University of Southern California Dissertations and Theses

The social meaning of sharing and geocoding: features and social processes in online communities

(USC Thesis Other)

The social meaning of sharing and geocoding: features and social processes in online communities

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE SOCIAL MEANING OF SHARING AND GEOCODING:
FEATURES AND SOCIAL PROCESSES IN ONLINE COMMUNITIES

by
Li Xiong
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(COMMUNICATION)
December 2012
Copyright 2012 Li Xiong
ii
Table of Contents
Acknowledgements ............................................................................................................. v

List of Tables ..................................................................................................................... vi

List of Figures ................................................................................................................... vii

Abstract ............................................................................................................................ viii

Chapter One: Introduction ............................................................................................... 1

Background ............................................................................................................. 1

Overview ................................................................................................................. 3

Practical Motivation ................................................................................................ 5

Theoretical Motivation ............................................................................................ 9

Chapter Two: Reputation Systems, Generalized Reciprocity and
Location Awareness ................................................................................. 17

Introduction ........................................................................................................... 17

Reputation Systems ............................................................................................... 22

The Use of Reputation Systems ............................................................................ 26

Location Awareness and Reputation Systems ...................................................... 29

Location Awareness and Generalized Reciprocity ............................................... 33

Chapter Three: Social Matching Systems, Community Attachment and
Location Awareness ................................................................................. 37

Introduction ........................................................................................................... 37

New Media and the Development of Online Communities .................................. 40

Tags and Social Matching Systems ...................................................................... 43

Location Awareness and Social Matching ............................................................ 48

Location Awareness and Community Development ............................................ 52

Chapter Four: Boundary Filters, Social Capital and Location Awareness ..................... 57

Introduction ........................................................................................................... 57

Structural and Technical Aspects of Social Capital .............................................. 60

Privacy and Social Capital .................................................................................... 66

Privacy in Online Communities ............................................................................ 67

Location Awareness and Privacy .......................................................................... 73

Chapter Five: Information Sharing in Online Communities .......................................... 78

Introduction ........................................................................................................... 78

Social Influence on Information sharing ............................................................... 80

iii
Social Shaping of Motivations for Contribution ................................................... 82

Online Information Sharing as ICT-enabled Collective Action ........................... 86

Hypotheses ............................................................................................................ 89

Chapter Six: Method ......................................................................................................... 99

Technical Overview .............................................................................................. 99

Design ................................................................................................................. 106

Subjects ............................................................................................................... 106

Procedure ............................................................................................................ 107

Material and Measurement ................................................................................. 110

Analysis............................................................................................................... 114

Chapter Seven: Results ................................................................................................... 118

Manipulation Check ............................................................................................ 118

Main effect .................................................................................................. 118

Moderation effect ........................................................................................ 119

Reputation Systems, Reciprocity and Location Awareness ............................... 120

Social Matching Systems, Social Identification, Interpersonal
Bonds and Location Awareness .......................................................................... 120

Privacy Controls, Social Capital and Location Awareness ................................ 122

Information Contribution .................................................................................... 123

Quality. ........................................................................................................ 123

Quantity. ...................................................................................................... 124

Group differences. ....................................................................................... 125

Difference in the shared focus .................................................................... 126

Difference in the use of geolocation features ............................................. 126

Chapter Eight: Discussion ............................................................................................... 129

Summary ............................................................................................................. 129

Reputation Systems and Reciprocity .................................................................. 133

The meaning of reputation systems ............................................................ 133

The meaning of generalized reciprocity ..................................................... 135

Building reputation systems online ............................................................. 137

Tags, Social Matching Systems and Community Attachment ............................ 138

Social matching systems in practice ........................................................... 138

The development of online communities .................................................... 141

Privacy and Social Capital .................................................................................. 144

The meaning of privacy online ................................................................... 144

The meaning of social capital online .......................................................... 146

The Social Meaning of Geocoding ..................................................................... 149

Geocoding as a social “add-on” .................................................................. 149

From geocoding to geotagging ................................................................... 154

iv
The Social Meaning of Sharing .......................................................................... 159

Quality and quantity of sharing ................................................................... 159

Sharing in different contexts of use ............................................................ 162

Chapter Nine: Conclusions ............................................................................................. 166

Summary ............................................................................................................. 166

Theoretical Significance ..................................................................................... 167

Practical Relevance ............................................................................................. 169

Limitations .......................................................................................................... 171

Future Research .................................................................................................. 173

Bibliography ................................................................................................................... 177

Appendices ...................................................................................................................... 201

Appendix A: Reliability, Convergent Validity and Discriminant
Validity for the Measurement Model .................................................................. 201

Appendix B: Measurement Items ....................................................................... 204

Appendix C: Un-hypothesized Paths .................................................................. 209

v
Acknowledgements
This dissertation research was supported with a grant from Annenberg Program
for Online Communities (APOC). I would like to thank Professor Dmitri Williams for his
advice, support and help intellectually, professionally and personally. I want to thank
Professor Margaret McLaughlin and Professor Andrea Hollingshead for their continuous
support and understanding. I would also like to thank Professor Ann Majchrzak for her
insightful suggestions and comments. Finally, I thank my wife Yiye Liu and my son Evan
for their priceless support and encouragement during a very long and interesting process.
vi
List of Tables
Table 1: Test Results for Supported Hypotheses ............................................................ 128

Table 2: Reliability ......................................................................................................... 201

Table 3: Convergent Validity .......................................................................................... 201

Table 4: Discriminant Validity ....................................................................................... 203

Table 5: Test Results for Un-hypothesized Paths ........................................................... 209

vii
List of Figures
Figure 1: Reputation Systems, Reciprocity and Location Awareness .............................. 36

Figure 2: Social Matching Systems, Interpersonal Bonds, Social
Identification and Location Awareness ............................................................. 56

Figure 3: Privacy Controls, Social Capital and Location Awareness ............................... 77

Figure 4: Perceived Quality and Actual Quantity of Sharing ........................................... 98

Figure 5: Screenshot of a Voilai.com posting ................................................................... 99

Figure 6: Screenshot of Reputation Features .................................................................. 100

Figure 7: Screenshot of Social Matching Systems ......................................................... 101

Figure 8: Screenshot of Privacy Systems ........................................................................ 102

Figure 9: Screenshot of the Filtering Interface ............................................................... 103

Figure 10: Main Interface of Voilah.com ....................................................................... 104

Figure 11: Main Interface of Voilai.com ........................................................................ 105

Figure 12: Theoretical Model ......................................................................................... 117

Figure 13: Overall Structural Model Results .................................................................. 127

viii
Abstract
This study examines how emergent communities might show different patterns of
uses and perceptions for communication and profile features when geolocation features
are used. It explores the ways that location awareness moderates the social and cognitive
processes that motivate people’s participation in the sharing of personal information
online. In a field experiment (N = 94), participants were assigned to use customized
websites with or without geolocation features that filtered information shared by and
shown to the users. Participants’ uses were captured with server logs, and their
perceptions of the respective user communities were assessed with two follow-up
surveys. The results suggest that the use of reputation systems has a positive influence on
the perceived reciprocity in a community, which is even stronger when geolocation
information is attached to user profiles and user-generated content. Second, the use of
social matching systems such as tag-based recommendations of interests and connections
is positively associated with perceived interpersonal bonds and social identities in a
community. Geolocation awareness also reinforces the influence of the use of social
matching systems on perceived interpersonal bonds. Finally, generalized reciprocity and
social identification positively contribute to the perceived quality of information
contribution. The implications for the development of online sharing communities
through the design and use of reputation, recommendation, privacy and geolocation
features are discussed.

