02 Whole
02 Whole
02 Whole
A thesis submitted in fulfilment
DOCTOR OF PHILOSOPHY
By
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
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
This dissertation aims to advance existing knowledge on social media content types by
nature of engagement behaviour in this context. This construct includes six discrete
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
x|Page
The results show that the four social media content types have distinct and independent
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.
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
remuneration and relational interaction. The results also showed significant moderating
effects of media richness, community size, and congruity of the social media content,
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
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
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
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: ____________________________________________
xii | P a g e
Publications
The following publications are based upon the research presented in this thesis, and may
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
Firstly, I would like to thank my supervisors Professor. John Fahy, Dr. Jodie Conduit,
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,
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
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
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,
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-
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:
manifestations that have a brand or firm focus, beyond purchase” (van Doorn et al.,
2010 p.254).
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
the brand (Hollebeek and Chen, 2014). This thesis categorises two negatively-valenced
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
Consuming SMEB: Users passively consume content without any form of active
Dormant SMEB: A temporary state of inactive, passive engagement by users who may
Detaching SMEB: Users take action to remove content of the brand appearing in their
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.
brand pages to enable high-level monitoring of the activities on the Facebook page.
allows researchers to quickly and easily capture content including web pages, online
approach to quantitative content analysis (QCA), suitable for this study due to its focus
commonly defined as a research technique for the objective, systematic and quantitative
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
engagement behaviour (SMEB) and distinguishes this from neutral and negatively-
valenced SMEB. It examines the role of social media content in the form 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
economic incentive or reward and a need for social and relational interaction.
contributes to the knowledge and understanding of the influence of social media content
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
relational content.
1|Page
Chapter 1: Introduction
The specific objectives of this research are:
2. To offer new insights into the characteristics, levels and valence of social media
engagement behaviour.
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
Development.
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
discussed in Chapter 2, Section 2.3, leading to the identification of the four categories of
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
social media users, consuming, dormancy and detachment are more passive and/or
individualised forms of engagement. The SMEB construct also recognises that SMEB
The relationship between social media content and SMEB is supported by the UGT
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
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
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
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
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
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.
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
4|Page
Chapter 1: Introduction
content to active participants who create vast quantities of user generated content
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
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,
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
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).
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
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
between the focal engagement object and/or other actors and result from motivational
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
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
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
engagement; and how social media marketing activities create customer engagement
(MSI, 2014).
touch-point in the service experience. It therefore does not reflect customer brand
Consistent with calls from previous researchers (Brodie et al., 2011; Vivek et al., 2012),
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
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
Skiera, Wege, & Zhang, 2013). Through understanding this form of data, managers can
measure and hence know radically more about their businesses and directly translate
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
demonstrating how this data can be effectively accessed and analysed in order to
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
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
2015). However, many practitioners have identified a lack of awareness and knowledge
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
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
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
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
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
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.
between social media content and SMEB in unrelated contexts. Further examination of
the relationships among the key constructs is recommended in various industry settings,
experienced by social media users would provide a more comprehensive insight into
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
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, &
In the next section, the outline of the thesis is presented, including the key topics
CHAPTER 2 provides a review of the literature regarding social media, UGT and
customer engagement. The emergence of social media and specific social networking
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
engagement.
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,
CHAPTER 4 presents the conceptual model of the thesis, outlining the specific
CHAPTER 5 describes the research design used to identify and examine the
relationships between the key constructs. This chapter provides the philosophical stance
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)
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
also demonstrates the moderating effects of community size, media richness and
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
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
14 | P a g e
Chapter 2: Literature Review
applied to user motivations for social media engagement. Relevant literature concerning
context.
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
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
literature. Social media has given consumers a rise in power, flexibility and visibility
regarding marketing content, changing the way individuals and organisations interact.
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
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
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
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
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
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
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
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
“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).
socialisation and navigation (Thelwall & Stuart, 2009). Networking involves social
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
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
description. For example, Ryan’s (2014) definition extends upon the description of
Kaplan and Haenlein (2010) to explain exactly how consumers share their opinions,
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
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
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
Social media take on many forms including blogs, business networks, enterprise social
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
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
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
Personal profiles established on SNSs include photos, video, audio files and blogs
(Kaplan and Haenlein, 2010). Consumers are increasingly becoming more active
loops and highly immediate communication (Brodie et al., 2013). The interactive
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 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,
strategies in this forum remains limited. This research focuses on one specific social
The world’s largest SNS, Facebook (Hughes, Rowe, Batey, & Lee, 2012), was
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
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
considering Facebook as the most attractive SMS for B2C marketing purposes
(Lillevalja, 2010). Over 40 million brands, globally have company pages registered with
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
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
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
with specific brands (Brodie et al., 2013), the extent to how much and how often this
brand pages has a measurable, beneficial influence on the brands they promote (Jahn &
marketers. Recent studies have indicated that less than five percent of customers engage
(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
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
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.
