. (1950). The sociology of georg simmel (Vol. 92892): Simon and Schuster.
Sinclair, L. (2014). Spending on digital marketing to outstrip traditional channels: study.
Sydney: The Australian.
Sirgy, M. J. (1982). Self-Concept in Consumer Behavior: A Critical Review. Journal of
Consumer Research, 9(3), 287-300
Skinner, E. A., & Belmont, M. J. (1993). Motivation in the classroom: Reciprocal
effects of teacher behavior and student engagement across the school year.
Journal of educational psychology, 85(4), 571
Smith, M., & Treadaway, C. (2010). Facebook Marketing: An hour a day: John Wiley
& Sons.
Smith, M. A., & Kollock, P. (1999). Communities in cyberspace: Psychology Press.
Smith, S., & Wallace, O. (2010). ‘What is Customer Engagement? [2010]. Retrieved
from http://www.wisegeek.com/what-is-customer-engagement.html
Smock, A. D., Ellison, N. B., Lampe, C., & Wohn, D. Y. (2011). Facebook as a toolkit:
A uses and gratification approach to unbundling feature use. Computers in
Human Behavior, 27(6), 2322-2329
So, K. K. F., King, C., & Sparks, B. (2014). Customer Engagement With Tourism
Brands Scale Development and Validation. Journal of Hospitality & Tourism
Research, 38(3), 304-329
Social Bakers. (2015). Free Social Media Statistics Retrieved from
http://www.socialbakers.com/statistics/
Speier, C., Valacich, J. S., & Vessey, I. (1999). The influence of task interruption on
individual decision making: An information overload perspective. Decision
Sciences, 30(2), 337-360
Sprott, D., Czellar, S., & Spangenberg, E. (2009). The importance of a general measure
of brand engagement on market behavior: Development and validation of a
scale. Journal of Marketing Research, 46(1), 92-104
Stafford, T. F., & Stafford, M. R. (2001). Identifying Motivations for the Use of
Commercial Web Sites. Information Resources Management Journal, 14(1), 22
Stafford, T. F., Stafford, M. R., & Schkade, L. L. (2004). Determining uses and
gratifications for the Internet. Decision Sciences, 35(2), 259-288
Steeves, N. (2013). Best practices: posting and analyzing effective facebook content.
Retrieved from http://www.nimble.com/blog/posting-and-analyzing-on-
facebook/
Stelzner, M. (2014). 2014 Social Media Marketing Industry Report. S. M. Examiner.
Stern, B., & Zaichowsky, J. L. (1991). The impact of'entertaining'advertising on
consumer responses. Australian Marketing Researcher, 14(1), 68-80
Stewart, K. (1998). The customer exit process-a review and research agenda. Journal of
Marketing Management, 14(4), 235-250
Stieglitz, S., & Dang-Xuan, L. (2013). Social media and political communication: a
social media analytics framework. Social Network Analysis and Mining, 3(4),
1277-1291