1
Chapter One: Introduction
Background
People are drawn to events that happen at a reachable distance, and tend to
communicate more with people who live nearby (Barnlund & Harland, 1963; Byrne &
Buehler, 1955; Hampton & Wellman, 2001; Mok, Wellman, & Basu, 2007; Takhteyev,
Gruzd, & Wellman, 2012). The interest in the local neighborhoods comes from the sense
of similarity, familiarity and relevance that is often associated with geographic proximity
(Festinger, Schachter, & Back, 1950; Guest & Wierzbicki, 1999). When the Internet
began to be widely adopted in the late 1990s, it expanded people’s choices for social
interaction, information access and media consumption (Katz & Rice, 2002). Tools like
email, forums and instant messaging made it possible for people to manage relationships
and acquire information beyond their geographical boundaries (Quan-Hasse, Wellman,
Witte, & Hampton, 2002). Because it eliminated the time and geographic boundaries, the
Internet seemed to foster communities based on shared interests and objectives, and not
simply on geographical or ethnic bonds (Wellman, et al., 2003). This phenomenon
initially made observers worry about the decline in people’s interest in civic actions and
participation in local communities (Nie, Hillygus, & Erbring, 2002; Putnam, 2000).
Furthermore, there was also the concern that local places for casual socialization and
gathering may disappear, because of our increasing reliance on digital media for
information and communication (Oldenberg, 1999).
However, evidence also showed that people often used the Internet to participate in
events in, for and about their local communities (Hampton & Wellman, 2001; Quan-
2
Hasse, et al., 2002). The Internet complemented rather than displaced community
activities and social interactions in physical settings (Hampton, 2007). It simply
facilitated people’s existing social practices and accommodated their interest in the local
environment. At the same time, it gave people more choices of relationships, information
and activities beyond their neighborhoods. That is, the Internet broadened people’s
horizons and facilitated their local interests at the same time. In Wellman’s words, the
Internet is a “glocalizing” technology (Wellman, et al., 2003).
This seemingly contradictory impact becomes even more salient now. On one
hand, Internet use reaches a new level of frequency, intensity and efficiency with the
rising popularity of social media (A. Smith, 2011c). People join various social
networking sites (SNS) and disclose a massive amount of personal details and activities
(boyd, 2006). By using social media, people can efficiently discover events of relevant
interests and connect with people who have shared interests. As a result, they can
organize or join many different kinds of collective actions from political petitions, group-
based sales deals, citizen journalism to celebrity fan clubs (Shirky, 2008). Sharing
information or thoughts becomes an easy yet powerful way of participation in online
collective actions (Shirky, 2011). The Internet now broadens our horizons even at a
greater scale.
On the other hand, exploring physical locations becomes more efficient
(Nakashima, 2008). With companies like Google and Microsoft actively improving and
promoting geospatial technologies, the awareness and use of geolocation functionalities
on the web have been on the rise (Zickuhr & Smith, 2011). Many web and mobile
3
applications are implementing the geotagging functionality to automatically detect users’
locations and provide customized content or services. For example, search engines like
Google and Bing can automatically refine search results based on the IP locations of
users. Geocoding refers to such a process where users’ geographic coordinates are
automatically captured and used to customize services (Goldberg, 2008). This feature
places a strong focus on explicitly presenting users with the specific geolocations. In
comparison, geotagging is more implicit process in which a broader set of information
from geographic coordinates to place names and others nearby is attached to other media.
For example, social networks like Facebook and Twitter can attach a geolocation tag to
whatever content people are publishing. Technologies like geotagging are driving the
trend of the “hyper-local web” (C. C. Miller & Stone, 2009). News agencies, community
organizers, governments and marketers alike are exploring various tools to deliver highly
customized messages and services to their customers over the Internet. Whether or not we
like it or even notice it, the Internet knows wherever we are and gives us refined
knowledge of where we are, and what we do at different places.
Overview
This study examines how emergent communities might show different patterns of
uses and perceptions for communication and profile features when geolocation features
are used or not. Specifically, it explores if and how communities of users engage in
different perceptions and practices when geolocation information is used to filter
information. Rather than showing a direct and straightforward effect, the goal is to
uncover the social-psychological processes that underlie the influence of geocoding
4
features on people’s behaviors and perceptions. Therefore, a basic premise is that location
awareness influences information sharing activities indirectly. This means that people
have to incorporate this new technology in their use of other features that directly
influence how they communicate and present themselves. If the use of certain features
supports some constructive social processes, location awareness will reinforce these
links. This study thus examines the ways that location awareness moderates the social
and cognitive processes that motivate people’s participation in the sharing of personal
information online. Perceived reciprocity, community attachment and social capital
formation are some of the constructive social processes that this study investigates. How
their associations with different features are moderated by geolocation awareness is the
main subject of analysis.
The study employed a field experiment. Participants were assigned to use
customized websites with or without geolocation features that filtered information shared
by and shown to the users. Participants’ uses were captured with server logs, and their
perceptions of the respective user communities were assessed with two follow-up
surveys. With this experiment, the study shows that new features like location awareness
make existing community features more effective in creating constructive social
processes in online communities. For example, users of reputation systems reported a
stronger sense of reciprocity when geotags were attached to user profiles and posts. The
study thus reveals the socio-technical process in which new features are appropriated in
the use of other features that facilitate users’ participation in collective actions.
5
Each chapter in the dissertation discusses the socio-technical processes
surrounding the use of a common feature in online communities, and explores the
potential influence of geolocation technologies. Then a final chapter addresses the
potential influences of these socio-technical processes on usage. Chapter Two tackles the
topic of generalized reciprocity and discusses how its association with reputation systems
is influenced by location awareness. Chapter Three explores the theoretical implications
of developing social attachment based on both interpersonal bonds and common
identities in the use of social matching features, together with location awareness.
Chapter Four examines the impact of the use of privacy controls on location-aware online
communities from the theoretical lens of social capital. Chapter Five lays out the
theoretical framework for why a reciprocal, cohesive and inclusive social environment
motivates information sharing in online communities. Chapter Six describes in detail the
methods of this study. Chapter Seven presents the results and discusses the meaning and
interpretation of the results in the context of the study. Chapter Eight discusses the broad
theoretical and social implications for the findings, pointing out the potential contribution
of my dissertation for communication theories and online community development.
Chapter Nine concludes this study by summarizing the key insights, identifying
limitations and laying out future areas of research.
Practical Motivation
The current media landscape is characterized by the social affordance of
information sharing with more people, and the technological capability of fine-grained
local focus. People can reach out for information and make friends beyond their local
6
neighborhoods. And yet they also have the means to know more about the events,
resources and people in their neighborhoods. The trends of hyper-social activity and local
focus complement each other, because both are becoming easier when mobile
technologies make the web universally accessible. Particularly as the mobile phone
becomes a versatile medium that extends and enriches the PC-based web, it complements
the trends of hyper social activity and location awareness (A. Smith, 2011a). Early
research on the mobile phone reveals that although it facilitates social interaction within
existing social groups, the mobile phone constrains the mutual awareness and
engagement of co-located acquaintances and strangers (R. Ling, 2008). With mobile
phones, intimate ties can stay in touch with one another through frequent, prolonged and
multi-channel conversations, creating a sense of “connected presence” (Licoppe &
Smoreda, 2006). As a result, mobile communication sustains the connections and bonds
in existing groups. In contrast, it provides limited capability for ad hoc social groups to
develop close and enduring relationships (R. Ling & Campbell, 2008). When the mobile
phone is equipped with web connectivity and geolocation capabilities, however, its use
will no longer be confined to communication in existing circles at the cost of attention to
casual, weak ties nearby. Rather, location may become an important context for
discovering diverse social ties in ad hoc communities based on shared interests
(Humphreys & Liao, 2011; Rheingold, 2002).
The affordance of tracking and exploring different places through geocoding
should also make it easier for people to cultivate ties, improve communication and
strengthen affinity, because it will be easier to perceive and configure geographic
7
proximity in searching for information and filtering communication with others. For
example, “checking in” at a popular restaurant on Facebook may strengthen existing
social ties because it keeps a user’s friends updated with her current activities. At the
same time, checking in at this restaurant may help the user uncover latent relationships
with people who have visited this restaurant before and share common interests. This
latent tie may extend well beyond the location of this restaurant and thus broaden the
user’s social confines. Because technology makes it possible for geographic and social
data to move across mobile and PC platforms, the mobile phone effectively satisfies
people’s interests in exploring local events and their needs for a broader set of
relationships and activities. If convergence is defined as the integration of different
technologies for seamlessly transmitting personal data (Pool, 1983), one of the key social
consequences of convergence of the web and the mobile devices is the integration of a
higher degree of social activeness in information sharing and relationship management
with an enhanced focus on people’s nearby environment.
The complementarity of local focus and hyper-social activeness suggests that
location awareness should make people care and know more about their local
neighborhoods and thus increase people’s information sharing activities. Yet reality does
not always support this proposition. Location-based services are still struggling to
identify their purpose, users and business models. Popular social networks like Facebook
and Twitter have stopped making checking-in a unique feature, but rather have made
location into a universal and implicit content tag (Fiveash, 2011). Hyper-local sharing
communities like Patch.com, EveryBlock.com and Hipster.com have seen dropping
8
interest and contribution from users, despite their initial hype several years back
(Kirkpatrick, 2011). The problem is that just by showing “where you are” and ‘where you
have been”, location-based services like Foursquare do not seem to create and sustain
communities in which people find it useful or interesting to share information about their
neighborhoods (Ingram, 2011). In other words, geocoding technologies alone do not
seem to sustain communities. Instead, there must be applications for geotags to make
socialization and information sharing more relevant and interesting. This is an interesting
problem for designers, developers, community organizers, users and researchers to solve.
The uncovering of the social mechanisms behind the features and processes that
foster communities is essential for understanding the current status and future of location-
based technologies. If location awareness enabled by geocoding alone does not seem to
increase usage and build a community, then what can be learnt from successful
communities such as eBay, Quora, Twitter or StackOverflow? What are the key features
that sustain mutual obligation, foster strong social identities, facilitate meaningful
communication and balance people’s need for privacy against the need for new ties and
new information? Will location features like geotags improve or undermine the
effectiveness of these features? Rather than examining straightforward effects of some
features on usage, this study tries to understand the potential processes associated with
these features, namely the processes in which individuals generate certain perceptions
and adjust actions in the use of those features (Sabherwal & Robey, 1995). Then the
relationship between usage and these processes is examined. This understanding is the
9
foundation to designing and building better solutions for effectively organizing location-
based information sharing communities.
Theoretical Motivation
The integration of hyper-social sharing and hyper-local focus across mobile and
web platforms presents an interesting opportunity for the study of new communication
technologies like geocoding. If people’s focus on their local neighborhoods does not
interfere with their exploration of broader interests and new social interactions, how will
the two trends complement and supplement each other? It is not the technologies that
result in behavioral change, but rather the structured access and perceptions of resources
and relationships that determine the outcomes of a new technology. Rather than testing its
effect, the study tries to understand the process of using a new technology. The socio-
technical process of using a technology – or the use practice and perceptions of a
particular feature – is the key determinant of its meaning (Emery & Trist, 1960). This is
the backdrop for many assumptions in this study, which helps it avoid technological
determinism (Bijker, 1995).
At the time of writing, geocoding is still a feature that developers, consumers and
businesses find controversial. Geocoding refers to the capability for applications and
websites to detect the user’s physical location, and use this information to customize and
filter relevant content (Goldberg, 2008). The detection could be either implicit or explicit,
and the customization may either confine or broaden the scope of communication and
information search. Therefore controversy sometimes arises, for reasons from privacy to
relevance and usefulness (Fiveash, 2011; Mitch, 2008; Zickuhr & Smith, 2011). The
10
uncertainties about this new technology make it possible to examine the social
construction process in which people discover the meaning of the technology and adjust
their use practices accordingly (Bijker & Pinch, 1984). A nuanced understanding of how
people use and perceive location awareness will then help us predict what types of social
relationships or activities will more likely be affected by geolocation technologies. With
the knowledge of social affordances and constraints in location-aware communities, it
will be easier to make predictions and observations on the impacts of new technologies or
simply new features that are pushed to us regularly. Rather than answer the question of
what new technologies do, this study aims to uncover what people do with new
technologies. Rather than framing location awareness as having disruptive effects, this
study tries to understand how this technology is appropriated in existing features and
practices.
Because geographic distance is related to the perception of social proximity, the
awareness of the geographic location of an information source makes people more reliant
on the validity and effectiveness of other community features like search, tagging or
recommendation systems. To put this logic simply, if a user called @yardsale sends an
alert about a yard sale in my local neighborhood, one will have a stronger urge to check if
@yardsale is reputable and shares similar interests with me. This is because a user who
holds a yard sale within one’s neighborhood matters more than another user who does it
5,000 miles away. With this sense of immediacy and relevance, one can determine if her
information is trustworthy and this yard sale is worth visiting. In other words, for an
online community that emphasizes mutual accountability and social identification,
11
geocoding features will have a reinforcing effect on the relationships between the
necessary community features and these social perceptions. By using geocoding features,
users may better configure communication to interact with friends and new ties that are
physically closer. As a sense of proximity can increase communication, propinquity may
constitute or substitute a sense of similarity in the early phase of community formation
(Barnlund & Harland, 1963). Yet more essential social cognitive factors identification
and interpersonal bonds may emerge over time in people’s use of other profiling and
communication features. Geocoding, as a feature that configures the basic sense of
geographic proximity, may improve the accuracy, relevance and effectiveness of these
features, and influence communication in an indirect way. The moderation effect of new
features like geocoding on pertinent socio-technical processes is the way this study
interprets the general observation that people “negotiate” the use of new technology
within the context of their existing practices (Bijker, Hughes, & Pinch, 1987).
Specifically, this study tries to understand what are some of the community
building features and processes, and how the use of geolocation awareness influences
these processes. The theoretical effort is to integrate Social Impact Theory with classic
CMC theories about community attachment (Prentice, Miller, & Lightdale, 1994), social
identities (SIDE) (Postmes, 2006) and interpersonal relationships (SIP) (Walther & Parks,
2002). Psychologists have proposed and empirically supported that physical propinquity
has a regular and consistent effect on mutual influences and friendship formation in
social interactions (Latane, Liu, Nowak, Bonevento, & Zheng, 1995; Barnlund &
Harland, 1963). People engage in more active interaction with social ties that are
12
geographically closer, because proximity increases the sense of immediacy and hence the
mutual influences between actors (Latane, et al. 1995; Takhteyev, Gruzd, & Wellman,
2012). The influence of propinquity may be moderated by other attributes of people. For
example, earlier research indicates that other attributes such as status and previous
familiarity may reduce the influence of geographic proximity (Barnlund & Fraland, 1963;
Monge & Kirste, 1980). What this line of research leaves unclear is how proximity is
integrated into other communication processes to influence the formation of new ties at a
community level. This problem becomes more obvious when the Internet redefines the
experience of proximity and makes this experience an integral part of community
experience (Latane & Liu, 1996). Technologies change the condition for people to
connect and communicate with one another across geographic boundaries. Therefore the
perception of social proximity will likely involve more dynamic communication
processes, which can be influenced by geographic distances, topics of shared interests,
network boundaries and many other processes. Social proximity will then have a more
complicated influence on the general development of affinity and identities in online
communities. It is necessary to understand the meaning of geographic propinquity as a
context for communication online.
Location awareness is the systematic and effective support for people’s holistic
social experience of places, resources and activities. It provides the necessary information
for people to determine their proximity to information, events and other resources. Yet
for location awareness to support meaningful social experience, there needs to be other
tools that configure social interactions. As Social Impact Theory suggests, people
13
evaluate mutual influences in their social interactions based on a combination of the
number, characteristics and immediacy of other social actors (Latane, 1981). The
geographic proximity of potential social ties is only one factor that involves immediacy.
It needs to be considered along with the number and relevant attributes of such ties in
specific contexts. Because people engage with and affect many different others in a
community, geolocation technology needs to be examined together with other community
features like messaging, reputation systems or social matching systems. The reason for
this integrated examination is that these community features support the semantic
knowledge of locations as well as the activities, people, and information about various
locations. The integration of location awareness and these features determines the new
scope of affordances and constraints for people’s social practices and experience in
emergent online communities. As a result, location becomes an additional layer of
contextual data that supports the functioning of a community. In other words, the sense of
geographic proximity becomes a psychological assessment of the necessity and relevance
of engagement within a social group (Latane & Liu, 1996). For example, people living
near downtown Los Angeles can share events, trends and stories about their local
neighborhoods simply by creating, replying to and commenting on different posts that are
tagged as within five miles of central Los Angeles. Furthermore, filtering features help
people securely and efficiently share such information with different kinds of users such
as friends and acquaintances. People may even use these features to interact with
strangers who happen to share common interests or objectives in knowing more about
this area.
14
By integrating with community features that support people’s social experience of
places, events and relationships, therefore, location awareness creates a new space for
shared activities and social interactions. Geographic proximity is a sufficient but not
necessary condition for people to assess the possibility and strategy for developing
relationships with others and grow attached to a community. Location information
provides a basic sense of accuracy and relevance, and proximity serves as an initial basis
for the feeling of proximity and similarity, when no other shared attributes are available
(Barnlund & Harland, 1963). Yet it is in the integration of location awareness with other
features of online communities that the broad social meaning of geocoding technologies
is defined. If an online community effectively implements these functionalities, people
are more likely to participate and contribute in this constructive social environment. In
short, community development is simply facilitated by easy and flexible configuration of
proximity, because social interaction, shared focus and coordination can be customized to
fit the community members’ needs. The theoretical question is the extent and scope for
which specific aspects of social interaction is influenced by the configuration of
geographic propinquity.
On the other hand, communication theories about virtual communities must
incorporate the influence of geolcoation on the development of affinity and identities
online. Most CMC theories were formulated at a time when people had to communicate
in very constrained media environments. So many people may not have known how close
they are to one another. The challenge now is that these conditions are no longer
prevalent or even possible. Communication media like Twitter enable almost
15
synchronous conversations, social networking sites such as Facebook encourage
interactions based on real identities, and social media sites like Flickr and YouTube are
built on the premise that sharing rich media is a natural part of online interaction. More
importantly, the division between online and offline interaction seems to be less relevant
when people stay connected with all kinds of computing devices and services. The
physical and the virtual become increasingly integrated. New technologies such as
geocoding make the physical setting even a more salient context for online interaction.
With increasingly rich communication of contextual and nonverbal cues, people’s online
practices will be different from before. Theories need to account for that change. How
will the experience of geographic proximity be absorbed in people’s communication in
online communities, and what are the exact technical features that facilitate or constrain
these activities? Will the sense of proximity influence information sharing and social
relationships, and what is the social process that increases or decreases affinity in
communities?
To answer these questions, it is necessary to examine the features and social
processes in communities that emerge when people share information of common
interest. These processes will in turn influence usage. Features are important technical
artifacts that configure the social architecture of online communities (Orlikowski &
Iacono, 2001). Social processes reflect the adaptive behaviors and perceptions in the use
of these features within particular social structures such as groups and communities
(DeSanctis & Poole, 1994). When people share information and communicate with others
online, communities may emerge as they gradually grow a sense of accountability,
16
interconnection and identification (Tapscott & Williams, 2008). This is a process in
which interpersonal affinity and shared identities gradually develop over long-term,
repeated communication. Social Impact Theory and earlier perspectives on propinquity
alone do not explain this process, because geographic proximity is just one factor that
influences the strength of mutual social impact (Latane, 1981; Latane & Liu, 1996;
Barnlund & Harland, 1963). Classic CMC theories like SIDE or SIP will not be adequate,
as they often lack the focus on specific features that configure the community processes.
This negligence is important, as the design and architecture of features inevitably
influences behavior (Lessig, 1999).
It is thus essential to consider what features contribute to these community-
building processes in informational exchanges. What are the social functions of features
that contribute to what Social Exchange theory would describe as interdependence in
information exchanges (Ekeh, 1974)? How do users develop community attachment
based on interpersonal affinity or collective identities (Prentice, et al., 1994; Spears, Lea,
& Postmes, 2006)? What are the technical and structural conditions for social capital to
accumulate in such information exchanges (Burt, 2000; Resnick, 2001)? And finally, how
does proximity influence communication processes and configure dynamic social
influences at a community level (Latane & Liu, 1996)? These are some of the key
theoretical questions that this study will try to answer.