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
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
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,
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
(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
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
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 &
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 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
Researchers have examined the psychological and behavioural aspects of internet users
(LaRose, Mastro, & Eastin, 2001; Lin, 1999). Items such as social escapism,
motivations for internet use (Korgaonkar & Wolin, 1999). Further, motivations such as
been suggested (Papacharissi & Rubin, 2000). The effect of these motivations have
attitude towards brands, purchase intention and satisfaction (Ko et al., 2005; Luo, 2002),
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
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, &
theoretical underpinnings of UGT, such as the need for social interaction, need for
entertainment, information seeking and sharing needs, and desire for reward or
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
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
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
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
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
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
information seeking are transferable to the field of customer engagement. Within this
needs to engage with a brand on social media. This engagement is most likely to
longer, reading details and threads and using multimedia features (Ko et al., 2005). This
engagement with the brand, rather than active engagement in the form of commenting
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 &
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,
Through the application of UGT (Katz & Foulkes, 1962) in the context of brand
communities and social media, authors have shown that consuming entertaining content
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
antecedents to customer engagement behaviour. Entertaining ads are said to lead users
Remuneration
the level of remuneration offered to the consumers has been studied as a driver of
Consumers may engage in social media use as they expect to gain some kind of reward,
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,
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
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
family and society, seeking support and substituting for real-life partnership (Muntinga
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 involves motivations such as getting peer support from others, meeting
interesting people, belonging to a community, talking about something with others and
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
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
discussion or forums. These studies suggest that the social gratification is a significant
34 | P a g e
Chapter 2: Literature Review
foundations, conceptualisation, dimensionality and related concepts.
(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
literature and practice (Brodie et al., 2011). In recent literature exploring customer
engagement, authors have focussed their attention on defining the concept, in addition
foundations of the concept, and outlines the definitions and dimensionality of customer
35 | P a g e
Chapter 2: Literature Review
2.4.1 Customer Engagement Theoretical Foundations
and the service-dominant (S-D) logic perspective (Ashley, Noble, Donthu, & Lemon,
Montgomery, 1999; Deshpandé, Farley, & Webster, 1993) in line with the relationship
which establish, develop and maintain successful relational interaction (Morgan &
Hunt, 1994). Within this broadened relationship marketing notion, the firm focuses on
provided with the value created by firms and act as receivers (Lusch, 2007), the S-D
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
36 | P a g e
Chapter 2: Literature Review
2.4.2 Engagement Conceptualisation
and contexts. For example, organisational behaviour literature suggests that engagement
as an initiative and adequate response to social stimuli (Jennings & Stoker, 2004).
(Hollebeek, 2011b), consumer engagement (Vivek et al., 2012) and simply engagement
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
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
(2011b) states that customer brand engagement is ‘the level of an individual customer’s
37 | P a g e
Chapter 2: Literature Review
levels of cognitive, emotional and behavioural activity in direct brand interactions’
engagement literature (Brodie et al., 2011). Despite the widely accepted three
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
38 | P a g e
Chapter 2: Literature Review
Table 2.3 Customer Engagement Definitions
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:
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
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)
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
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
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
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
varied interpretations of what customer engagement is, and is not. Specifically, scholars
have focussed attention on providing a clear distinction between the relatively new
commitment, interactivity and flow. This section of the literature review addresses these
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
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
objects and subjects (Mollen & Wilson, 2010). Involvement has been suggested as an
Kyner, 1973). Loyalty does not reflect part of the engagement concept itself. Rather it is
customers (Brodie et al., 2011). However, some discrepancy occurs regarding its
conceptual relationship with engagement. Authors have also proposed that satisfaction
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
encapsulate this notion, only reflecting the behavioural dimension. Additionally, Vivek
(2009) suggests that participation refers to customers’ connections with the firm in
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
interest and relevance. However it does not incorporate behavioural and cognitive
consequence of customer engagement (Brodie et al., 2013; Chan & Li, 2010).
engagement. While there is little consensus about the definition of interactivity in the
their actions” (Mollen and Wilson, 2010. p.5). This concept appears to capture the
psychological and motivational elements as captured within the emotional and cognitive
customer engagement.
attention, a clear mind, mind and body unison, effortless concentration, complete
psychological state (Mollen & Wilson, 2010). Flow may act as an antecedent to
engagement.