303 | P a g e
References
Stieglitz, S., Dang-Xuan, L., Bruns, A., & Neuberger, C. (2014). Social Media
Analytics. Business & Information Systems Engineering, 6(2), 89-96
Stone, P., Dunphy, D. C., Smith, M. S., & Ogilvie, D. (1968). The general inquirer: A
computer approach to content analysis. Journal of Regional Science, 8(1), 113-
116
Sujan, M. (1985). Consumer knowledge: Effects on evaluation strategies mediating
consumer judgments. Journal of Consumer Research, 31-46
Swanson, D. L. (1987). Gratification seeking, media exposure, and audience
interpretations: Some directions for research. Journal of Broadcasting &
Electronic Media, 31(3), 237-254
Szmigin, I., Canning, L., & Reppel, A. E. (2005). Online community: enhancing the
relationship marketing concept through customer bonding. International Journal
of Service Industry Management, 16(5), 480-496
Taheri, B., Jafari, A., & O'Gorman, K. (2014). Keeping your audience: Presenting a
visitor engagement scale. Tourism Management, 42, 321-329
Taylor, D. G., Lewin, J. E., & Strutton, D. (2011). Friends, fans, and followers: do ads
work on social networks? Journal of Advertising Research, 51(1), 258-275
Terry, D. J., Hogg, M. A., & White, K. M. (1999). The theory of planned behaviour:
self‐identity, social identity and group norms. British Journal of Social
Psychology, 38(3), 225-244
Thelwall, M., & Stuart, D. (2009). social network sites. Social Computing and Virtual
Communities, 263
Thomson, M., MacInnis, D. J., & Park, C. W. (2005). The Ties That Bind: Measuring
the Strength of Consumers' Emotional Attachments to Brands. Journal of
Consumer Psychology (Lawrence Erlbaum Associates), 15(1), 77-91. doi:
10.1207/s15327663jcp1501_10
Tinsley, H. E., & Weiss, D. J. (1975). Interrater reliability and agreement of subjective
judgments. Journal of Counseling Psychology, 22(4), 358
Tsai, H.-T., Huang, H.-C., & Chiu, Y.-L. (2012). Brand community participation in
Taiwan: Examining the roles of individual-, group-, and relationship-level
antecedents. Journal of Business Research, 65(5), 676-684
Tuten, T. L. (2008). Advertising 2.0: social media marketing in a web 2.0 world:
Greenwood Publishing Group.
Ulusu, Y. (2010). Determinant factors of time spent on Facebook: brand community:
engagement and usage types. Journal of Yasae University. Vol. 18 (5), 2949-
2957
van Doorn, J., Lemon, K. N., Mittal, V., Nass, S., Pick, D., Pirner, P., & Verhoef, P. C.
(2010). Customer engagement behavior: theoretical foundations and research
directions. Journal of Service Research, 13(3), 253-266
Verhoef, P. C., Reinartz, W. J., & Krafft, M. (2010). Customer engagement as a new
perspective in customer management. Journal of Service Research, 13(3), 247-
252
Verleye, K., Gemmel, P., & Rangarajan, D. (2013). Managing engagement behaviors in
a network of customers and stakeholders evidence from the nursing home sector.
Journal of Service Research, 1094670513494015
Viera, A. J., & Garrett, J. M. (2005). Understanding interobserver agreement: the kappa
statistic. Fam Med, 37(5), 360-363

304 | P a g e
References
Vinography. (2012, 14 September). Social Media and the Wine Industry: A New Era.
Retrieved from
http://www.vinography.com/archives/2012/02/social_media_and_the_wine_ind
u.html
Vivek, S. D. (2009). A scale of consumer engagement. The University of Alabama.
Vivek, S. D., Beatty, S. E., Dalela, V., & Morgan, R. M. (2014). A Generalized
Multidimensional Scale for Measuring Customer Engagement. Journal of
Marketing Theory and Practice, 22(4), 401-420
Vivek, S. D., Beatty, S. E., & Morgan, R. M. (2012). Customer engagement: Exploring
customer relationships beyond purchase. The Journal of Marketing Theory and
Practice, 20(2), 122-146
Vollmer, C., & Precourt, G. (2008). Always on: Advertising, marketing, and media in an
era of consumer control: McGraw Hill Professional.
Wallace, E., Buil, I., De Chernatony, L., & Hogan, M. (2014). Who Likes You and
Why? A typology of Facebook Fans. Journal Of Advertising Research
Waters, R. D., Burnett, E., Lamm, A., & Lucas, J. (2009). Engaging stakeholders
through social networking: How nonprofit organizations are using Facebook.
Public Relations Review, 35(2), 102-106
Weaver, K., & Olson, J. K. (2006). Understanding paradigms used for nursing research.
Journal of advanced nursing, 53(4), 459-469
Weber, R. P. (1990). Basic content analysis: Sage.
Whiting, A., & Williams, D. (2013). Why people use social media: A uses and
gratifications approach. Qualitative Market Research: An International Journal,
16(4), 362-369
Wilson, D., & Quinton, S. (2012). Let's talk about wine: Does Twitter have value? .
International Journal of Wine Business Research, 24(4), 271-286
Wimmer, R. D., & Dominick, J. R. (1994). Mass media research: An introduction
Belmont, CA: Wadsworth.
Wine Australia. (2015). Grape Varieties. Retrieved from
http://www.wineaustralia.net.au/en/grape-varieties.aspx
Winebiz. (2014). Wine Industry Statistics. Retrieved from
http://www.winebiz.com.au/statistics/
Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of consumer
research, 341-352
Zeithaml, V. A. (1988). Consumer perceptions of price, quality, and value: a means-end
model and synthesis of evidence. The Journal of marketing, 2-22
Zuckerman, M. (1979). Sensation seeking. Corsini Encyclopedia of Psychology

305 | P a g e

You might also like