17
Chapter Two: Reputation Systems, Generalized Reciprocity and Location
Awareness
Introduction
The main goal of this study is to theorize the use of features in online
communities when there is a stronger focus on geographic proximity in the sharing of
personal information. Specifically, it examines the social cognitive processes that are
associated with the use of several common features in online social networks. It then
examines how the use of new technologies like geolocation influences these processes.
Such processes can be embodied by the perceptions of attachment to communities,
generalized reciprocity, and social capital in a community. These perceptions indicate the
perceived salience of constructive social processes have been shown to promote
participation and contribution in online communities or organizations (e.g. Bagozzi &
Dholakia, 2006; Kankanhalli, Tan, & Wei, 2005; Wasko & Faraj, 2005). However, it is
not clear what specific technical artifacts support or hinder these processes. As a result, it
is often difficult to apply the idea that online behavior is the consequence of social
architecture of code and design (Lessig, 1999).
Drawing from theories of collective action, this study will specify the relationships
between the uses of particular features and the socio-psychological processes that
facilitate effective participation in online communities. Rather than providing a
comprehensive account of how communities work, the study focuses on the use and
outcomes of several features essential to many online social networks. Both the main
effect of feature use on social processes and the moderation effect of geotagging are
given sufficient weight and consideration. The rationale is that to understand how a new
18
technology like geocoding is appropriated into existing practices, it is first necessary to
analyze the relationships between some more mature and prevalent features and their
intended consequences (Rogoff, 1995). The first such relationship is between reputation
systems and reciprocity.
Reciprocity is an important concept in all communities. According to Social
Exchange Theory, it indicates the extent to which individuals are mutually accountable
for their actions (Ekeh, 1974; Shumaker & Brownell, 1984). In a reciprocal community,
contribution is rewarded with recognition and appreciation, and the retrieval of resources
is justified with equitable contribution. An early example of reciprocity in online
communities is the use of feedback systems in eBay (Resnick & Zeckhauser, 2002).
Numerous studies have documented that when individuals systematically exchange favor
or mistreatment with feedback and when this feedback is systematically recorded and
reflected in users’ reputation profiles, a strong sense of generalized reciprocity is created
(Hou, 2007; Resnick, Zeckhauser, Swanson, & Lockwood, 2006). Generalized
reciprocity refers to the degree to which each member of a community is accountable and
recognized for their activities at the collective level (Bolton, Katok, & Ockefels, 2004;
Heinz & Rice, 2009). In comparison, bilateral reciprocity specifies the extent to which
resources and activities are exchanged in both ways within pairs of people. When an
online community systematically provides feedback or recognition for contribution,
members tend to have a stronger motivation to participate and contribute high-quality
knowledge (Moon & Sproull, 2008). The underlying mechanism is that social loafing or
free-riding becomes difficult when individual actions are visible and identifiable (Harkins
19
& Szymanski, 1989; Karau & Williams, 1993). Generalized reciprocity therefore
becomes an important social process that makes communities like eBay a secure
environment for transactions and interactions.
To create and sustain generalized reciprocity, there needs to be a commonly
accessible archive of social interactions. An official acknowledgement of one’s
contributions accompanied by material rewards makes good behavior a more desirable
option. If one does not behave in compliance with the community norms, this “bad”
behavior is penalized with a cumulative bad reputation in the community, which alerts
others about the potential risks of interacting with them (Ekeh, 1974). The use of this
reputation archive and the belief in this equitable distribution of recognition and rewards
need to be shared on a community level. This way, every interaction or transaction is
guaranteed with the same degree of accountability (Shumaker & Brownell, 1984).
Generalized reciprocity thus exists in communities where everyone is rewarded or
penalized for his or her actions. In such communities, mutual obligation and respect is
systematically promoted and expected in everyday interactions. Especially in online
communities where there is a lack of shared history or a lack of expectation for future
interaction, generalized reciprocity has proven to influence people’s perception of social
accountability for actions (Wasko & Faraj, 2005; Wasko & Teigland, 2002).
The basis of generalized reciprocity is an effective profile system that verifies the
identities of individuals and archives their actions. A persistent identity is necessary for
anyone’s activities to become public record on a community level (Ma & Agarwal,
2007). In online communities, a profile system can include users’ relevant demographic
20
attributes for basic individual identification. Increasingly, many communities implement
features that aggregate and display users’ activity histories in their profiles. For example,
communities like eBay or Yelp display the history of a member’s main activities such as
selling things or writing business reviews as a prominent part of their profiles. Since
these communities rely on effective transaction or relevant information contribution from
users, such activity streams become the objective and systematic records for the quantity
and quality of a member’s contribution to the public good in the communities.
It is worthwhile to explicitly note that reputation features, as with many other
features, have a powerful influence on behaviors and perceptions in any online
communities. This is because features are designed and programmed to implement
community operators’ assumptions, intentions and norms. Simply put, code is the law
that governs behaviors, and features are the coded tools for executing the laws (Lessig,
1999). For example, if Yelp intentionally displays every member’s past reviewing
activities, it is because the Yelp developers instill the law – or assumption – that such
numbers indicate reviewers’ past contributions and help other users better evaluate such
contributions (Pattison, 2010). The exact meaning of a feature may vary across use
contexts, but it is the interaction between the intention of a technology and users’
appropriation that ultimately determines the social process of use in any social structures
(DeSanctis & Poole, 1994). As reputation systems simplify and improve the verification
of others’ histories online, user appropriation may not be necessary or relevant due to the
extra cognitive labor. From this perspective, the implementation of reputation features
21
should have a considerable influence on the social process in which community members
evaluate others’ contribution and justify their own activities.
The positive effect of the quantitative measures of a community member’s past
and current activities on generalized reciprocity has been documented in e-commerce
communities like eBay (e.g. Bolton, et al., 2004). In information-sharing communities
such as Quora or more general-purpose social networks like Twitter, users’ activities are
also evaluated with quantitative measures such as reputation scores or influence indices
(Morris, Counts, Roseway, Hoff, & Schwarz, 2012). Some rudimentary measures of
users’ social networks, such as the number of friends or followers, have shown to
influence people’s general impression of others (Tong, Van der Heide, Langwell, &
Walther, 2008; Walther, Van der Heide, Hamel, & Shulman, 2009). The activity metrics
serve as a community-wide voucher for experience, popularity or influence. It becomes a
basic recommendation for other users to access and adopt information. In other words,
quantified personal history implemented by many online communities can be turned into
reputation metrics that facilitate identification and verification.
Because everyone is identifiable and verifiable within a specific community, there
will be an increased sense that one’s contribution or use of collective resource pool is
accountable. This perception may result in a stronger sense of generalized reciprocity and
obligation (Ma & Agarwal, 2007). The expectation of reciprocation and commitment to it
constitute the relational mechanism that explains the influence of profile systems on the
overall salience of trust and security in online communities. Even though the profile may
not be tied to an offline identity, the record of past activities can still facilitate this
22
community process. This is possible when the record is associated with a persistent
identity in the community context, and when the recorded activities are relevant and
specific to the purpose of the community (Bolton, et al., 2004; Ostrom, 1990). For
example, an eBay seller does not have to fully disclose her real life name and address, yet
her persistent reputation information still helps buyers evaluate her trustworthiness. A
programming expert can use any pseudonym to answer questions on the coder
community StackOverflow. His contribution is still verifiable as long as he maintains a
persistent username. Real world identification will facilitate the verification of persistent
identities, but is not always necessary for the verification of community-specific activity
histories.
The concept of reputation systems and why it is related to generalized reciprocity
in online communities will be discussed in detail next.
Reputation Systems
Reputation is “what is generally said or believed about a person’s character or
standing” (Simpson & Weiner, 1989, p. 1227). It is an essential piece of information
about the benevolence or credibility of an individual (Whitmeyer, 2000). A reputation
system in online communities is the collection of features that allow users to post,
aggregate and view feedback for transactional activities such as purchases (Resnick,
Zeckhauser, Friedman, & Kuwabara, 2000). For a reputation system to be effective, it
needs to help users “distinguish quality attributes among potential transaction partners”
(Zhou, Dresner, & Windle, 2008, p. 789). In other words, a reputation system provides
information about the attributes of actors that are relevant to the context of practices in a
23
community. For a community like eBay, such information may indicate the capacity of
buyers or sellers for completing commercial transactions and honoring agreements
(Resnick, et al., 2006). For communities that are based on sharing information of
common interests, in comparison, reputation systems can often indicate the quality and
quantity of individuals’ contribution of useful knowledge or feedback (Moon & Sproull,
2008).
Several studies have examined the relationship between online reputation systems
and transaction outcomes in the e-commerce context (See Zhou, et al., 2008 for a
review). These studies generally show that there are two conditions for reputation
systems to work. First, ranking mechanisms need to be effectively implemented. They
should accurately reflect individuals’ past behaviors, facilitate easy feedback (N. Miller,
Resnick, & Zeckhauser, 2005) and minimize intentional gaming of the ranking (Fan, Tan,
& Whinston, 2005). Second, identity verification mechanisms must exist to facilitate the
accumulation of meaningful knowledge of transaction partners sources based on relevant
demographic and contextual attributes (Forman, Ghose, & Wiesenfeld, 2008).
Built on these mechanisms, effective reputation systems bring such benefits as
greater transaction security for buyers (Zhang, Ghorbani, & Cohen, 2007), higher selling
prices for sellers (Ba & Pavlou, 2002), higher trust of the community as a secure and
friendly place (Fuller, Serva, & Benamati, 2007), and more effective sharing of
information (Ardichvili, Page, & Wentling, 2003). Researchers argue that buyers rely on
positive reputation feedback because such information certifies both the benevolence and
credibility of sellers (Ba & Pavlou, 2002). Trust is established and strengthened among
24
transaction partners. As individuals gradually reinforce their belief in reputation
information as a systematic, verifiable record of behavior during repeated interactions
(Bolton, et al., 2004). The key to the verifiability of such record lies in systematic
solicitation, aggregation and display of relevant reputation information for persistent
identities. Exactly because identities are persistent in given communities, the “shadow of
the future” makes it more likely for people to uphold a minimal level of accountability
and trust in their interactions (Axelrod, 1985).
The gist of this line of research is that the combination of accurate, objective
ranking information with personal verification data provides users with the necessary
tools to build trust and engage in social interactions securely online (Resnick, et al.,
2000). What remains unclear is the common socio-cognitive process in the use of
reputation systems. Put in a more broad way, this process should be prevalent in in
communities that may have different social architectures, shared purposes, and
consequently different interpretation of reputation. When an online community is based
on sharing information about local restaurants, for example, the creation, meaning and
use of reputation information will be different from a commerce-oriented community. In
the former, reputation rankings can be based on different types of metrics such as the
number of friends, the number of reviews and the badges that indicate recognition for
review quality from the community operator. Furthermore, there are also many different
ways for users of different communities to provide feedback and evaluations of others’
reputation. Rather than actively providing feedback on transactions, people can provide
their evaluation or appreciation of others in a more casual and less committed way.
25
Adding one as a friend or following one on Twitter may be a token of appreciation, a
token of recognition for the latter’s past activities (e.g. constantly posting funny jokes), or
a symbol of current relational status (e.g. being a follower of one’s follower). Moreover,
commenting or forwarding one’s Twitter post shows a certain degree of agreement to or
alignment with its statement, or at the very least, endorsement of its relevance (Morris, et
al., 2012). The aggregation of such relational and activity statistics forms the basis of a
reputation system that reflects the collective feedback on one’s standing in an online
community. A useful analogy is Google’s PageRank algorithm, which denotes the
relevance or “reputation” of web pages based on their embedded incoming and outgoing
links (Jøsang, Ismail, & Boyd, 2007). By acting on others’ content and interacting with
them, we leave traces of statistics that reflect our assessment or perception of them. This
implicit feedback becomes useful as reputation information once it is aggregated and
presented in a meaningful way. Yet it is our common understanding and endorsement of
such feedback that gives it meaning.
A common mechanism is that reputation becomes a type of profiling information
that fosters verifiable identity and accountability in communities. The effectiveness of a
reputation system depends on the degree to which good behavior is observed, honored
and reciprocated (Resnick & Zeckhauser, 2002). In other words, there must be
community-level trust, reliance and use of the system that collects attributes and metrics
in order for users to judge the relevance and value of others’ standings. The community
needs to develop and sustain a systematic belief that such reputation systems are effective
and trustworthy guidelines for interactions. It is the repeated use and systematic display
26
of personal reputation metrics and a reliance on it that leads to the reinforced belief about
generalized reciprocity in the community.
The Use of Reputation Systems
Reputation and other pieces of personal information in a user profile are heuristic
cues for knowing and assessing others in online communities. Heuristic cues refer to “any
variable whose judgmental impact is hypothesized to be mediated by a simple decision
rule” (Eagly & Chaiken, 1993, p. 327). Some researchers have begun to employ a social
cognitive perspective to explain the effect of information technologies on
communication, technology acceptance and information adoption (c.f. Ferran & Watts,
2008; Metzger, Flanagin, & Medders, 2010; Robert, Dennis, & Ahuja, 2008). Faced with
information overload on the Internet, individuals have to verify the true identities of
information sources in order to determine their credibility and trustworthiness (Ma &
Agarwal, 2007). Their adoption of information can be influenced by heuristic cues such
as geographic proximity rather than by the inherent quality of the information content
(Forman, et al., 2008).
The explanation for these phenomena draws from the Heuristic-Systematic Model
(HSM) (Chaiken, 1980; Chaiken, Liberman, & Eagly, 1989). Simply put, a
comprehensive approach that relies on systematic scrutiny of arguments and evidence can
consume and demand more cognitive resources. A limited approach that relies on
heuristic cues – simple decision rules or schema based on relevant facts about the
information source such as demographics, experience or status – can be cognitively less
demanding. The choice of the exact processing mode depends on the availability,
27
accessibility and perceived reliability of judgmental cues. Because humans always seek
solutions that require the least effort, heuristic processing is often preferred, especially
when information validation itself is cognitively demanding (Eagly & Chaiken, 1993, pp.
326-327). In other words, people often voluntarily rely on relevant attributes when they
assess information from others.
Unlike with traditional media, the authenticity and usefulness of information
cannot be readily judged in an online community. It requires extra effort and time for
individuals to develop trust of information sources online (Briggs & Smyth, 2006). As
discussed above, reputation is one special type of decision cue about information sources
that indicates benevolence or credibility of people in a specific context of information
exchange. In a community in which the main purpose is the sharing of information of
relevant interests, judgment of content quality can be a daunting task due to the sheer
amount of content. Information adoption is therefore largely determined by the attributes
of content contributors. The capacity for making good contributions can be verified by
objective certifications. A cumulative numerical score of actors’ past behavior is shown
to influence other community members’ willingness to conduct transactions with them
(Resnick, et al., 2006). For this belief in quantified reputation to apply a broader range of
communities, there must be decision rules that guide the collective perception and peer
evaluation of people’s standing in the community. In this scenario, the decision rule is
that the quantified history of people’s past contribution can systematically validate their
reputation.
28
So how can this decision rule be systematically reinforced? Studies of online
communities have suggested that people tend to base their impression of others on
“warranting” information, or information about someone that is immune to manipulation
(Walther, et al., 2009). Such information can come from an objective source such as a
third-party website, or from the collective perception of others (Walther & Parks, 2002).
The feedback or the implicit rankings of the contribution of an information source can
become a reputation cue that has high “warrant”. For example, a high number of
comments, shares, or “likes” for a person’s post on Facebook indicates peer interest and
appreciation of the relevance of this person’s post, which is difficult to manipulate by the
creator of the post. Just as such rankings collectively determine the quality of a news
entry on social news sites like Digg, the standing or character of a person in a community
can be indicated by peers’ implicit acknowledgement of interest in and appreciation of
this person’s activities. Because such feedback or rankings can be automated and
verified, individuals do not need to make extra efforts to validate them. As a result,
people can evaluate others in online communities based on the systematic metrics that
objectively reflect the communities’ collective assessment of relevant attributes.
The use of reputation metrics thus depends on the trust in the relevance and
accountability of community-level assessment. People must trust in the reputation
systems in a community, believing that the systems will not be gamed, and that feedbacks
and rankings reflect true perceptions and promote future benevolence (Zhou, et al., 2008).
Effective aggregation and presentation of relevant reputation information enables users to
verify their own and others’ identities and activities, facilitates self disclosure and
29
meaningful communication, and reinforces strong community norms of disclosure,
transparency and trust (Forman, et al., 2008; Ma & Agarwal, 2007). By integrating
relevant reputation information, a community-wide profile system makes individuals’
actions identifiable, verifiable, and traceable, thus encouraging contribution and
penalizing free riding. As a result, individuals not only make decisions based on their
direct experience with others, but also derive confidence in the wisdom of the crowd –
the generalized trustworthiness of the reputation heuristics that are generated and verified
by the community (Surowiecki, 2005). The use of a well-implemented reputation system
that collectively verifies relevant reputation information will thus contribute to the
construction of generalized reciprocity in an online community. Based on this logic, the
following hypothesis is proposed.
Hypothesis 1: Community members’ use of reputation systems that implement
objective and relevant measures of users’ identifiable and verifiable activity histories will
have a positive influence on their perception of generalized reciprocity in the community.
Location Awareness and Reputation Systems
When a geocoding feature is used in an online community, users’ current
geographic locations can be automatically captured and displayed in their profiles.
Furthermore, this geocoding feature can be used to filter and sort others’ postings based
on users’ locations. This is how geolocations can be turned into contextual tags that
indicate the geographic proximity for communication and social interactions. For
example, on Twitter, Foursquare or Facebook, it is possible to tell not just where a user is
based, but also where a message is sent or an activity takes place. Every time a person
30
makes a post or sends a message, the geographic coordinates are automatically captured
and recorded in these communities. The dynamic disclosure of a member’s location may
serve as a common basis of geographic homophily. After all, geographic locations like
birthplaces, schools and workplaces still form the basis for many ties that emerge and
evolve in our social life. There are some well-documented effects of geographic
similarity on identity verification and information evaluation in online communities
(Guest & Wierzbicki, 1999; Hampton, 2007). For example, people tend to trust product
reviews more if the reviewers are from the same geographic region (Forman, et al.,
2008). Twitter users also tend to communicate more with others who are geographically
closer (Takhteyev, et al., 2012).
Location disclosure is particularly important for online communities that rely on
the exchange of geographically bounded information or local services like Yelp or
CitySearch. The disclosure of locations in members’ profiles makes it efficient for
individuals to quickly filter and verify information in such communities. Location
information does not make it easier or more necessary to tell a member who she is in real
life. But it does make it easier to distinguish authentic or relevant information based on
where she is. For example, a location tag for a restaurant reviewer called Andrew K. on
Yelp does not make one more interested or capable of verifying if this reviewer indeed
has a first name of Andrew. Nevertheless, this location tag will help one determine if his
reviews about restaurants within the same neighborhood are genuine or relevant.
However, it would be unreasonable to assume that location awareness alone will
simply make people more identifiable, verifiable or more trustworthy in online
31
communities. On one hand, as new features are integrated into people’s existing use
practices, it is reasonable to assume that location information is also integrated into a
holistic set of personal metadata that facilitate mutual verification and evaluation
(Forman, et al., 2008). On the other hand, geographic proximity does not necessarily
create affinity, as the latter depends on many other communication processes that shape
people’s perception and interpretation of others (Latane & Liu, 1996). Proximity must be
integrated with the norms that govern the perception of other people’s quality and
credibility in the community. For example, if the norm is to trust people who share
similar interests in a community, then the same preference for similarity will be applied
to the processing of location demographics. Existing purposes, preferences, and social
architectures in a community determine how useful or valuable location features should
be. Knowing where a friend is or where a friend is doing something may not matter to a
community that is focused on organizing flash mobs or sharing merchant deals. So the
central issue of location awareness in online communities is how location information
integrates with people’s social activities and how it influences the process in which
people’s social and demographic attributes are perceived and interpreted.
Putting location in a person’s user profile is nothing new for online communities
(Drori, 2002). What is new with geocoding technologies is the tagging of essential
activities such as posting, commenting and messaging with geolocation data, and the
effect of “geotagging” on the verification of users’ reputation information. What is new
with geocoding technologies is the tagging of essential activities such as posting,
commenting and messaging with geolocation data, and the effect of “geotagging” on the
32
verification of users’ reputation information. When any piece of information being shared
is tagged with geographic coordinates, this additional contextual information may make
information query and retrieval more efficient and relevant. Furthermore, this contextual
awareness may also increase people’ sense of mutual interdependence and initiate new
relationships based on the shared sense of social presence (Jones, et al., 2007; Schmandt
& Marmasse, 2004).
This suggests that geotagging may serve as a community-wide tool for
systematically ascertaining the metadata of other members and reaffirming the
community norms that regulate interpersonal communication. In other words, geolocation
information provides a relevance or immediacy check for people’s profiles. For example,
the geotags for a person’s activities may confirm or counter what this person claims to
have done. As a result, geolocation awareness provides an extra layer of verification
mechanisms: it integrates and interacts with existing verification systems that aim for
promoting mutual trust and accountability. When a person has a high reputation and lives
nearby, it makes sense to give her opinions some special weight. First, her opinions might
contain relevant information to one’s own neighborhood. Second, there is a higher
possibility for future interactions. It is possible for a person to fake his or her current
locations with so-called “location spoofing” technologies. Yet location spoofing may be
less of a threat if the main purpose of geotagging is to facilitate the verification process,
rather than substitute it. For example, just showing up at a place many times does not
necessarily make on an expert in the information about that place. The expertise still
needs to be established on meaningful contribution of relevant information, and effective
33
communication among the community members. The faked frequency of activities in a
place, in other words, will not be enough to establish reputation in a community that
prizes quality content and meaningful communication. Therefore location awareness
affects the extent to which reputation systems are perceived and used as a reliable way to
assess the risks and benefits for sharing information with others. From the perspective of
Social Impact Theory, geographic proximity simply reinforces the social influence of
individuals by verifying and validating their relevant characteristics (Latane, 1981). In
summary, geotagging makes reputation systems more effective in verifying identities.
The stronger verification mechanisms simply reinforce the community-level sense of
reciprocity.
Location Awareness and Generalized Reciprocity
It is also reasonable to assume that in a location-aware online community,
individuals need to rely more on effective reputation and feedback systems to interact
with others. Location is a sensitive piece of private information that may compromise
personal safety and privacy if used unnecessarily and inappropriately (Rawassizadeh,
2011). For individuals to willingly disclose location data in information sharing, an
online community must demonstrate the need for the value of integrating such sensitive
data in its everyday interactions (Raento & Oulasvirta, 2008). The benefits might be the
retrieval of useful and relevant information, and the filtering of users for more targeted
social interaction. For example, Yelp would not be able to provide accurate restaurant
guides without asking users to report their locations. Zaarly, a location-base community
for hiring or offering errands, would only show the posted content by first detecting and
34
acquiring users’ current locations. The risks of disclosing location information must be
effectively mitigated by a reputation system that objectively records the community-wide
assessment of other members’ benevolence or trustworthiness.
For a community to be useful, however, people need a trustworthy, reciprocal,
cohesive and inviting social environment. Generalized reciprocity is a social process that
facilitates participation. It can be promoted by a well-designed profile system in which
reputation is a key metric. This reputation metric indicates implicit or explicit
endorsement of other users’ relevant activities and attributes, as illustrated by the like and
vote functions in communities like Facebook and StackOverflow. When geotags are
used, the disclosure of personal location information may increase the sense of proximity
and hence the potential mutual attraction, liking and bonds. Yet the voluntary disclosure
of geolocation information also increases the need for making sure people behave well
and respect one another in the community. There needs to be an equivalent belief in such
systems being fair, valid, objective and immune from manipulation. This belief places a
higher expectation on the community as a social environment that honors the equitable
exchange of information and fosters reciprocity. Therefore, for geographic proximity to
become part of the community experience, individuals must ensure that the disclosure of
geographic information is equitable, secure and supplementary to the purpose of the
community. This requires the implementation of an effective reputation system that
reinforces mutual accountability and generalized reciprocity. This enhanced
accountability, reciprocity and security forms the basis of a social environment that
makes participation rewarding and meaningful.
35
Thus, location awareness should have a reinforcing effect on the relationship
between the perceived use of reputation systems and the generalized reciprocity in an
online community. This makes sense especially when geographic proximity is relevant to
the focus of a community. For example, when a reviewer is shown to frequently visit a
neighborhood where she has posted many business reviews, other users will consider her
opinions more credible, and place higher trust on the reputation metrics in the reviewer
community. A community based on trading used textbooks will need to ensure that
everyone trades with only people nearby to reduce scam and fraud. As location awareness
changes the credibility of reputation systems, geotags change the relevance and
immediacy of people’ reputation status. As a result, perceived generalized reciprocity
will be different in the community when geographic proximity can be automatically
evaluated and configured. Useful contributions relevant to the focal locations and
common interests will always carry the expectation of being recognized and rewarded.
Effective access to relevant information is exchanged with relevant contributions that are
recorded and visible to the whole community. It is therefore hypothesized that,
Hypothesis 2: The influence of community members’ use of reputation systems on
their perceived generalized reciprocity will be stronger when geolocation features are
used.
The hypotheses are illustrated in the diagram below.
36