Discussion surrounding the closely related concepts to customer engagement has given
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
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
(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
engagement, and constructs that have been categorised as both antecedents and
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)
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)
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)
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)
50 | P a g e
Chapter 2: Literature Review
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
cognitive and emotional aspects (Brodie et al., 2011; Hollebeek, 2011b; Patterson et al.,
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)
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
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
consideration, as customers and firms rapidly and easily interact online (Verhoef,
Reinartz, & Krafft, 2010). The achievement of customer engagement is said to incur
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
recent years has precipitated a paradigm shift, significantly altering the way customers
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.
engagement. The chapter began by discussing the emergence of social media, its
from the foregoing literature presented in this chapter that a significant challenge for
marketing academics and practitioners has emerged, with recent academic enquiry
research has explored the conceptualisation, related concepts, and antecedents and
with social media are only beginning to emerge. In particular, there is a need to develop
This chapter examined the relevant literature regarding social media, UGT and customer
recent UGT literature; information seeking, entertainment seeking, desire for economic
reward or remuneration, and desire for social and relational interaction. Through
53 | P a g e
Chapter 3: Social Media Engagement Behaviour
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
In this chapter, the focus is on one touch-point; social media. Extending from the
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-
of the nature of engagement at different intensities and with different valence (Brodie et
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
54 | P a g e
Chapter 3: Social Media Engagement Behaviour
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.
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
between two levels of engagement, lower and higher. Lower engagement describes
higher engagement occurs in cases when customers more actively process the role of the
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
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
behaviour.
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
Extant literature pays little attention to negatively-valenced behaviours that involve such
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
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
Studies addressing the processes and levels of social behaviour in the online context
engagement behaviours offers managerial clarity regarding exactly how users engage,
both positively and negatively, within social media platforms. The construct includes a
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
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
Creating
Users who create original content within social media platforms exhibit the highest
(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
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
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
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
knowledge resources with suppliers and/or customers that is conducive to the firm
(Meeus, Oerlemans, & Hage, 2001). Social media platforms have been credited with
tacit knowledge (Ribiere & Tuggle, 2010), enhancing information and knowledge
and experiences (Brodie et al., 2013; Jaakkola & Alexander, 2014). Users share their
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.
ways of using products and brands (Sashi, 2012). This form of content creation is aimed
(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
other acquaintances (Vollmer & Precourt, 2008). Further, advocating focal brands and
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
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).
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 &
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
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).
products, services, brands or brand meanings (Brodie et al., 2013). Users engaged in co-
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
Affirming: Affirming denotes the specific creation of content by users with the aim of
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
whereby users and customers recognise and communicate the brand’s skills, worth and
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
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
search, identifies relevant social media platforms, and posts a comment. The nature of
the posted comment and resultant conversations characterise the creation of knowledge
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
Contributing
contribute to existing content. However, they do not create any additional or new
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
Consuming
reading reviews, discussion and comments, viewing photos, watching videos and
clicking on content and links. Consuming reflects the minimum level of positively-
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
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
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
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,
temporary state of inactive engagement (Brodie et al., 2013). This state does not
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
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
Detaching
yet privately removing themselves from social media brand pages through selecting to
hide brand related content, or ‘unlike’ and ‘unsubscribe’ from the page. Detachment
a decision to terminate their interaction with the brand, meaning there is a temporary or
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.
scholarly research through the concept of relationship ending (Duck & Perlman, 1985).
more temporary in nature. Consumers may choose to re-join the community, following
Destructing
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
70 | P a g e
Chapter 3: Social Media Engagement Behaviour
destruction, or diminishment of value by providers and customers (Plé & Cáceres,
and the proximity of others, and attitudes to the business in general (Lau & Ng, 2001).
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
The proposed construct of SMEB contributes to the literature through the development
explored the customer engagement cycle, a construct for understanding both positive
and negative SMEB has not yet been developed. The typology captures both positively-
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
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
72 | P a g e
Chapter 4: Conceptual Model Development
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
This chapter begins by introducing the conceptual model of social media content and
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
media content derived from UGT as antecedents to SMEB. Social media content
content. The impact of each of these content types on each of the types of SMEB
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
content. The conceptual model depicts the relationships between the four categories of
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.
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
4.3 Hypotheses
4.3.1 Informational Content
content to viewers and advertising acceptance has been well documented (Bauer &
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
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
conversing with other users in the same way as entertaining or relational content.
SMEB (H1).
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
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
76 | P a g e
Chapter 4: Conceptual Model Development
information overload refers to the delivery of too many communications, causing an
that they can easily respond to (Hiltz & Turoff, 1985). In both traditional and electronic
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
predicted by H1.
H1a: High levels of informational content weaken the relationship with positively-
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
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).
engage with the content are met. Hence, brand fans may exhibit a more positive
Based on this foundation, it can be hypothesised that entertaining brand posts will
It is further hypothesised that with greater levels of entertaining content, the relationship
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-
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
& Gurhan Canli, 2013). Customers may engage in social media as they expect to gain
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
community members, a stronger effect has been observed for passive compared to
active members (Aksoy et al., 2013). Monetary rewards decrease active online
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
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
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
(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;
In line with H1a and H2a as previously discussed, with increased levels of content it can
H3a: High levels of remunerative content weaken the relationship with positively-
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.,
develop attitudes, norms and community language (Brodie et al., 2013). Customers who
camaraderie with other users (Chen, 2011). Customers highly motivated by socialisation
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
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
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
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
this rationale, it has been suggested that when faced with conflicting processing
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
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
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
Media Richness
Media Richness Theory (Daft & Lengel, 1986) is based on the assumption that the goal
states that media differ in the degree of richness they possess, measured by the amount
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 &
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
provide both ‘rich’ and ‘lean’ advertising and marketing content, within a single media
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
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
Michahelles, 2013; De Vries et al., 2012) and clicking (Cho, 1999; Lohtia et al., 2003) it
H6: The strength of the relationship between social media content and social media
Congruity
The extent to which social media content is congruent or incongruent to the brand may
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
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
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
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).
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
H7: The strength of the relationship between social media content and social media
Community Size
Research in sociology has shown that increased community size has a negative effect
greater tendency for social engagement (Dholakia et al., 2004). Members of a smaller
higher level of brand engagement. In larger communities, scholars have suggested that
In smaller online communities, users may join more often for friendship and
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
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
that the strength of the relationship between social media content and SMEB will
H8: The strength of the relationship between social media content and social media
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
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.
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
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
coding, sampling, training, and reliability testing and reporting decisions are presented.
The chapter then presents the methods adopted for hypothesis testing: binary logistic
2. To offer new insights into the characteristics, levels and valence of social media
engagement behaviour.
behaviour.
In order to achieve these aims, two phases of research design are required. Phase 1
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
developed in Chapter 4.
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
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.
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
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
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
this forum with little guidance and empirical understanding (Stelzner, 2014).
Regardless, over 80 percent of brands are said to use the social media platform
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
of how to effectively engineer content, managers will fail in their endeavours to attract
Whilst studies have demonstrated that social media has been accepted and is widely
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.
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
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
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
Facebook Insights
enable high-level monitoring of the activities on the Facebook page. Facebook Insights
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.
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
NCapture
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
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.
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.
Facebook Page Insights also allowed access to specific post metrics for each of the
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
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
Content analysis is most commonly defined as a research technique for the objective,
(Berelson, 1952). The objective, systematic and quantitative nature of content analysis
remains central to other definitions of the concept, including Carney (1972) who
Dunphy, Smith, and Ogilvie (1968) who describe content analysis as a research
(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
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
numeric value according to valid measurement rules, and the analysis of those
relationships involving those values using statistical methods, in order to describe the
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
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
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.
Neuendorf (2002); determination of the appropriate theory and rationale for the
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
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.
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
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.
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
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
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
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
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
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
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
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
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;
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
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
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
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.
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
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
Relational content includes the use of emotion. In order to develop a custom dictionary
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
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
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
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].
The dependent variable of the study, SMEB merges both positively- and negatively-
user typologies and social media user categorisations (See Chapter 2) provides a useful
measures. The construct of SMEB consists of three positively-valenced, one neutral and
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.