Figure 1: Reputation Systems, Reciprocity and Location Awareness
37
Chapter Three: Social Matching Systems, Community Attachment and Location
Awareness
Introduction
Previous studies stressed that effective collaboration requires the pairing of
interests, needs and goals based on a mutual understanding of common ground and
shared contexts (Brandon & Hollingshead, 2004; Tiwana & McLean, 2005). The key to
achieving a shared appreciation of context is technology that supports individuals’ efforts
to access multiple perspectives and identify common assumptions (Boland, Tenkasi, &
Te'eni, 1994). For the past decade, scholars and observers have consistently shown that
the Internet facilitates the organization of online communities based on shared interests,
needs and goals (cf. Castells, 2003; Shirky, 2008; Wellman & Gulia, 1999; Wellman, et
al., 2003). Applying this perspective in online communities, it is clear that online
communities must facilitate the discovery and sustaining of interpersonal compatibility
and community-level commonality. However, few studies of online communities employ
a similar approach to discuss specific features that facilitate this process in user
interactions and practices at the community level. Classic CMC theories rarely consider
specific features that gradually change how people explore common interests and engage
in conversations. This negligence makes it difficult to extend the idea that code
configures behavior in the online environment in new contexts of use (Lessig, 1999).
This chapter discusses the association between features such as tags and the social
matching processes in online communities.
Under the label of social matching systems, features like recommendations and
suggestions facilitate the social matching and pairing process. In an extensive discussion
38
of the design of social matching systems, Terveen and McDonald (2005) argue that a
community system must accommodate meaningful social interaction between users who
have matching attributes and objectives. In addition to pairing users based on static
attributes such as age and gender, a social matching system must understand users’
preferences and goals in order to facilitate their exploration of new relationships and
information. It must also provide sufficient support for emergent communication and
actions. In other words, an effective social matching system should not only identify and
discover content and connections that match users’ interests or attributes, but also
facilitate community actions that achieve aligned goals and fulfill broader interests.
The social Q & A website, Quora, can illustrate how such a system works. Users
receive constant recommendations on both who to follow based on shared interests, and
what topics to follow based on users’ past histories (Rivilin, 2011). Therefore, users not
only can identify the content and people that match their interests. They can also take
concrete actions such as starting private conversations, building relationships and
contributing new content. Similar features, or rather the pairing, matching and
communication processes facilitated by these features, prove to have turned Twitter from
a group messaging application into a social network that facilitates the sharing of
information and the construction of groups based on many different topics (Levy, 2009).
In such social matching processes, a community can emerge from the development of
interpersonal ties based on bilateral connections and the reinforcement of shared purposes
through conversations and common activities. Social matching systems are the key to
initiating such processes.
39
Furthermore, social recommendation and suggestion features make opportunistic
matching possible. Opportunistic matching means that initially unconnected individuals
can be joined into ad hoc actions based on shared interests. When social interaction spans
across structural clusters that are based on different traits or backgrounds such as
workplaces, gender or political affiliation, social cohesion and identification will be
stronger in the community (Burt, 2007). This is because matching people based on many
more kinds of traits increases the overall density of connections in the community
network, and provides opportunities for ad hoc collective actions (Burt, 2002). For
example, retweets that contain hashtags on Twitter help some users identify topics of
shared purposes beyond their immediate social circles – their followers and the people
they are following. These users then can join more types of ad hoc conversations and
advance a common agenda with a much broader set of ties. The snowballing growth of
participants in these conversations engenders new relationships, and helps the users
identify otherwise hidden or under-explored ties.
By using social matching systems, an online community can thus facilitate the
construction of both interpersonal ties and shared identities. This constructive process
then increases its members’ overall attachment to the community and reliance on it as a
source of information and social support (Shirky, 2011). In other words, social matching
systems enable online communities to develop on both dyadic or group-level
interpersonal affinity and community-level identification of shared purposes and
objectives (Ren, et al., 2007). When geotags are used in a community, the accurate
perception of proximity will help users more effectively configure their social
40
experiences with contacts or topics that are recommended by the matching systems. This
is because proximity has a strong influence on people’s perception of others’ relevance
and closeness, and on their perception of belongingness to given social groups (Latane &
Liu, 1996). This chapter explains the implications of social matching features and geotags
for online communities in detail.
New Media and the Development of Online Communities
At the early phase of the Internet as a mass communication medium, anonymity
was prevalent, and communication was hampered by the lack of audio and visual cues
(Baltes, Dickson, Sherman, Bauer, & LaGanke, 2001; Kollock, 1999b). The Internet
appeared to be a deficient environment for the creation and sustaining of meaningful
interpersonal ties (Nie, et al., 2002). Yet many online communities arose, in which people
complied with emerging social norms and built meaningful and rich relationships over
time (Katz & Rice, 2002). One line of thought (Social Information Processing, or SIP)
suggested that online social interaction could accommodate meaningful relationships,
because people strategically use stylized nonverbal cues and present themselves in
creative ways. Although initially communication can be simple and constrained, over
time people are able to validate their impression of one another by transforming
nonverbal messages into useful social information. This way, asynchronous and
anonymous media actually improved the relationships in some online groups (Walther,
1996).
Another line of thought (Social Identity Deindividuation, or SIDE) held that
precisely because of asynchrony and anonymity in an impoverished communication
41
environment, it is difficult for individuals to maintain and express their self-identities.
When there are sufficient cues that facilitate group formation, identification and
comparison, individuals will be more sensitive to the norms in the groups. As a result, a
lack of individuating cues actually contributes to a more salient social identity of the
groups (Spears, et al., 2006). Both perspectives assume that asynchrony, anonymity and
lack of nonverbal cues are the prevalent conditions in online communication, and that
these conditions make social behaviors and interpersonal communication on the Internet
different from the physical world.
Because these perspectives lack the specificity about what features configure such
social processes, they have limited explanatory power when communication conditions
change. The new media environment invalidates the assumptions in these models,
because people may no longer remain anonymous by default in their use of many online
social networks and search engines (Harris & Llanso, 2011). With the rise of
microblogging and media sharing, communication is much more interactive, multimedia
and synchronous in many communities (K. Moore, 2011). As a result, there are fewer
barriers for people to quickly start relationships, and express their personal identities. In
other words, deindividuation is no longer a necessary condition for online communities to
establish their identities. Social information is immediately exchanged and verified
without too much personal effort and time. The Internet has progressed in a way such that
theories like SIDE and SIP no longer adequately explain the identification and
communication processes in some online communities like Twitter. These theories must
be updated with a closer look at how emergent communities use specific features to foster
42
identities and improve communication. In other words, classic CMC theories like SIP and
SIDE must incorporate an ecological perspective to examine the use of different features
in social media as a community building practice (Barnes, 2008).
Social identification is a social process in which individuals become affiliated with
and attached to groups (Hogg & Turner, 1985). Groups – as a social entity based on the
alignment of individual goals and commitments – define the range of actions, means of
communication, and meanings of social practices (McGrath, 1984). Technology affects
individuals mainly by complementing, supplanting or supplementing existing social
practices within group boundaries (Brandon & Hollingshead, 2006). New media
technologies change the way people identify themselves and explore relationships in
online communities. This is possible when new technologies reshape the patterns of
social interaction, and reconfigure the structures of goals and resources in these groups.
For such groups to become an engaging social environment, a strong sense of social
identity needs to be cultivated in meaningful social interactions (Bagozzi & Dholakia,
2006; Roberts, Hann, & Slaughter, 2006). This social process is necessary for groups to
function and for individuals to explore meaningful experiences on the Internet. New
technologies will change this social process, because people may engage with different
groups more easily and flexibly. Social identification processes based on shared interest,
goals or topics may be more fluid and volatile than those based on more stable attributes
such as gender, age or political affiliation. Yet collective identities can still be widely
observed in interest- or topic-based online groups (Klimmt & Hartman, 2008; Tanis,
43
2008). Consequently, the assumptions in SIDE or SIP must be updated with the new
patterns of online community development.
Tags and Social Matching Systems
Traditionally, communities can be constructed on different forms of attachment
(Ren, et al., 2007). Individuals can be attracted to the community based on interpersonal
attraction, which can be developed in social interactions with other community members
(Spoor & Kelly, 2004). Individuals can also be attracted to the community based on
commitment to a collective identity, which can be created by self-categorization and
intergroup comparisons (Prentice, et al., 1994). With the rising use of tags, group tools
and other recommendation features in online social networks, both common identities
and interpersonal bonds increasingly become salient. Communities that used to be built
on dyadic interaction such as Twitter can gradually cultivate strong collective identities in
subgroups or sub-networks formed on trending topics (Papacharissi & Olivera, 2012).
Communities that prioritize strong collective identities over interpersonal connections
such as fan clubs and open-source software contributors can also foster close
relationships (Jenkins, 2006; Tanis, 2008). In other words, the increasing ease of
developing interpersonal relationships especially among weak ties through social
recommendation features makes it possible for a community to develop and
accommodate more interest groups. Each group may foster a collective identity among its
members, yet the encompassing social networking sites can also cultivate a general sense
of shared purpose. For example, LinkedIn can brandish a shared goal of improving social
connections among professionals, and Twitter can foster an identity based on the free,
44
constant broadcast of thoughts, opinions and events with anyone. The key is to uncover
the specific processes associated with the use of social matching systems.
A well-designed social matching system enables individuals to more easily locate
potential partners with shared interests (Terveen & McDonald, 2005). Discovery tools
enable the reconfiguration of social interactions among weak and latent ties, which in
particular increases the chance that new relationships among different subsets of groups
will be established. Current online communities afford weak ties the ability to engage in
rich communication and transparent disclosure of personal information. Theoretically,
such immediate and open exchanges of social information not only increase interpersonal
affinity, but also contribute to social identification at the group level (Heinz & Rice,
2009). When communication happens more often and in more meaningful ways, it is
easier for people to grow more familiar and bonded with more peers. They can then
engage in frequent and open discussions to improve their perception of the shared goals
for collective actions. In parallel to potential development of intimate social connections
among the new ties, there is a dynamic identification process that is less vulnerable to
environmental or personnel change (Harrison, Mohammed, McGrath, Florey, &
Vanderstoep, 2003). For example, on the social Q & A website Quora, people can follow
specific topics of questions and answers, and have direct, instant conversations with
others who have contributed. As a result, people can naturally identify themselves
together with the contributors as the groups of enthusiasts for topics like social network
analysis or entrepreneurship in Los Angeles.
45
People build ties not only through direct communication, but also through
interactions over interest groups, projects and events. Recommendation and suggestion
features thus provide both the content and context for persistent conversations and
collective actions for people to develop interpersonal bonds (Spoor & Kelly, 2004).
People become familiar and attached to others with the use of such tools, because it is
easier to identify commonality and interpersonal compatibility. For example, the social
coding site GitHub is home to thousands of open-source software projects. These projects
can be categorized by the programming languages they are written in. Users therefore
participate in any of these projects and join different groups of contributors that are
segmented by languages. Yet there is no visible barrier that prevents anyone from
communicating across these language camps. On the contrary, the site provides extensive
features for users to explore “interesting” code repositories based on how often these
repositories are being downloaded, revised or subscribed to. Members of this site can
therefore easily develop collaborative relationships with other contributors. At the same
time, they can identify themselves both as fans of particular languages and advocates of
the general principle of code sharing (McMillan, 2012). Active interpersonal interaction
becomes possible when community features encourage richer communication about
diverse topics (Ren, et al., 2007). So first, it is proposed that the use of social matching
systems should have a positive influence on the development of interpersonal bonds.
Hypothesis 3: Community members’ use of social matching systems with relevant
tags has a positive influence on their perception of interpersonal bonds in the community.
46
With the help of tags and related recommendation features, emergent groups can
be efficiently operated with a clearly communicated set of shared missions. At the same
time, these groups can accommodate conversations among different types of
interpersonal ties. As documented in numerous online activist groups and fan clubs, such
groups can become legitimate platforms for organizing for collective action (Shirky,
2008). The key assumption is that in these groups, social recommendation and suggestion
features facilitate the identification of common grounds for building social ties. In such
groups, they also support communication about the rules, norms and forms of collective
actions among different subsets of the group. For example, a simple recommendation of
“users to follow” can automatically suggest latent ties that share similar objectives or
interests. A tag that automatically filters relevant content for people, on the other hand,
can help users identify potential topics of interests and explore social ties based on the
shared interests in the topics (X. Li, et al., 2008). These features can be integrated with
messaging functions to provide necessary communication features for initiating
conversation and planning actions. As a whole, they serve as the social matching systems
that can positively influence the building of interpersonal bonds and the social
identification process in an online community.
At the behavioral level, an online community may shift its shared foci by
accommodating new trends, topics and events. The collective identities may be reflected
in many different conversations and actions. In such a dynamic environment, an effective
social matching system is the key to effective organization of collective activities,
because of its archiving, search and communication features. As long as the community
47
can organize new conversations and actions, dynamic groups can evolve to serve some
shared purposes and allow wide participation in common actions. The key to such
community functions is the systematic support for the constant re-structuring of actions
based on emergent needs and agenda, and reinforcement of a general principle of
interdependence, mutual accountability and community sanction (Connaughton &
Shuffler, 2007; van Knippenberg, De Dreu, & Homan, 2004). This is how social
matching systems in general can support the behavioral functions of most online
communities despite personnel shifts, technical change or social trends (Terveen &
McDonald, 2005).
Granted, self-categorization cues in this social identification process are weaker
than those based on demographic attributes or organizational affiliations. Shared interests
in sharing information about Pixar movies, for example, will be much harder to identify
and maintain than common gender, sexual orientation or religious beliefs. Nevertheless, it
is possible that the facilitation of identification and communication of shared interest will
contribute to a basic sense of membership and belongingness. This is because this
identification and communication process can constantly reinforce the cognition of
commonality among the members. It is certainly possible for the community members to
follow many different topics and join many different groups. Yet because all groups and
topics are generated and operated within the technical, semantic and social architecture of
the general community, the fundamental functionality of a community-wide social
matching system should still be the reinforcement of perceived identification with the
community. That is, the usefulness and effectiveness of social recommendation and
48
suggestion systems in a community for identifying with many different interest groups
actually make it more attractive and interesting to be a member of the community.
Hypothesis 4: Community members’ use of social matching systems with relevant
tags has a positive influence on their perceived identification with the community.
It is increasingly possible that a community can cultivate a strong sense of
identification and belongingness, while fostering meaningful and deep communication
among its members (Tanis, 2008). The use of social recommendation and suggestion
features make it easier for community members to join different circles and groups to
fulfill different interests needs. Meaningful bonds may be built between the members in
rich and sustained conversations, as well as through potential actions or meet-ups that
further increase interpersonal affinity. The use of such social matching features also
increases the “stickiness” of the community, improving members’ identification with the
community at both cognitive and behavioral levels. This dual development of social
identification and interpersonal bonds suggests that these two perceptions could be
correlated, with the shared topics and interests as the cognitive cues for community
membership (Hogg & Turner, 1985).
Hypothesis 5: Community members’ perceived identification with the community
is positively correlated with their perceived interpersonal bonds in the community.
Location Awareness and Social Matching
Location awareness may enrich and extend how a community works. First, it
supplements interpersonal interactions with contextual cues about people’s locations that
are made possible with the use of geocoding features. Community users can filter or
49
customize communication based on their desired locations and ranges. Second, location
awareness can help community users discover what other users are already nearby. The
knowledge of nearby peers provides a new common frame of reference that can increase
the sense of commonality and identification (Oulasvirta, Petit, Raento, & Tiitta, 2007).
For example, geo-location tags for a status update on Twitter or Facebook can create a
sense of togetherness among people who happen to be physically around. So this
supplemental layer of geographical proximity increases interpersonal affinity for both
kinds of social ties. It may become a social icebreaker for latent social ties and a reminder
for ad hoc meet-ups or conversations among existing friends. At the same time, weak ties
can benefit from this serendipitous matchmaking.
However, geographic proximity alone will not be sufficient for creating and
sustaining group cohesion and social attachment, because it needs to be tied into other
communication and identification processes that contribute to affinity (Latane & Liu,
1996). People who have no shared interests, or are outright hostile towards one another,
may not develop a sense of affinity based on geographic proximity. Fans of a college
football team, for example, may not have a strong sense of commonality with the rival’s
supporters even if they can all “check in” on Foursquare or Facebook on their
smartphones. The functions and meanings of urban places are constantly defined by
peoples’ actions and interpretations (Paay & Kjeldskov, 2008). As a result, the perception
of being together at these places as well as people’s sense of geographic proximity need
to be incorporated in concrete collective actions that are driven by shared goals, needs
and interests.
50
This suggests that location awareness, as with any other new technologies, has its
own affordances and constraints. Although it may be easier for people to be better aware
of and updated with friends’ locations, geocoding might not suit everyone’s life style and
social needs. The risk of having one’s private information compromised may be high,
which often brings unexpected consequences for some users. Because geographic data is
a very sensitive type of personal information, the use of such data in an online
community must be justified with sufficient utilitarian and psychological benefits for
most, if not all, members. Disclosure and display of every community member’s current
locations without implementing necessary communication and identification features will
not make the community more useful and attractive. Therefore, location awareness
influences the social identification and communication processes in an online community
mainly by supplementing the use of social matching systems. In other words, geotagging
rather than geocoding is the feature that improves a community’s existing functions and
foci.
A social matching system in an online community defines the common focus for
its members. For example, this system may consist of a list of recommended users with
whom to start conversations or a list of recommended topics on which to start
conversations and ideally, plan actions together (Terveen & McDonald, 2005). When
geolocation information is used to further filter and refine these shared topics, events and
interests, individuals may have a more accurate assessment of personal relevancy. In turn,
it is more efficient for people to determine the necessity and benefits for joining a
particular group and engage in different types of group actions. For example, a trending
51
topic such as “#occupyLA” on Twitter can be geotagged so each relevant tweet is
attached with geolocation data. As a result, users can have a better sense of where these
conversations or actions are exactly happening in specific locations. These users will then
have a more accurate assessment of the feasibility of joining the protests.
Because it configures the perception of proximity, an important perception that
influences interpersonal social impact, this use of location awareness simply makes the
shared purposes more salient, relevant and meaningful for a community. It makes a social
matching system more effective in identifying the collective focus for a community and
efficient in facilitating conversations among people with similar purposes. With a table of
geotagged local crime statistics, for example, a stronger shared focus on safety can
emerge among people who live in the same neighborhood, and conversations can be
organized based on the common experience of the neighborhood (Palan, 2012). As a
result, location-specific social matching as a whole can make the community more
valuable and relevant for geographically dispersed groups. Websites like MeetUp.com
may be the home for thousands of interests groups that are organized around physical
gathering over some events or topics. Yet it is the spirit and experience of face-to-face
engagement in relevant physical locations that draws users to the site and drives them to
participate in all kinds of discussions and activities (Heiferman, 2009). Location
awareness simply makes the organization of physical engagement more efficient and
makes the experience more relevant and meaningful.
52
Location Awareness and Community Development
The use of location awareness in online communities hence has interesting
implications for the development of common identities and interpersonal bonds. On one
hand, social matching systems that are enhanced with geotagging features can strengthen
existing interpersonal ties and initiate new relationships based on shared interests.
Location awareness is a unique contextual cue that implicitly increases people’s sense of
social presence and psychological engagement (Oulasvirta, et al., 2007). Geometric
distance simply determines the likelihood, scope and outcome of communication (Latane
& Liu, 1996). First, the addition of a communication feature such as geotags has the
potential of activating social contact between previously unconnected ones and enriching
communication among previously weakly connected individuals (Haythornthwaite,
2006). For example, a Facebook group organized for a protest may use the “check-in”
function to help otherwise unfamiliar members locate one another and start conversations
both virtually and physically. Location awareness simply provides additional contextual
cues for individuals to redefine and refine their search for potential conversations or
actions. With proximity as a moderating factor, a higher salience of interpersonal
compatibility and closeness can improve individuals’ engagement with others (Ravaja, et
al., 2006).
Second, geotags can also make users more accountable for their actions, since
proximity increases the likelihood of future interaction (Monge & Kirste, 1980). The
increased sense of future engagement not only makes it more likely for people to like one
another, but also makes it imperative for everyone to take responsible and answerable
53
actions (Darley & Berscheid, 1967). By making communication more relevant,
transparent and secure, location awareness improves social affinity among people who
share common interests or agenda. This transparent disclosure of geographic information
might not be ideal for everyone in the community. But for those who prefer exploring
new ties based on common interests, geolocation awareness will make their use of social
recommendation and suggestion features more efficient. Therefore location awareness
provides users with a more accurate understanding of the compatibility and similarity in
interests and objectives with other community members.
This logic suggests that geolocation features can make social matching systems
more effective at fostering interpersonal bonds. When people use and rely on such
systems, they will become more aware of the presence of others who not only share
similar interests, but also live at reachable, tangible and verifiable places. It will be easier
for community members to become connected with others over secure, open and relevant
conversations. It is therefore proposed that:
Hypothesis 6: The influence of community members’ use of social matching
systems on their perceived interpersonal bonds in the community will be stronger when
location features are used.
On the other hand, location awareness increases the effectiveness of social
matching systems in promoting social identification at both cognitive and behavioral
levels. For a community to develop social cohesion, there must be a mechanism for
individuals to equally participate in the development of shared goals and purposes. At the
cognitive level, a community can strengthen its common identity by promoting self-
54
categorization based on shared attributes, encouraging inter-group social comparison, and
creating situations that demand high interdependence (Brewer, 1979; Jetten, Duck, Terry,
& O'Brien, 2002; Tajfel & Turner, 1986). An effective social matching system may
facilitate social categorization by providing a context for individuals to explore
commonality, establish boundaries and engage in collective actions. For example, the
common identities for Twitter users can be represented by an automatically generated list
of “trending” topics. Geotags may refine the basis for perceived commonality, because
they can help users narrow down the list of topics and connections based on proximity.
Furthermore, location information can be used to set up a territorial boundary for each
such topic, facilitating the social categorization processes that emphasize intra-group
similarities and identifying inter-group differences (Cheok, Teo, Cao, & Thang, 2005; E.
Smith, Consolvo, & Lamarca, 2005; Valenti & Rockett, 2008). Finally, tangible location
information of participants helps the community create a more identifiable and secure
communication environment, which increases the overall accountability and
interdependence among the members of the community.
At the behavioral level, the additional geographic context makes the contexts for
actions more relevant and salient for community members (Bradley & Dunlop, 2005).
Geographic proximity has a regular influence on people’s perceived likelihood and
importance of interaction with others (Latane, .et al 1995; Monge & Kirste, 1980).
Therefore, geotags for a trending topic or a planned action will help people evaluate the
needs and necessity for aligning themselves with these activities. As discussed earlier,
location awareness can increase the perceived psychological involvement with the tasks
55
or events for ad hoc groups (Jones, et al., 2007). Proximity provides a sense of homophily
that helps people overcome uncertainty and unfamiliarity in the interaction with
strangers. When a community needs to engage a wide range of users in a broad set of
common actions, a strong mechanism must ensure that each action is interesting, feasible
and engaging for a sufficient number of participants despite the potential unfamiliarity
and uncertainty. Geotags can help users perceive the proximity of these groups, and thus
facilitate the mental construction of a necessary common ground. This commonality
helps community members maintain basic interests in community activities that have a
minimum level of geographical relevance. For example, a location-aware social matching
system can enable a member of a local bikers’ community to propose new routes or
events that are tagged with location information. The supplemental location data about
these routes or events will make the social focus more relevant to more people.
Therefore, at both cognitive and behavioral levels, location awareness should make
the positive influence of the perceived use of social matching systems on collective
identities in a community even stronger. It is therefore proposed that:
Hypothesis 7: The influence of community members’ use of social matching
systems on their perceived identification with the community will be stronger when
location features are used.
The theoretical model for this chapter is provided in the diagram below.
56