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
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
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 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
Congruity Operationalisation
Small Less than 1,500 fans
Medium 1,500 to 10,000 fans
Large Over 10,000 fans
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
114 | P a g e
Chapter 5: Research Design
Table 5.14 Image Coding Scheme
quantitative content analysis techniques, eliminating coding errors and enabling the
content coding schemes, coding rules may be formalised, permitting perfect coding
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
community member comments and images embedded within the social media content,
In consideration of the nature of the sample, the research questions of the study and the
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
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
=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
=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
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
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
Training and initial reliability is required for human coding of social media content. A
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
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
118 | P a g e
Chapter 5: Research Design
formulas have been developed in recent literature for measuring inter-coder reliability.
reliability which many researchers regard as more useful than percentage agreement
International, 2015):
Calculate the expected frequency by which the agreement between users could have
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
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
1. Total units of agreement between the two users (TA) minus the expected
units (TU) within the source minus the expected frequency (ΣEF) of the
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
Korenstein, & Meade, 2008; Viera & Garrett, 2005) and thus is utilised in this study.
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
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
Social media content was coded with informational, entertaining, remunerative and
relational custom dictionaries. Descriptive results were generated using the analyse,
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
content category.
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
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
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
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
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.
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.
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.
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
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
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
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
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
128 | P a g e
Chapter 5: Research Design
outlined in this chapter allowed the data to be prepared for hypothesis testing.
Hypotheses 1 to 5 were analysed using binary logistic regression using IBM SPSS
a categorical (nominal or nonmetric) variable, and the independent variables are metric
or nonmetric variables (Hair, Black, Babin, Anderson, & Tatham, 2006). Logistic
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
Logistic regression does not assume a linear relationship between the dependent and
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.
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
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.
A summary of the coding for the control variables can be found in Table 5.26.
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
To test H6, H7 and H8 concerning moderation effects, Process Model 2 was applied.
addition to their integration in the form of a conditional process model (Hayes, 2013).
standard errors, t and p-values and confidence intervals using OLS regression and/or
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
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,
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
In order to construct this model, indicator coding of the moderator variables was
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
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
The results of the conditional process analysis and consequent hypothesis testing are
presented in Chapter 6.
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
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
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
remunerative social media content types on SMEB are also tested. A series of adapted
media richness, congruity and community size. These effects are subsequently
discussed. The chapter concludes with an overview of the hypotheses and summary of
results.
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
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
The results for Hypotheses 1 to 4 are presented in the following sections, with
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
consumer response, it was hypothesised that informational social media content will
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,
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
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
Creating behaviour occurs when users engage with brands and other users by creating
comment. Creating epitomises a highly active level of SMEB. Table 6.1 shows that the
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
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.
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
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
an Exp(B) of 1.91 indicating that when posts have informational content, they are 1.9
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
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
139 | P a g e
Chapter 6: Results
Table 6.2 Logistic Regression Showing Effect of Informational Content Presence
on Contributing Behaviour (Shares)
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
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.
through reading reviews, discussion and comments, viewing photos, watching videos
and clicking on content and links. Consuming reflects the minimum level of positively-
active contribution to or creation of content. Table 6.4 shows the effect of informational
informational content are 2.27 times more likely to result in consuming behaviour,
141 | P a g e
Chapter 6: Results
Table 6.4 Logistic Regression Showing Effect of Informational Content Presence
on Consuming 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
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)
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
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
increased in April (1.56 times more likely) and July (1.90 times more likely) compared
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,
Table 6.6 shows the results of the binary logistic regression predicting the occurrence 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
Table 6.7 shows the summarised results for the main effects between informational
positively predicted the occurrence of creating, contributing (likes and shares) and
present.
The summarised results in Table 6.7 show that informational content presence is a
investigating the level of informational content within a post, and how this may alter the
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
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
145 | P a g e
Chapter 6: Results
Hypothesis 1a suggested that high levels of informational content weaken the
occurs it is likely that a reduction in decision quality and attention will occur (Speier et
al., 1999).
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.
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
and positively-valenced SMEB in the form of creating and contributing. As the level of
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
contributing behaviour in the form of shares, the odds of sharing behaviour increased in
147 | P a g e
Chapter 6: Results
increased. For informational content levels of 6 or greater, there was no significant
H1a is therefore partially supported. The results in Table 6.8 show that a maximum
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
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
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
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
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
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
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.
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
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
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
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
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
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
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
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
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,
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
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
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
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
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
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
156 | P a g e
Chapter 6: Results
Hypothesis 2a suggested that high levels of entertaining content weaken the relationship
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
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
shares occurring. This suggests that there is a maximum level of entertaining content
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
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,
Tables 6.17 through 6.22 show the results for entertaining content and creating,
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
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,
form of creating.
Table 6.18 presents the results of the logistic regression concerning remunerative
Exp(B) of 1.77 indicating that when posts have remunerative content, they are 1.77
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
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
Table 6.20 presents the results of the binary logistic regression predicting consuming
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
Table 6.21 presents the results of the logistic regression, testing the effect of
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
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
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
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
Table 6.22 indicates that the presence of remunerative content is not a statistically
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,
There was no significant effect for post day of the week or post time of the day in
164 | P a g e
Chapter 6: Results
Table 6.22 Logistic Regression Showing Effect of Remunerative Content on
Detaching Behaviour
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
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.
This result suggests that remunerative content can facilitate the occurrence of active,
165 | P a g e
Chapter 6: Results
no significant relationship between remunerative content and consuming behaviour.
for example ‘share/like to win’ or ‘share/like for discount/promotion’ etc., causing the
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.
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
H3a: High levels of remunerative content weaken the relationship with positively-
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
In the following section, the results regarding remunerative content level and its effect
H3a suggested that high levels of remunerative content weaken the relationship with
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
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
behaviour. Hypothesis 3a: high levels of remunerative content weaken the relationship
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
hypothesised that relational social media content which stimulates interaction amongst
168 | P a g e
Chapter 6: Results
Tables 6.25 through 6.30 show the results for relational content and creating,
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
Table 6.25 shows the results of the binary logistic regression predicting the occurrence
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
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
169 | P a g e
Chapter 6: Results
Table 6.25 Logistic Regression Showing Effect of Relational Content Presence on
Creating Behaviour
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)
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
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
Users click on photos, read posts, click on links and watch videos when they engage in
consuming SMEB.
consuming behaviour, with the odds ratio of 3.73 indicating that for posts with
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
172 | P a g e
Chapter 6: Results
Table 6.28 Logistic Regression Showing Effect of Relational Content Presence on
Consuming Behaviour
Relational content presence within social media is not a statistically significant predictor
users who are exposed to the social media post (which contains relational content) and
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),
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
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
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
occurring.
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
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
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
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
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,
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
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
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-
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
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
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%.
This section presented the results of binary logistic regression which detailed the effects
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
occurring also increased. Lower levels of entertaining content were required in order to
information overload effect. The level of 4 for entertaining content had the greatest
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
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
181 | P a g e
Chapter 6: Results
The prediction of contributing behaviour in the form of likes was also dependent on the
occurrence of contributing behaviour in the form of likes occurring, while the level of
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
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
content within a post did however have an effect on the likelihood of consuming
behaviour occurring.
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
consumer. Thus:
The results regarding the interaction effects are summarised in Table 6.34 and described
183 | P a g e
Chapter 6: Results
Table 6.34 Interaction Effects Summary
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
results indicate that the odds of creating behaviour occurring when the post contains one
informational content decreased the odds of creating behaviour occurring by 70% and
These findings suggest a negative effect of combined conflicting content types such as
presence of informational and entertaining content within a post were previously found
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
results indicate that combining entertaining content with informational content at certain
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
185 | P a g e
Chapter 6: Results
behaviour occurring is decreased by 68%.
SMEB in the form of dormancy occurring, the interaction reduced the odds of
and entertaining content has a negative effect on the likelihood of active, positively-
occurring.
H5b predicted that the simultaneous presence of informational and relational content
presence of informational and relational content within a post did not have a statistically
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
behaviour occurring are increased by 2.3 times. When the level of informational content
content, the odds of dormant behaviour occurring increase by 3.71 times. This level of
increased the odds of dormant behaviour occurring by 65.48 times. There was no
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.
‘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
SMEB. H5c is therefore not supported. This demonstrates that there is no benefit of
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.
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
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
supported. This finding indicates that it is not necessarily detrimental to the brand to
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
The results, presented in Table 6.34, indicate a significant effect of the simultaneous
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
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
The simultaneous presence of entertaining content and relational content within a post
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
However this effect was dependent on the specific levels of entertaining and relational
189 | P a g e
Chapter 6: Results
The simultaneous presence of relational and remunerative content did not have a
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
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)
190 | P a g e
Chapter 6: Results
H5b: The presence of informational and relational content facilitates negatively-
valenced social media engagement behaviours.