Figure 2: Social Matching Systems, Interpersonal Bonds, Social Identification and
Location Awareness
57
Chapter Four: Boundary Filters, Social Capital and Location Awareness
Introduction
Social capital is a helpful concept for understanding the process and outcomes of
social interactions. It refers to the productive resources that are embedded in social
relationships (Coleman, 1988). Although it is often not an explicit goal, social capital can
be the residue or outcome of patterned social practices. For example, the interaction with
close ties may make people feel emotionally safe and secure, contributing to a bonding
social experience (Aries & Johnson, 1983). Communication with weak ties may expose
one to new information and alternatives (Granovetter, 1973). Social capital accumulation
is a multidimensional process. It involves the structural configuration of relationships and
communication, the distribution of social resources such as information, trust and
psychological support, and the relational and practical outcomes of social behaviors in
groups and organizations (Nahapiet & Ghoshal, 1998). Because people’s relationships
with these ties can constantly change, however, it is rare that these different forms of
social capital are mutually exclusive (Williams, 2006). With the use of new
communication technologies, it is increasingly likely that one can access different social
capital from the same kind of ties. For example, people can still access new perspectives
or information in strong-tie networks (Aral & Van Alstyne, 2011), and strong
connections and affection could be developed among weak ties (Donath, 2007).
The prerequisite for both bonding and bridging social capital to develop in
people’s social networks is decentralized, constant communication that spans group
boundaries (Aral & Van Alstyne, 2011; Burt, 2002; Nahapiet & Ghoshal, 1998). The key
58
to such communication is the active use of tools and features to maintain connection with
strong ties and explore interactions with weak and latent ties (Haythornthwaite, 2002).
One such feature is the sharing tool that is accompanied by relationship settings. For
example, Google’s social networking service Google+ enables people to share
information with self-defined “circles” of ties, such as friends, followers or acquaintances
(C. C. Miller, 2011). People’s use of communication and privacy tools thus helps them
configure different group boundaries. The extent to which they are aware of the features
to configure communication and use them determines what kinds of social capital they
access in the everyday interactions with different ties in online communities (Madden,
2012).
The study of people’s access to different forms of social capital is essential for
understanding how people share and access information in different social circles. Social
networks like Twitter and Facebook become popular not simply because people can share
information with others they know. Such networks are popular also because they help
people join conversations, actions and events that happen outside their strong-tie circles
of friends, colleagues and families (Bakshy, et al., 2012). Privacy has risen to become
people’s major concern when they share information online (Ellison, Vitak, Steinfield,
Gary, & Lampe, 2011). When people are concerned with privacy online, they prefer to
control the communication of personal information within groups of strong ties for the
sense of security and comfort (Madden, 2012). Nevertheless, people have been
increasingly active in sharing personal information that is available for public scrutiny
(Rawassizadeh, 2011). A potential reason is the benefits that sharing networks afford.
59
Past and current evidence suggests there are practical benefits when people communicate
and interact with weak ties, such as new information or alternative perspectives (Bakshy,
et al., 2012; Granovetter, 1973).
The perception and use of privacy features therefore indicates one’s preference for
communicating with different ties. For example, if people choose to limit their sharing
within friends or even select groups of friends on Facebook, it suggests that they prefer to
send and receive information, news and opinions within these very circles of intimate
ties. The benefits for limiting social interactions only within these circles could be
security and comfort. There could be benefits for reaching out of these circles such as
new connections and new information. Yet such benefits may not always weigh
favorably against the benefits of interacting with existing strong ties, so it might make
more sense to confine communication within the strong-tie circles. In this scenario,
privacy control is a social preference for individuals to access different forms of social
capital.
This phenomenon suggests that privacy control is a calculated choice for different
benefits and risks in engagement with different social circles. People’s use and perception
of privacy features have important implications for their access to both forms of social
capital, and consequently, for their intention to share information with different people.
Practically for online community operators, designing effective privacy features is a
necessary yet challenging task, because it directly influences how users interact with
others in the communities. For communication researchers, it is also interesting to
understand what motivates people’s sharing behaviors despite the prevalent concern with
60
privacy online. When geotags are used, the affordance of proximity should make it easy
and efficient to configure the interaction with different social circles, because proximity
has a simple yet remarkable effect on people’s tendency for interpersonal communication
(Latane, et al., 1995; Takhteyev, et al., 2012). The calculation and choice of privacy
control will therefore be facilitated by the use of geolocation features. This chapter is an
attempt to examine the use of privacy features and geotags.
Structural and Technical Aspects of Social Capital
Organizational sociologists like Burt have emphasized the structural basis of social
capital. According to Burt, social capital is “the advantage as a result of location in the
structure of social relationships” (2007, p. 45). Specifically, this location is determined by
people’s membership in single or multiple groups. In this definition, groups are the
collections of individuals with shared attributes and goals. Social capital within these
groups can be embodied by strong identities, mutual obligation and psychological
security. There are structural holes across the social clusters, or disconnections across
enclosed groups. Social capital can therefore also be cultivated in the bridging
relationships that span these structural holes. Compared to social capital based on
structural closures, social capital in a “brokerage” relationship tends to be embodied by
early access to more diverse information and control of the diffusion of information
(Burt, 2002). Social capital therefore has a structural component that is determined by the
interconnections between groups (Burt, 2000).
It is natural to infer that across the structural clusters there are different forms of
productive resources. According to Resnick (2001), productive resources may include
61
communication paths, shared knowledge and values, a shared sense of collective identity,
and social obligations and norms. These resources can be available within any groups,
depending on the range and intensity of communication across relationships. For groups
of strong ties, frequent communication about topics or events of common interests is
essential for sustaining a bonding experience. As a result, it will be easier for strong-tie
groups to become the source of bonding social capital, which is characterized by
emotional support and psychological security (Burt, 2007; Williams, 2006). When
bonding social capital is salient, a stronger sense of interdependence and obligation can
arise in the groups. Consequently, a more trusting social environment is more likely to
emerge which further facilitates intensive and meaningful communication within strong-
tie networks.
Outside the strong-tie groups, people can access another type of resources when
they engage with weak ties. When workers interact with colleagues outside of their
departments or units, for example, they have the opportunity to access new information
from other divisions and serve as social brokers among the disconnected clusters (Burt,
2007). Similarly, when people extend their communication outside their strong-tie circles
in a large network, they also have a chance to forge new ties and learn new perspectives
(Heinz & Rice, 2009). Engagement with weak ties therefore makes it possible for people
to access more diverse information (Granovetter, 1973). The access to a diverse range of
perspectives and fresh opinions is essential for building bridging social capital (Putnam,
2000).
62
There is no guarantee that everyone’s strong-tie circles will be more
interconnected simply because someone actively mingles with all of these circles. Yet at
a community level, bridging social capital can foster more decentralized and participatory
communication, increase mutual awareness of the stance and values of each other, and
create a social environment that is rich in the norms of reciprocity and openness
(Nahapiet & Ghoshal, 1998). This suggests that it is generally easier for bridging social
capital to accumulate when people actively engage with their weak ties. It is possible for
strong ties to facilitate the building of bridging social capital (Aral & Van Alstyne, 2011).
There are, however, strong tendencies for enclosures of strong ties to serve as the “echo
chamber” that reinforces existing beliefs and emphasizes consensus (Gentzkow &
Shapiro, 2011). Social interactions with weak ties outside one’s strong-tie networks will
offset this potentially negative influence of social enclosure and expose one to alternative
ideas.
Social capital must be considered with new technologies that shift social structures
(Haddon, 2004). It has been noted that media technologies influence social behaviors by
resetting the group boundaries that configure access to information and communication
(Meyrowitz, 1985). For example, the use of the mobile phone blurs the division between
front-stage/public behavior and backstage/private behavior so that a person can talk about
private matters on a commuter train (R. Ling, 2008). Traditional boundaries between
performers and the audience, or between leaders and followers, are brought down when
people use electronic media to access information across groups (Meyrowitz, 1985).
Indeed, as Bimber et al (2005) argue, the most obvious impact of the Internet on
63
collective actions is that the boundary between public and private resources is blurring.
Sharing one’s private knowledge and information on the Internet to the entire public is
literally just at one’s fingertip (Clayton, 2011).
Thanks to new social networking tools like Twitter and Facebook, there is greater
flexibility in communication with a broad range of ties from family members, close
friends to acquaintances, colleagues and strangers (Hampton, et al., 2011). There are
more ways for close relationships to remain connected with one another. For example,
family members can share videos or photos to conveniently keep everyone updated with
important life events (K. Moore, 2011). Close friends often use microblogging tools to
coordinate meetings or just to remain “in the loop” (A. Smith, 2011b). With such
repeated and intensive communication, there is greater potential for strong ties to
reinforce the bonds and increase the sense of “connected presence” (Licoppe & Smoreda,
2006). Yet social networking technology also provides ways for people to improve
communication with weak ties and initiate communication with latent ties
(Haythornthwaite, 2002). Rich profiles and publicly viewable interaction logs enable
individuals to more accurately assess the “social signals” of weak ties with whom there is
limited interaction previously (Donath, 2007). Tools like hashtags, geocoding and
recommendation systems help users discover potential ties based on shared interests,
locations or preferences (Banerjee, et al., 2009; X. Li, et al., 2008; Yu & Fei, 2009).
Communication in many online communities becomes more intense and diversified, so
users can access a greater range of opinions, events and ties (Bakshy, et al., 2012; Kumar,
et al., 2010).
64
On the other hand, the mobile phone also has new features that can be used for
interaction with different ties. Individuals can switch between voice and text messaging
to discuss different things with different ties (Healey, White, Eshghi, Reeves, & Light,
2007). The increasingly sophisticated capability of context awareness further facilitates
the interaction with different people for different purposes. For example, users of a
location-aware mobile application Dodgeball can form ad hoc groups with strangers to
explore public places of common interests together (Humphreys & Liao, 2011). Friends
can use tools like Foursquare to keep updated about one another’ whereabouts and plan
for potential meet-ups based on co-location (Wortham, 2009). Mobile technology
provides greater flexibility and convenience for people to engage with a greater range of
ties (R. Ling, 2008).
Social networking and mobile technologies do contribute to an enhance capability
for people to communicate with a diverse range of ties. However, the boundary between
different social circles does not necessarily disappear (Bimber, et al., 2005). People are
still constantly aware of their online privacy, and they are more active in configuring the
boundaries for sharing their content within different circles (Haridakisa & Hanson, 2009;
Madden, 2012). The key issue is to what extent people are using specific features to
customize the scope and boundaries of communication. For example, users may set the
privacy level on YouTube so that their own content is only viewable to certain groups of
friends (Haridakisa & Hanson, 2009). This setting will influence the content of their
interaction with friends who belong these groups, and acquaintances or strangers who do
65
not (Benevenuto, et al., 2008). The awareness and use of such customization features will
thus have an impact on the users’ access to different forms of social capital.
It should be noted that the awareness and use of privacy controls in online social
networks are not always efficient or transparent (Barnes, 2006). People might not even be
aware of how to control their privacy settings, and understand the meanings and
implications of these settings. Engineers and marketers might not even have the best
intention to protect people’s privacy. Yet again, privacy control is still a new technology
that has its unique combination of affordances and risks. For some, privacy management
might be an effective tool for them to configure their online experience (Madden, 2012).
For others, the balance between openness and security might be a difficult technical issue
(Debatin, .et al, 2009). It is necessary to point out that the awareness and use – or the
unawareness and non-use – of privacy features have both intended and unintended
consequences in different social contexts. From a theoretical issue, privacy is considered
as a dynamic calculus of the social and utilitarian benefits in people’s information
exchange with others. So the use of privacy controls should be related to people’s
perception of such benefits. Privacy settings have practical implications for the
boundaries of communication between the users, their different social circles and the
public. To fully understand the consequences of the awareness or unawareness of privacy
settings, the specific structures of communication, information access and social capital
must be considered in the general context of information sharing in online social
networks.
66
Privacy and Social Capital
The privacy setting is a prevalent feature in online communities that helps users
customize the boundary condition for others’ access to their information and activities
(Debatin, Lovejoy, Horn, Brittany, & Hughes, 2009). For example, a Twitter user can
configure her settings so that her content is only viewable to people she approves of. This
user can also customize her list of followers so that she only sees the tweets and retweets
from a selected few. In other words, people can use privacy settings as a tool for
configuring the boundaries for the communication with different ties (Madden, 2012).
The extent to which people are aware of their communication choices will have a
tangible influence on the overall structural openness of a social network. This is because
the flexibility and ease for information to flow across social clusters can influence the
decentralization and density of a communication network (Aral & Van Alstyne, 2011;
Burt, 2000). As a key aspect of human groups, a boundary condition not only determines
how individuals access a community, but also shapes the range and scope of
communication (Ancona & Caldwell, 1992). An enclosed group will have more exclusive
communication on a highly focused set of topics. In comparison, a more open group will
have more inclusive communication about topics that newcomers bring in (Heinz & Rice,
2009). People’s awareness of their own privacy settings will influence their awareness of
the structure of communication and the access to different ties in a community. For
example, social networks like Path were designed to help users share information in
strictly defined and guarded social circles, in contrast to more open and loose networks
like Facebook. As a result, users will have a different perception of openness, diversity
67
and convenience for Path than other social networks like Facebook and Twitter (Donald,
2011). In this sense, the design of the boundary settings is the code that governs behavior
and resources across different social circles (Lessig, 1999). Because the actual boundary
of privacy disclosure does have a tangible influence on the scope of communication, it is
particularly necessary to examine how the perception of privacy is integrated with
people’s communication and sharing activities in online social networks.
As discussed earlier, social capital involves the structural configuration of
relationships and communication, the distribution of social resources such as information,
trust and psychological support, and the relational, informational and perceptual
outcomes of social behaviors at dyadic, group and organizational levels (Coleman, 1988).
The use of interactive communication, location awareness and interest discovery
functionalities in online communities makes it possible for people to know and interact
with other people from many social circles, such as workplaces, schools, and
neighborhoods. Conversations with these different ties depend on the awareness and use
of features that control boundaries on an individual basis. For each user of an online
community, the privacy setting is a boundary control that is related to the scope of
communication with different ties, and thus related to their perception of different forms
of social capital (Ellison, et al., 2011). The idea of privacy will be reviewed next from
this perspective.
Privacy in Online Communities
Ethics research holds that privacy is simply “the claim of individuals, groups, or
institutions to determine for themselves when, how and to what extent information about
68
them is communicated to others” (Westin, 1967, p. 7). With rising popularity of
electronic commerce, researchers started to discuss the factors that promote people’s
propensity for communicating private information online (Culnan & Bies, 2003; Minch,
2004; Raento & Oulasvirta, 2008). In this context, privacy is defined as the intentional
disclosure of personal information in expectation of certain rewards or benefits (Culnan
& Bies, 2003; Raento & Oulasvirta, 2008). According to Social Contract theory
(Donaldson & Dunfee, 1994), first, the risks and benefits for individuals in the
transactions between institutions must be balanced (Culnan & Bies, 2003; Donaldson &
Dunfee, 1994). That is, individuals’ disclosure of private information is rewarded with
something of value. The transaction between individuals and institutions must comply
with this equitable exchange of resources and benefits. Second, individuals must be
provided sufficient control over the disclosure of personal information (Gilliland, 1993).
In other words, privacy is the negotiation of intentions and techniques between
individuals and institutions for assuring an equal share of the control over the amount and
content of personal disclosure. In summary, an expectation of positive outcomes and the
necessary means for individuals to exert control over personal information are the
necessary prerequisites for individuals to disclose personal data.
Research on online privacy has generally adopted a trust-risk framework, in which
the trust of information-collecting institutions offsets the perceived potential risks in
online information exchange (Luo, 2002; Sirdeshmukh, Singh, & Sabol, 2002). The
consensus is that trusting beliefs, or the perception of the dependability of institutions for
protecting private information, reduce the perceived uncertainty and vulnerability in
69
commercial transactions and lower the perceived risks (Jarvenpaa & Tractinsky, 1999;
Morgan & Hunt, 1994). However, the effect of trusting beliefs and risks may be less
relevant when people use commercial social networks to share personal information. For
example, although consumers have very low trust on the capacity of popular social
networks like Facebook for protecting personal data, millions of people still willingly use
Facebook on a daily basis (Duncan, 2010). People may not perceive direct benefits from
their commercial interaction with Facebook as a firm that sells consumer data to
advertisers. They do, however, expect and value the benefits from sharing their everyday
life with friends and colleagues (Hampton, et al., 2011). The perception of benefits and
the capacity to control information disclosure need to be considered in the specific
context of user practices. There are supplemental mechanisms that guarantee tangible and
immediate benefits so that people can minimize the risks of disclosing personal
information.
When people are aware of the specific features that control information disclosure,
they will have a clearer perception of the benefits of disclosing such information. For
example, users of location-based mobile services have different perceptions of benefits,
when they can configure the channel for their location information to be shared with the
service provider (Xu, et al., 2010). This finding suggests that people’s awareness of the
means for controlling information disclosure and their perception of social benefits are
the key elements of the mechanisms that balance beneficial outcomes for the disclosure
of private information online. The social experience in close-knit groups based on
familial, professional and friend ties can provide a strong sense of trust, cohesion and
70
security, thus creating bonding social capital (Coleman, 1988; Williams, 2006). In
comparison, reaching out of one’s strong-tie groups and engaging with unfamiliar ties
may expose one to new perspectives, information and opportunities (Burt, 2002;
Granovetter, 1973). As shown above, the use of various communication and organization
technologies in online communities makes it possible for individuals to engage with both
strong and weak ties (Haythornthwaite, 2006). In communities like Facebook or Twitter,
features such as privacy controls are designed and implemented to configure behavior
(Kim, 2000; Lessig, 1999). For example, if a user uploads a photo of her son’s birthday
party on Facebook, she has the choice of sharing this photo with some of her friends, all
her friends or the whole user community. She may also configure her Twitter settings so
that the photo she shared is only viewable by her followers. If people are aware of the
choice for configuring different privacy settings, they will be aware of the balance and
boundary of their social interactions with their strong-tie and weak-tie networks.
Different forms of social capital can develop together in either strong or weak ties (Aral
& Van Alstyne, 2011; Ellison, Steinfield, & Lampe, 2007), which require repeated, open
and meaningful communication over time.
The discussion above suggests that privacy in online communities is a form of
boundary control for communication with different ties. Because communication with
different ties has an impact on people’s access to social capital, the awareness of privacy
configurations may be related to people’s perception of different forms of social capital
online. This integration of the privacy perspective and the framework of social capital
illustrates a potential reason for online communities to become meaningful spaces of
71
communication and social capital. When people are aware of how they can configure the
scope of their communication with different ties, they will have a clear idea of the
benefits they can obtain by their disclosure of personal information in a community
(Raento & Oulasvirta, 2008).
On one hand, if people are willing and capable to initiate and join diverse
conversations with many different groups outside their strong-tie circles, the community
will have a greater potential for facilitating diverse participation and fostering bridging
social capital (Tanis, 2008). The overall increase in the intensity and diversity of
communication at a community level will even make it possible for strong ties to provide
new information, ideas and opinions, overcoming the “weakness” of strong ties (Aral &
Van Alstyne, 2011). On the other hand, if people are capable of strategically limiting the
scope of their communication, they can focus on fostering stronger trust and cohesiveness
with specific ties and cultivating bonding social capital. It is even possible to extend
one’s existing strong ties and build new intimate relationship, when new relationships are
formed over shared interests and are reinforced with sustained communication and shared
activities (Heinz & Rice, 2009). The affordance of controlling communication boundaries
in online communities therefore explains the phenomenon that people value an open
atmosphere that facilitates sharing, yet still pay attention to the privacy issues in their
communication of personal information (Madden, 2012).
Privacy control is therefore related to people’s perception and awareness of social
capital in their engagement with different circles. The use or conscious control over one’s
informational privacy indicates that the person is capable and conscious of strategically
72
configuring the boundaries of her social networks. Following this logic, it is proposed
that:
Hypothesis 8: Community members’ use of privacy controls that configure the
disclosure of personal information to strangers or friends will have a positive influence
on their perception of bridging social capital in the community.
Hypothesis 9: Community members’ use of privacy controls that configure the
disclosure of personal information to strangers or friends will have a positive influence
on their perception of bonding social capital in the community.
It is reasonable to assume that people’s perception of bonding and bridging social
capital should be positively correlated. People are constantly engaged in many different
circles and groups in a community. The regular access to both strong and weak ties
makes it more likely that a community can foster bonding social experiences through rich
and intensive communication while also exposing one to a diverse set of perspectives
(Bakshy, et al., 2012). Even though people may desire different kinds of social capital in
their interaction with strong and weak ties, the boundaries between these ties may be
fluid: people become more familiar with weak ties over repeated and sustained
interactions, or they may be alienated from some strong ties in important transition
periods in life. Human networks are dynamic and changeable enough, so that it makes
little sense to maintain an arbitrary boundary between strong and weak ties. As a result,
different forms of social capital should coexist and correlate with one another in people’s
use of online communities. It is therefore proposed that:
73
Hypothesis 10: Community members’ perception of bonding social capital is
positively correlated with their perception of bridging social capital in the community.
Location Awareness and Privacy
Location awareness might bring usability benefits for online community users,
such as personalized and location-specific topics and services (Cuneo, 2008). More
importantly, geolocation technology affords users the capability to customize their
experience based on the proximity of events and people. Because proximity has a
fundamental influence on the way people perceive relational closeness and the
immediacy of others’ social impact, location tools like geotags can facilitate people’s
engagement with different ties, groups and events (Latane & Liu, 1996; Latane, et al.,
1995). Yet the wide use of geolocation technology for detecting and recording people’s
physical locations also dramatically increases the concern with privacy online, because
location information is a very sensitive piece of personal data. Voluntary disclosure of
one’s whereabouts needs to be based on a clear recognition of the usability and social
benefits (Junglas & Watson, 2008). Users also need to be convinced that the risks to
personal security and privacy are adequately mitigated with technical, regulatory and
institutional measures (Xu, Teo, Tan, & Agarwal, 2010).
Because geotags become part of interpersonal communication in online
communities, the use of location features in online communities is also an issue of
balance between the benefits of personalization, customization and convenience and the
control over disclosing personal locations with the appropriate parties. Yet because it
configures the perception of geographic proximity, geolocation technology also
74
reinforces the influence of privacy controls on the scope of people’s communication. For
example, if a geotag is attached to a user’s broadcast of daily activities, this user will
want to be more selective and careful in configuring the boundaries for the broadcast.
This user may also use the “proximity” scale to further refine the communication within
specific circles, for example, with only her friends who live within a five-mile radius. In
this sense, geolocation awareness is a moderating factor in the use of privacy controls for
achieving personal benefits. It magnifies the positive outcomes of configuring one’s
communication boundaries in a way that matches one’s intentions and goals.
On one hand, geolocation features may make the awareness of privacy controls a
more important influence on the perception of bridging social capital. In a community,
shared physical locations may provide a basis for individuals to explore local events and
ties that may exist beyond their intimate social circles. The combination of geographic
proximity and shared interests makes it possible for ad hoc groups to emerge (Humphreys
& Liao, 2011). Yet the discovery of co-located interests relies on an inclusive, diverse
and open social environment. This social environment is possible when the users desire to
access new information, perspectives and relationships from nearby. For users to explore
relationships beyond their existing social circles, they need to have a stronger awareness
and tendency to configure the boundary of their information disclosure. For example, if a
user of Foodspotting - a mobile application for sharing photos and recommendations for
foods - intends to find new friends nearby who share interests in bacon sandwiches, she
will have to know how to configure her privacy settings so that she could publish her
location and her past interests in the food. When location features help users customize
75
communication based on their current locations, it makes more sense for them to be
aware of the means to actively reconfigure their interaction with new ties. Because
geolocation features provide a proximity cue, users should be primed to feel a stronger
opinion and interest about the latent and weak ties around them. In other words, a weak
or latent tie from one’s neighborhood matters more than a similar social tie 5,000 miles
away. Geotags, in this sense, serve as a filter that makes the influence of privacy controls
more obvious and tangible. It is therefore proposed that:
Hypothesis 11: The influence of community members’ awareness and use of
privacy systems on their perception of bridging social capital will be stronger when
location features are used in the community.
On the other hand, the connection between the use of privacy controls and access
to the bonding social experience in strong-tie groups may also be stronger in a location-
aware community. Location awareness provides an additional set of contextual cues that
increase the mutual social presence within groups of people with histories of interaction
(Dey & Häkkilä, 2008). Geographic proximity simply has a reinforcing effect on people
with existing bonds, commonality or shared objectives (Monge & Kirste, 1980).
Therefore, location awareness may be more useful and meaningful when strong ties such
as friends or colleagues already share a stronger sense of mutual engagement and
interdependence (Hampton, Livio, & Goulet, 2010). For such strong ties, location
awareness will provide a sense of social and territorial togetherness that reinforces
communication and participation in collective activities such as serendipitous meetings or
travels (Wortham, 2009). Even for situations that do not require being at the same
76
location, a simple tag of physical locations will improve communication between friends
(Banerjee, et al., 2009). For example, messaging between friends and families could be
enhanced when each other’s location is automatically disclosed, leading to richer
communication about potentially new topics (Counts, 2007). As a result, people are more
aware of each other’s physical and situational environments. They can then communicate
with one another with better understanding of each other’s situations. In this scenario,
location awareness simply makes the disclosure of private information a more important
ingredient of the bonding social experience among strong ties. It also makes the active
and effective use of privacy features a more important condition for accessing bonding
social capital. It is therefore proposed that:
Hypothesis 12: The influence of community members’ awareness and use of
privacy controls on their perception of bonding social capital will be stronger when
location features are used in the community.
The hypotheses in this chapter are illustrated in the diagram below.
77