The results in Table 6.34 indicate partial support for H5e, while H5a, H5b, H5c, H5d
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
modifying the PROCESS model to estimate a simple moderation model with a 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
Hayes (2013) PROCESS macro for SPSS provides a means for estimating a model with
(media richness, congruity and community size). Hayes (2015) method of ‘hacking’
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,
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
192 | P a g e
Chapter 6: Results
Table 6.35 Richness Operationalisation
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
SMEB, by media richness were achieved. Among the tests of moderation, richness was
behaviour in the form of shares. The results of this moderation effect are presented in
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
behaviour in the form of sharing occurring. Table 6.36 shows the test of interaction
Table 6.36 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Shares) by Media
Richness
The “Both” row in Table 6.36 provides a test of the null hypothesis that media richness
behaviour in the form of shares. The null hypothesis can be rejected, F (2, 1823) =3.99,
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),
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
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
each of the three groups. In this model, the conditional effect of X (informational
in Table 6.38.
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
(shares) is 0.07 with a standard error of 0.02. This is statistically different from zero, t =
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
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
For posts with more than 2 pieces of information, the most effective media type in order
This section demonstrates that media richness moderates the relationship between
informational content and contributing behaviour in the form of shares. The effect of
richness (categorised as video posts in this study). This finding indicates that while a
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
SMEB. Similarly, media richness did not moderate the relationships between
H6: The strength of the relationship between social media content and social media
6.4.3 Congruity
The extent to which social media content is congruent or incongruent to the brand may
informational, remunerative and relational posts may vary in their degree of 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
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
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
SMEB by congruity were achieved. Among the tests of moderation, congruity was
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.
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
to 5 are used as the independent variable in this model which tests the moderation of the
(2015) PROCESS model was applied. Table 6.40 shows the test of interaction between
Table 6.40 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Informational Content on Contributing Behaviour (Likes) by Congruity
The “Both” row in Table 6.40 provides a test of the null hypothesis that congruity does
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
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
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,
199 | P a g e
Chapter 6: Results
Table 6.41 PROCESS Data for Visualising Conditional Effect of Informational
Content on Contributing Behaviour (Likes)
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 three groups. In this model, the conditional effect of X (informational content) on Y
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
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
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
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
201 | P a g e
Chapter 6: Results
informational content that are related to the category (medium congruity) have a greater
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
However, as the level of information within a post increases, medium and low 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
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
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
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
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)
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
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
Therefore, posts with entertaining content that are related to the brand or specific
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
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
content and other SMEBs. Further, there were no statistically significant results to
indicate that congruity moderates the relationships between relational and remunerative
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
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
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.
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
Community size has been found to negatively impact the level of interactions between
al., 2004). Social media users are therefore likely to be more connected to a smaller
community size (small, medium and large) moderates the relationships between social
H8: The strength of the relationship between social media content and social media
206 | P a g e
Chapter 6: Results
Using the SPSS code to construct PROCESS Model 2 for a multi-categorical
community size were achieved. Among the tests of moderation, community size was
on dormant behaviour. The results of these moderation effects are presented in the
following sections.
The presence of entertaining content within social media posts was found to be a
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
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
Table 6.46 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Likes) by Community
Size
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
content depends on community size. This means that community size significantly
(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
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
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
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
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
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
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
Comparatively, when the community size has more than 1,500 fans, providing higher
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.
large). The levels of entertaining content (1, 2 and 3) used as the independent variable in
Table 6.49 shows the test of interaction between community size and contributing
Table 6.49 Partial Output From PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Contributing Behaviour (Shares) by Community
Size.
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
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).
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
Table 6.51.
The output in the first row of Table 6.51 indicates that the conditional effect 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
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
users, which reduces the number of shares and likes made on the content.
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
Table 6.52 Partial Output from PROCESS Model 2 Examining Moderation of the
Effect of Entertaining Content on Dormant Behaviour by Community Size
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
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).
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
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
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.
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
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
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
Community size was found to moderate the relationships between entertaining content
behaviour. There were no statistically significant results to indicate that community size
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
demonstrated a benefit for large and medium community sizes, with the presence of
Comparatively, for small community sizes, entertaining content within a post was
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
Instead, the relationship between social media content and SMEB was positively
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
moderating variables.
The media richness of a post was found to positively moderate the effect of
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
creating behaviour, with highly congruent posts (related to the specific product/brand)
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
reduced.
219 | P a g e
Chapter 6: 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
There was partial support found for Hypothesis 1a which suggested that high levels of
facilitates active, positively-valenced SMEB was partially supported. There was partial
support for Hypothesis 2a which suggested that high levels of entertaining content
which predicted that the presence of remunerative content would facilitate passive,
supported, suggesting that high levels of remunerative content weaken the relationship
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
valenced engagement behaviours. Finally, partial support was found for the moderation
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
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
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
construct, with cognitive, affective and behavioural components (e.g. Brodie et al.,
2011). This thesis focused exclusively on furthering the understanding of one dimension
definition and preliminary investigation into the nature of engagement behaviours (E.g.
Jaakkola & Alexander, 2014), there still remains much to understand about these
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
(Hollebeek & Chen, 2014; van Doorn et al., 2010) and engagement intensity (Malthouse
The development of the SMEB construct and corresponding six behaviours enhances
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
moderate intensity; detaching and contributing. Finally, the construct demonstrates the
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.,
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
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
(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.
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
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
explanation of why and how individual users interact and engage with various forms of
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
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
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
7.2.3 Establishment of the Relationship between Social Media Content and Social
Media Engagement Behaviour
categories and SMEB. An association between content and user actions such as ‘liking’
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
levels, and interactions, and SMEBs, demonstrating that content should be designed in a
(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
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
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
The presence of informational content within social media posts was found to predict
previous literature which has stated that internet users increase their usage patterns as a
(Stafford et al., 2004). Similarly, the presence of entertaining content within a social
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
also identified. The presence of relational content had a significant effect on all
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
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
the specific amount of content delivered to users through social media posts is an
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 content delivered to users increased, high media richness levels were
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
creating behaviour, with highly congruent posts (related to the specific product/brand)
having the greatest effect on creating behaviour. This finding demonstrates the benefit
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
reduced.
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
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
literature, such as the mere number of likes, shares of comments used as engagement
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
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
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
categorise, code and measure content through the development of the process for
automatically coding the presence of content attributes and elements. The codes and
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
interaction which were not included in this study. Factors such as personal identity
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
investigation of social media engagement within this thesis fails to consider users
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
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
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
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
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
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
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.
A valuable area for further research concerns the dynamic nature of the levels of
progression through, or within the six behaviours is not considered. This could be
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,
that the customer engagement process may vary for existing, compared to new
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
The notion of user progression through various stages of SMEB has parallels with
such as customer bonding (Szmigin, Canning, & Reppel, 2005), relationship ending
(Duck & Perlman, 1985), and customer exit processes such as relationship termination,
Pura, & Van Riel, 2004) would be valuable in further research exploring user
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
by a range of factors beyond social media content. For example, customer expressions
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
beyond social media content which may facilitate positively- and negatively-valenced
(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),
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
would also provide further insight into the relationship between content and SMEB.
Further examination of content gratifications beyond those tested within this thesis
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
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
motives and resultant coding of social media content within this thesis encapsulated the
four main gratifications observed in recent literature stemming from the UGT
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
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
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
the brand, product or social media brand page, and future purchase intentions.
Another valuable area for further investigation concerns the adoption of the three
and examining the influence of social media content on cognitive and affective
consequently could be a better predictor of future behaviour. While this thesis provides
239 | P a g e
Chapter 7: Discussion and Conclusion
would provide a far more comprehensive insight into the overarching customer
engagement concept.
wine brands, the results have use for managers in similar settings. While implications
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
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
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
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
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
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
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
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
suggest that if managers provide entertaining social media content, users consuming
behaviour will change. The findings also demonstrated the importance of the
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
242 | P a g e
Chapter 7: Discussion and Conclusion
The provision of relational content had some of the strongest effects on predicting
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
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
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
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.
informational content and entertaining content will increase the likelihood of users
remaining dormant and therefore essentially ‘scrolling’ straight past the content.
exclusively from each other, rather than attempting to provide posts that contain both
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
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
Posting entertaining content that has nothing to do with the product or brand is therefore
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.
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
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
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
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
(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
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
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.
concept, this thesis extends the ideas through empirical quantitative enquiry. It explores
248 | P a g e
Appendices
Appendices
Hello______,
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
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
252 | P a g e
Appendices
253 | P a g e
Appendices
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
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
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
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
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.
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
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
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