Figure 3: Privacy Controls, Social Capital and Location Awareness
78
Chapter Five: Information Sharing in Online Communities
Introduction
Information sharing is an important activity for online communities, and effective
exchange of relevant information plays an important role in community building
(Tapscott & Williams, 2008). Social networks like Twitter rely on its millions of users to
constantly share thoughts or events that hold common interests for different groups of
people (A. Smith, 2011b). Other online communities that have more focused purposes,
such as Yelp.com, survive and thrive when amateur reviewers voluntarily share
experiences and evaluations of local businesses (Pattison, 2010). Successful information
sharing is often associated with successful knowledge integration, reuse and application,
which enable organizations to accumulate intellectual capital and gain competitive
advantage (Heinz & Rice, 2009; Markus, 2001). Similarly, online communities depend
on the contribution of relevant, interesting and useful information from their users in
order to become sources of relevant information and meaningful social interactions
(Haythornthwaite, 2006).
It is worthwhile to examine the social processes that motivate voluntary
contributions. Effective information sharing entails the recognition and contribution of
domain knowledge as well as effective retrieval from the knowledge pool (Robert, et al.,
2008). That is, sharing knowledge or information is inherently a two-way process,
requiring individuals to engage one another in meaningful communication and exchanges
of private information. An organization and a successful community both benefit from
the effective organization of collective actions and social interactions that facilitate
79
diverse flows of meaningful information. In organization studies, it has been documented
that various information and communication technologies (ICTs) such as intranets
(Hollingshead, Fulk, & Monge, 2002), knowledge management systems (KMS) (Heinz &
Rice, 2009) and discretionary databases (Fulk, Heino, Flanagin, Monge, & Bar, 2004)
can facilitate knowledge sharing. Despite the obvious utility and apparent ease of
contributing information, however, knowledge sharing in organizations is not always
effective. For example, individuals might be unaware that they have something to share
(Sieloff, 1997). They might be simply unwilling to disclose their unique knowledge for
fear of losing competitive advantage (Cabrera & Cabrera, 2002). Hierarchies and inherent
differences in backgrounds, goals and experience often make it difficult to coordinate
knowledge contribution across different departments. Understanding the motivations
behind contributions despite these barriers is an important task for organizational
researchers (Olivera, Goodman, & Tan, 2008).
Similarly, online communities often face the problem of free riding and deficient
contribution (Shirky, 2008). The use pattern of websites from Wikipedia to YouTube and
Twitter has shown the same power-law distribution: most of the content is created by a
small fraction of users, and consumed by a majority of users who never contribute
(Kumar, et al., 2010; Mislove, Viswanath, Gummadi, & Druschel, 2010; Yu & Fei,
2009). Non-contribution and free riding might not be a big problem for these
communities on the surface because of their tremendous global user bases. Yet for
communities that cater to niche interests or focus on narrower purposes such as the
hyperlocal news community EveryBlock.com or the question and answer site Quora.com,
80
user participation determines their livelihood (Ingram, 2011). The interesting questions
for both organizations and online communities are why and how people voluntarily share
something that they deem relevant and important.
The investigation of effective information sharing across different contexts is
particularly relevant, as new communication technologies are gradually implemented in
both consumer and enterprise markets. For example, microblogging tools built on the
same functionalities of instant, multimodal, multidirectional and interactive
communication are not only implemented by companies to facilitate internal
conversations (e.g. Yammer). They are also used by Twitter users to discuss news events
and organize civic actions (A. Smith, 2011b). The technologies that power Wikipedia are
becoming the necessary tools for organizations to build knowledge repositories with
greater accessibility and higher user engagement (Tapscott & Williams, 2008). How and
why effective information sharing happens in both business enterprises and online
communities can be examined with the common theoretical lens of ICT-facilitated
information sharing. Regardless of the organizational forms of knowledge sharing,
common social and cognitive processes underlie the use of new technologies for
organizing collective actions. When new technologies like geolocation awareness afford
easy and efficient customization of content based on proximity, it is necessary to examine
how this new condition of communication facilitates or constrains participation.
Social Influence on Information sharing
Information sharing is often framed as a problem of collective actions (Fulk, et al.,
2004; Heinz & Rice, 2009; Hollingshead, et al., 2002). The logic is that it requires
81
contribution at individual cost for the production of public goods that are non-excludable
(i.e., everyone can use it) and non-rivalrous (i.e., one’s use will not make it less usable for
others). The fundamental issues for collective actions are the motivation for contribution
against the choice of free riding, and coordination of participation among discrete
interests (Hardin, 1982; Olson, 1965). Although knowledge in organizations is different
from the information that is shared in online communities, the problems of non-
contribution in information-sharing communities can be considered from a similar
perspective. While collective action involves a cost/benefit calculus from the perspectives
of economists and political scientists, for communication researchers collective action is
a social cognitive problem. It is the problem of identifying how people voluntarily justify
and optimize the efforts of contribution. The key is to identify why people want to share
something that they think is relevant and interesting for others to know.
Although information sharing is an individual action, the willingness and capacity
to contribute information is often shaped by the social environment that people live in (K.
Ling, et al., 2005; Wasko & Faraj, 2005). Technologies affect the information sharing
process in different ways. Some technical features facilitate the social cognitive
mechanisms that improve motivation, such as reputation metrics and reward systems
(Roberts, et al., 2006). Other features mitigate structural barriers to communication, such
as discovery and messaging features (Nardi, 2005; Zammuto, Griffith, Majchrzak,
Doughetry, & Faraj, 2007). In specific, a technology is more effective for information
sharing if it facilitates social identification with the group or organization, increases
interpersonal affinity, reinforces reciprocity, and enables decentralized and open
82
participation (Heinz & Rice, 2009). Technology is less effective if it lacks
communication, storage and retrieval features for facilitating these social processes. The
processes of identification, reciprocity and inclusiveness have a similar impact on
participation in online communities, as demonstrated in the studies on various
communities from social support groups and eBay to online fan clubs and open source
software developer groups (Bagozzi & Dholakia, 2006; Roberts, et al., 2006).
Consequently, the extent to which geocoding promotes these social processes determines
the impact of this new technology on information sharing activities. The positive
influence of social attachment, reciprocity and bridging social capital on sharing
motivations is discussed next.
Social Shaping of Motivations for Contribution
Information sharing is a social activity. When people join a group, they engage in
the exchange of unique information that each one possesses or has access to. The
exchange process demands the collective efforts of identifying information domains,
mobilizing information flow, and coordinating the contribution process and outcomes
(Robert, et al., 2008). To contribute to the information good, people need to make efforts
to create relevant and valuable knowledge. Because individuals need to constantly align
themselves with the norms and objectives in their social groups, it can be very
burdensome to contribute information. For example, on the code sharing website GitHub,
once a contributor is committed to writing and sharing new code for a project, they must
adhere to the goal, coding style and expectations that are set by the project contributors
(McMillan, 2012). For such voluntary, and sometimes time-consuming work, individuals
83
must have strong motivations, or perceptions that justify the allocation of mental
resources with sound evaluation of desired outcomes (Kanfer, 1990). People’s
motivations might be driven by both internal and external factors (Roberts, et al., 2006).
They are motivated to contribute information not only because they feel capable and
willing to, but also because they identify with the embedding social norms and values,
and abide by the structures and protocols of communication (Heinz & Rice, 2009). It is
often the internalized extrinsic motivation - such as the innate satisfaction of status
recognition - that promotes contribution behavior (Olivera, et al., 2008; Robert, et al.,
2008). This kind of extrinsic motivation must increase individuals’ identification with the
community’s norms and values. Therefore social influence often plays an important role
in individuals’ evaluation of such rewards.
Participation in information sharing could be motivated by social processes that
define the context and scope of personal intentions (Schmitz & Fulk, 1991). People are
dependent on one another in their social lives. They often set goals that are relevant to
their own social groups or classes (Mischel, 1968). People’s behaviors could be easily
influenced by others (Bandura, 1985), or rather, by their perception of a particular
situation in their engagement with others (Mischel, 1973). In the field of communication
research, studies have shown that social influence has a strong effect on the adoption and
use of new technology (Fulk, Schmitz, & Steinfeld, 1990). When others’ actions can be
observed, social processes like reciprocity and social identification all have significant
effects on participation in collective actions like knowledge contribution (Fulk, et al.,
2004; Kankanhalli, et al., 2005). Individual dispositions could have a large influence on
84
the decision to join a community, yet the social environment in the community may have
a great influence on the goals and meaning of participation (K. Ling, et al., 2005). For
example, even though a user may have joined Yelp to express her dissatisfaction in a
restaurant review, her continued participation will depend on the extent to which she
perceives Yelp as a community that offers shared identities and connection with friends.
First of all, when individuals perceive a social environment that rewards mutual
obligation and cultivates a sense of reciprocity, they are more likely to share information
with others (Kankanhalli, et al., 2005; Robert, et al., 2008). This is because perceived
risks for collective actions – such as people free riding on others’ contribution – are
reduced, and perceived benefits – such as opportunities for rewards – are increased in a
social environment that values reciprocity and mutual obligation (Bock, Zmud, Kim, &
Lee, 2005; Kankanhalli, et al., 2005). Second, people are more likely to contribute, when
individuals share a strong sense of purpose and feel aligned with the shared identities the
embedding groups (Bagozzi & Dholakia, 2006). This is because a strong shared identity
facilitates the group categorization process and increases the social cohesiveness in
repeated and intensive communication (Bagozzi & Dholakia, 2006; Burt, 2007). With
strong identities come with strong commitment to the collective goals such as solving a
programming problem or completing a political petition (Bagozzi & Dholakia, 2006;
Shirky, 2008). Finally, information sharing can also be motivated by frequent and
meaningful social interaction at both group and organizational levels. When people are
actively involved in diverse communication across structural closures or group
boundaries, they are more likely to share unique knowledge with otherwise disconnected
85
groups and uncover unspecified information needs (Burt, 2007). A more decentralized
communication network is more participatory, increasing individuals’ interest in others’
information and their own self-worth (Dennis & Garfield, 2003). Decentralized
communication also makes it possible for people to access multiple angles of
interpretation and understand the contexts and situations of others, which improves the
quality and relevance of communication (Faraj & Sproull, 2000; Kudaravalli & Faraj,
2008).
In sum, mutual accountability, social identification, and decentralized
communication are some of the key social processes that help turn an organization or
community into a participative social environment (Nahapiet & Ghoshal, 1998). These
processes shape the development of shared mental models about the collective goals,
build the trust that is based on reciprocity and social obligation, and facilitate open and
participatory communication. Yet still, the idea that social influence shapes individual
contribution to online communities demands detailed examination. This is because new
features in online communities like social networking and instant messaging have
fundamentally changed how social influence works (Heinz & Rice, 2009). These features
may reinforce some community-level social processes that increase members’ intention
to contribute information. These processes may be the key to understand how the social
influence of “persuasive technologies” works in this new context (Fogg, 2002). The
constructive social processes that influence individuals’ participation can be examined in
the new context of collective action online.
86
Online Information Sharing as ICT-enabled Collective Action
As discussed earlier, information sharing is inherently a collective-action problem
(Fulk, et al., 2004). It involves voluntary collective efforts from multiple individuals, and
the outcome of sharing is a public good that is non-excludable and non-rivalrous. It is
believed that the Internet has lowered individual cognitive cost for participation in
collective action. Bimber et al argue that because the Internet has made participation and
coordination so easy, collective action is virtually “a set of communication processes
involving the crossing of boundaries between private and public life” (2005, p. 367). The
basic idea is that since contribution of information online is so effortless (leaving a
comment on a blog, saving the browser cookie for Google’s search algorithms,
retweeting or commenting a piece of breaking news, etc.), participation in collective
actions no longer requires intensive coordination of intentional efforts of committing
considerable resources (Shirky, 2008). Instead, collective action is determined by the
process of “interaction and negotiation of individuals’ communicative and information
environment” (Bimber, et al., 2005, p. 372). In other words, online collective action is
essentially a process in which people communicate with one another and identify the
degree of alignment of goals and interests within certain social structures.
Technologies influence online collective actions on two dimensions: by
reconfiguring interpersonal communication and by balancing the coordination structure.
Numerous studies have covered how the awareness and use of different features influence
the social dynamics of communication in the context of online communities. For
example, scholars show that detailed profiling and interactive communication tools help
87
individuals more effectively verify information sources and increase mutual
understanding in online communities (Ma & Agarwal, 2007; Postmes, 2006). Enhanced
self-expression and identification tools contribute to a more efficient attribution process
where the negative effect of situational invisibility – common in an online environment -
on information sharing is minimized (Cramton & Orvis, 2003). Community-wide
reputation and feedback systems make individuals personally identifiable and
accountable for their previous activities. Such systems not only help community members
make more informed decisions about knowledge sources. They also make sure
individuals’ previous and future contribution be rewarded with positive feedback and
higher reputation ranking (Resnick, et al., 2000). As a result, profiles with reputation and
other background information help users develop a higher level of trust and a stronger
sense of social presence, thus effectively mitigating uncertainty in information exchanges
(Pavlou, Liang, & Xue, 2007). In other words, new communication technologies help
individuals communicate with more people in more efficient ways. Yet the use of
technologies like reputation systems can also reinforce the norms about reciprocity and
mutual obligation in this emergent process (Fuller, et al., 2007).
It is in the use of communication, recommendation and filtering features that
constructive social processes promote effective communication and efficient structuring
(Heinz & Rice, 2009). Such social processes ensure equitable participation in collective
actions and maintain an appropriate level of mutual obligation and commitment.
Therefore, the effect of new technologies on collective actions can be understood in two
respects – interpersonal communication and the structure of organizing. To create a
88
positive social environment and overcome the structural barriers to sharing, an online
community must deploy technologies that facilitate communication across different ties,
while promoting mutual obligation and interdependence in dynamically formed social
groups (Katz, et al., 2004). This is the general principle for examining how effectively
different new features in online communities foster the positive constructive social
processes. Indeed, this view is echoed in Flanagin et al’s model (2010). This model
proposes that collective actions online involve the combination of interpersonal and
impersonal interaction and the integration of hierarchical and decentralized modes of
coordination. Technologies facilitate such collective actions by optimizing different
modes of communication and balancing the structures of coordination in different
contexts. Phenomena like open-source software communities, flash mob organizations
and information-sharing networks all illustrate the importance of implementing different
technologies for fostering diverse communication and flexible social structures (Flanagin,
Stohl, & Bimber, 2006).
Features such as reputation systems, recommendation systems and privacy
controls, if effectively implemented, can optimize communication and structuring
processes in online communities (Forman, et al., 2008; Raento & Oulasvirta, 2008; Zhu,
2010). Specifically, reputation systems can increase mutual obligation, individual
accountability and general trust (Fuller, et al., 2007). Therefore they instill a sense of
mutual interdependence in a community where people constantly connect and disconnect
with one another. Privacy controls help individuals configure the boundary and context
for the interaction with different kinds of ties (Hoadley, Xu, Lee, & Rosson, 2010). Such
89
systems may help people control and secure their communication with strong ties, but
may also facilitate wide-reaching conversations and actions with hidden and weak ties
(Debatin, Lovejoy, Horn, Brittany, & Hughes, 2009). The use and awareness of privacy
systems may affect the access of social capital (Ellison, et al., 2011). Recommendation or
discovery systems based on keywords and tags can make the identification of shared
interests and goals among peers more efficient (Schifanella, Barrat, Cattuto, Markines, &
Menczer, 2010). As a result, these systems may reinforce shared interests that strengthen
people’s emotional and cognitive attachment to the community (Prentice, et al., 1994).
Therefore, the key to understanding the impact of new features on online collective
actions is the extent to which these features influence the social processes of attachment,
reciprocity and social capital in the online community. Namely, with community-wide
features that facilitate identification, communication and participation, strong common
identities can be established in emergent, loose networks of practice. Diverse,
decentralized communication is guaranteed with a sense of mutual accountability and
trust (Heinz & Rice, 2009). Reciprocity, community attachment and social capital are
some of the most important constructive processes that contribute to the relational,
cognitive and structural resources in a community (Robert, et al., 2008).
Hypotheses
In this study, location awareness is a feature that may affect existing socio-
technical practices in online communities. Because location and other contextual
information can be conveniently exchanged without too much effort, individuals can
optimize their mental resources to coordinate activities that are more central to the tasks
90
or goals (Oulasvirta, et al., 2007). For example, Facebook and Twitter now both support
automatic geotagging, so that users can attach location information to their postings. A
citizen journalist can use Twitter to report a protest without having to spend time on
reporting where she is. In this sense, location awareness should improve the efficiency
for individuals to contribute pertinent information.
However, to predict whether such new features make information sharing more
effective, it is necessary to understand that the social use of such features is shaped by
existing communication practices and norms of engagement (Haythornthwaite, 2006). In
order for online communities to cultivate a more secure and trusting environment,
location awareness features need to be effectively integrated with existing features such
as reputation systems, social matching systems and privacy control systems. Effective
integration facilitates socio-technical practices that lead to higher reciprocity, stronger
attachment, and broader access to social capital. For this assumption to work, it is
proposed that the overall effectiveness of collective actions – as indicated by the quantity
and perceived quality of information sharing – is indeed associated with the social
processes identified above. As noted earlier, this assumption is drawn from the discussion
about the motivation processes for collective actions across different communities and
organizations. Its premise is based on the fundamental principles of collective actions in
the new media environment (Bimber, et al., 2005). Its logic is deduced from the analysis
of the common drivers of participation in information sharing in communities and
organizations (Bagozzi & Dholakia, 2006; Kudaravalli & Faraj, 2008). Hence,
reciprocity, community attachment and social capital – the consequences of integration
91
and appropriation of various technological affordances in online communities – should be
related to the actual quantity and perceived quality of knowledge sharing in an online
community.
First, reciprocity is an important aspect of a community environment that promotes
contribution (Kankanhalli, et al., 2005). If one expects his or her contribution to be
reciprocated with rewards from the community or recognition from others, this sense of
reciprocity will be positively associated with a stronger perception of mutual engagement
and accountability (Ma & Agarwal, 2007). As the consequences of good and bad
behaviors are transparent, a community-wide sense of reciprocity helps establishing a
social environment in which members interact with others more securely and efficiently.
Furthermore, system-wide reciprocity ensures that reputation systems or similar
mechanisms exist to effectively record and regulate community members’ behavior. This
systematic reciprocation provides the basis for satisfying contributors’ needs for
reputation and status (Roberts, et al., 2006). As a result, rewards in the form of reputation
and status can be fairly distributed in a highly reciprocal environment. For example, the
computer programming community StackOverflow employs a sophisticated reputation
system that awards points to members based on the relevance of their questions and the
quality of their answers. These points accrue to establish the reputation of a user in this
community. In this sense, extrinsic rewards become the basis of a community member’s
need for status recognition. The sense of reciprocity therefore serves as the cognitive
basis for participation motivation. It is therefore proposed that:
92
Hypothesis 13a: The quality of individuals’ contribution will be positively related
to their perception of generalized reciprocity in an online community.
Hypothesis 13b: The quantity of individuals’ contribution will be positively related
to their perception of generalized reciprocity in an online community.
Second, effective information sharing depends on an active alignment of individual
and collective goals. Intentions for contribution can be increased when individuals share
a strong sense of collective goals and a strong alignment with a common identity
(Bagozzi & Dholakia, 2006). When individuals are attracted by the collective purpose of
the community, they will be more likely to expect the satisfaction of both extrinsic needs
such as money and internalized needs such as status and reputation (Roberts, et al., 2006;
Wasko & Faraj, 2005). Yet contribution may also be attributed to interpersonal bonds.
For example, people share stories on Twitter not only because such stories contain topics
of mass interest, but also because they have the potential to initiate conversations
between people’s followers or friends (A. Smith, 2011b). Both social identification and
interpersonal affinity can contribute to social attachment in current online communities,
an important factor for individuals’ motivation to participate and contribute.
With new features like tags, content filters and instant messaging, current online
communities can easily foster rich communication among people, so that they could
develop stronger bonds over repeated high-quality communication about common
objectives (X. Li, et al., 2008; Zhu, 2010). Thanks to the affordance of recommendation
and discovery features, shared goals and purposes can be more effectively articulated and
refined at the community level (Shirky, 2008; Yu & Fei, 2009). These communication
93
and recommendation features effectively connect the community at both dyadic and
global levels, because the pairs of nodes who share interests gradually evolve into a
social network of ties (Eagle, Pentland, & Lazer, 2008; Haythornthwaite, 2006). In
repeated and extended communication about shared interests, objectives and agenda,
individuals may perceive a social environment that is rich in both interpersonal affinity
and social cohesion. That is, people may feel connected with other members in specific
and with the “bigger picture” of the community at the same time (X. Li, et al., 2008;
Shirky, 2008; Utz, 2008). In consequence, they may be more attached to the community
and more willing to participate in information sharing. For this reason, interpersonal
bonds and social identification are both salient constructive social processes that promote
participation.
Hypothesis 14a: The quality of community members’ contribution will be
positively related to their perception of interpersonal bonds in the community.
Hypothesis 14b: The quantity of community members’ contribution will be
positively related to their perception of interpersonal bonds in the community.
Hypothesis 15a: The quality of community members’ contribution will be
positively related to their perceived identification with the community
Hypothesis 15b: The quantity of community members’ contribution will be
positively related to their perceived identification with the community.
Finally, social capital can influence participation in community actions in different
ways. Above all, the volume of information sharing increases when there is more
frequent and meaningful social interaction with diverse ties. Because it is easier to engage
94
with new groups and form new ties, people are more likely to share unique information
with otherwise disconnected clusters of knowledge domains and uncover unspecified
information needs from them (Burt, 2007; Granovetter, 1973). It is also increasingly the
case that people can access new information even within the clusters of close friends and
other strong ties (Aral & Van Alstyne, 2011). This is because new technologies like
social networking makes the flow of information relatively easily across different social
groups, and contributes to a more open environment for expanding people’s circles of
communication (Bakshy, et al., 2012). A more open communication network is more
participatory and inviting, increasing individuals’ willingness to share information and
enhancing their perceived self-worth (Dennis & Garfield, 2003).
On one hand, because there is more communication across different group
boundaries, the overall density of connection and frequency of information exchanges
will be higher. On the other hand, bridging social capital inherently entails the access to
new perspectives and information embedded outside individuals’ strong-tie groups. As a
result, the motivation for contributing quality information may also be higher than when
people consider themselves engaging in closed and redundant communication (Burt,
2007). Bridging social capital will therefore be positively related to both the quantity and
quality of participation in information sharing.
Hypothesis 16a: The quality of individuals’ contribution will be positively related
to their perception of bridging social capital in a community.
Hypothesis 16b: The quantity of individuals’ contribution will be positively related
to their perception of bridging social capital in a community.
95
In comparison, bonding social capital – or relational and emotional support
normally associated with enclosed groups of strong ties – does not necessarily contribute
to people’s motivation of participation in the same ways. When individuals intentionally
use features such as privacy controls to configure communication in their strong-tie
groups, they prefer bonding social experiences that bring them comfort, security and
familiarity within these social circles (Szmigin & Reppel, 2004). At the community level,
the focus on communication within clusters of strong ties makes it less necessary to
explore unfamiliar territories. Communication with weak and latent ties is likely, but a
relatively closed community is less receptive to more diverse conversations and
alternative perspectives (Burt, 2007). Bonding social capital may still contribute to the
volume of communication and the quantity of contribution, because individuals may
communicate frequently within their strong-tie groups if they perceive the social benefits
of doing so (Aral & Van Alstyne, 2011). Yet bonding social capital may not contribute to
the motivation for contributing quality content. This is because it may be more difficult
for new perspectives and alternative opinions to be accommodated in communication that
is perceived to reinforce the bonding social experience. For example, groups of strong
ties may choose to discuss familiar topics and conform to existing channels, norms and
styles (Dutta-Bergman, 2004; Licoppe & Smoreda, 2006). It is therefore proposed that:
Hypothesis 17a: The quality of individuals’ contribution will be negatively related
to their perception of bonding social capital in a community.
Hypothesis 17b: The quality of individuals’ contribution will be positively related
to their perception of bonding social capital in a community.
96
Generalized reciprocity, community attachment and social capital are some of the
constructive social processes that facilitate information sharing. They may contribute to
an optimal balance between interpersonal and impersonal communication, as well as
between bounding structures and emergent networks (Flanagin, et al., 2006). In a
location-aware community, the affordance of configuring content and communication
based on users’ current locations turns proximity into an important condition for
developing social affinity. Geographic proximity has a regular and positive influence on
the intensity and frequency of interpersonal communication (Latane, et al., 1995;
Takhteyev, et al., 2012). Since the use of reputation features depends on the effective
communication of personal activities and data, geographic proximity should improve the
effectiveness of such features for reinforcing the norms of recognizability, accountability
and reciprocity. In a similar vein, geographic proximity should also improve the quality
of communication about shared interests and increase the relevance of privacy controls.
These positive effects of geographic proximity can all be made possible when
geolocation features are integrated within the common practices of a community. It is
therefore reasonable to assume that the use of geolocation features should promote the
constructive processes in the community. In other words, features like geotags should
facilitate reciprocity, attachment and social capital. As a result, geolocation awareness
should make the community easier to use and increase participation. It is therefore
proposed that:
Hypothesis 16a: The perceived quality of individuals’ contribution will be higher
when location features are used in a community.
97
Hypothesis 16b: The increase in the quantity of individuals’ contribution will be
higher when location features are used in a community.
Nevertheless, it is unreasonable to assume that reciprocity, attachment and social
capital have equal influences on individuals’ participation in a location-aware
community. When features like geotags help users customize content and social
interaction based on their geographic neighborhoods, the foci and objectives of the
community may change as a consequence (Beale & Lonsdale, 2004; Bristow, et al.,
2004). When a user chooses to interact only wither people from their neighborhoods, the
affordance of proximity, familiarity and similarity may make reciprocity a less powerful
process for fostering mutual accountability. Geotags may already provide a form of social
identification, for example, “the residents of the zip code 90210”. In this scenario, social
identities associated with other kinds of commonality such as interests or gender may no
longer influence participation. Because geographic proximity makes face-to-face
interactions more feasible and efficient, the definition of social capital may also be
broadened or narrowed to reflect people’s preference for physical actions. For example,
people might consider yard sales within neighborhoods as a form of bridging social
capital. In comparison, they may prioritize physical meetings over casual phone
conversations as a valid form of bonding social capital. In other words, the perception of
geographic proximity may interact with other necessary processes such as reciprocity and
interpersonal bonds to influence individuals’ intention to share information in a
community. Therefore, a research question is proposed below:
98
Research question 1: Will reciprocity, community attachment and social capital
have different influences on participation when location features are used in a
community?
The hypotheses in this chapter are summarized in the diagram below.

Figure 4: Perceived Quality and Actual Quantity of Sharing

99
Chapter Six: Method
Technical Overview
A field experiment was conducted in which the main stimulus is access to
geotagging features. For this experiment, a web application was built which allowed
users to post, view and share posts about their personal needs and offers of help, advice,
and personal belongings. This web application was integrated with Twitter, so that tweets
of similar nature were queried, categorized into “offers” or “needs”, and displayed along
with postings made within the web application. In a nutshell, this web application was
designed as a mash-up of Twitter and Craigslist. It was a web-based community that
helped users share personal information, and was enhanced with messaging, social
networking, reputation, content filtering and privacy features form Twitter. A screenshot
is provided below that shows a standard posting.

Figure 5: Screenshot of a Voilai.com posting
In order to create “real-world” user experience, reputation, social matching and
privacy systems were designed. Additional content filtering interfaces were then provided
on the top of these systems. First, a reputation system was developed based on the
number of posts a user made and the number of responses the user’s posts got. If the user
100
authorized the web application to link with her Twitter account, the reputation score
would also be combined with the ratio of the number of her followers and the number of
people she followed. All users’ reputation scores were then normalized and converted
into ranking percentages, indicating the overall reputation standings of the users relative
to the whole community. The idea was to calculate any users’ reputation based on their
past contributions to the community in terms of both quantity and quality (Farmer &
Glass, 2010). For example, the user “symbolism” (pseudonym) has posted 13 personal
needs and his needs had received 11 responses in total. These activities were converted to
a reputation score of 9, which was higher than the scores of 82.5% of all users. As a
result, her profile would show such words “symbolism: Voilah reputation ranking
82.5%”.

Figure 6: Screenshot of Reputation Features
Second, a social matching system was developed in which the posts were
organized into different categories and “interests” based on the hashtags and keywords in
them, such as “apartment to sublease”, “need advice”, “#CS101”, etc. For example, posts
101
that contain “apartment to sublease” would be put into the category “Housing”, and posts
that have “#CS101” would be organized by the interest “Coding”. These categories and
interests were counted and compared with users’ own posts, and then presented to users
as ordered items in a navigation drop-down list.

Figure 7: Screenshot of Social Matching Systems
Third, a simple privacy system was designed. This privacy system allowed users to
configure if their own profile and content should be visible to friends, followers or just
strangers and if they wanted to view content from these different ties. Friends in this
context referred to symmetrical ties. For example, if user A followed user B on Twitter,
the relationship from user A to user B was called “follower”. If user B also followed user
A, the relationship from A to B then became “friend”.
102

Figure 8: Screenshot of Privacy Systems
Finally, an interface was designed to display to the users as soon as they logged in
to the application. This interface asked the users to set the filters for sharing based on the
contributors’ reputation scores, similarity in interests and relationships to the user. For
example, a user could choose to see only content from contributors whose reputation
ranking were at least 50% (Good), shared at least 50% of the same interests (Sharing
some interests), and were at least the followers of the user. The purpose of this interface
was to make sure the design of reputation, social matching and privacy features was
visible to all users. It also emphasized that the use of this web application was
considerably configured by these features.
103

Figure 9: Screenshot of the Filtering Interface
The manipulation of the geolocation feature was implemented by creating two
versions of the same web application. Voilah.com was the version without geolocation
functionalities. This means that the query to the Twitter API did not specify a geographic
location. When users made posts on Voilah.com, their locations were not captured,
recorded or displayed along with their posts. Voilai.com was exactly the same as
Voilah.com, but had geotagging functionalities. In specific, a user’s geographic location
was automatically detected from either the IP address or the GPS coordinates, depending
on the device the user used to visit Voilai.com. This location information was then
attached to the query to Twitter for searching tweets in the categories of offers or needs.
As a result, for example, a user from Los Angeles, California was greeted with a message
at the top of the page, “See what people are sharing within 25 miles of Los Angeles, CA”.
104
Furthermore, all the tweets and posts made from Voilai.com were also attached with a
location tag to indicate where these tweets or posts were created. When the user replied to
any of these posts, his or her own physical location was automatically attached to her
responses. Finally, any user could also change the location and radius for displaying
offers and needs.

Figure 10: Main Interface of Voilah.com
105

Figure 11: Main Interface of Voilai.com
It took a Danish freelance programmer and this author about nine months to build
Voilah.com and Voilai.com. The development of the two websites cost about $4,000 in
grant and personal funding. Significant programming efforts were made to ensure that the
appearance and functionalities of both sites were similar either accessed from mobile
devices or desktop computers. In specific, geolocation features were implemented by
106
setting up a dedicated geospatial database and integrating data with the Google Maps
Application Programming Interface (API), a public database of Internet Protocol (IP)
addresses, and the HTML5 Geolocation API. This ensured that users’ locations could be
accurately captured either with the GPS coordinates from a mobile device or with the
physical addresses associated with IP addresses. A constant connection with the Twitter
API was established so that relevant tweets were continually collected and saved
whenever users refreshed the sites. Reputation, social matching and privacy features were
implemented with scripts in the Python programming language. Based on user actions
from the web browser, these scripts aggregate, filter and display data from dedicated
databases of both users and content. Finally, both Voilah.com and Voilai.com were
deployed on dedicated Linux-based web servers operated by Amazon Web Services,
which had ensured 100% uptime during the study.
Design
The study employed a three-group, one-posttest design (Cook & Campbell, 1979).
Participants were randomly assigned to one of the three groups, Voilah, Voilai and
Twitter. For Voilah and Voilai groups, participants were asked to use the designated
website several times a week for four weeks. For the Twitter group, participants were
asked to continue using Twitter in their regular ways for four weeks. At the end of the
four-week period, all participants were asked to complete a final web survey.
Subjects
94 participants completed the study. Among them, 30 were randomly assigned to
the Voilai.com group, 31 were in the Voilah.com group, and 33 were in the Twitter
107
group. Of these participants (N = 94), 48.9% were male and 51.1% female. The average
age was 23.2 years. Ethnically, 22.2% of the participants were Caucasian/White, 67.8%
Pacific Islander/Asian, 2.2% African-American, 4.4% Latino or Hispanic, and 3.3%
unidentified. About 93.5% of the participants were students. Prior to participating in the
study, the participants had used Twitter for an average of 18.4 months (SD = 13.9).
Procedure
Recruitment advertisements were emailed to three undergraduate classes at the
School of Communication at the University of Southern California. To increase the
geographic and demographic diversity of the participant pool, the advertisements were
also sent to the engineering school at the same university, and were shared on the
author’s pages at Facebook and Twitter. A $20 Amazon.com gift certificate and an entry
in an iPad2 sweepstakes were offered as the incentives for completing the study. To
standardize incentives across groups, participants in the Voilah and Voilai groups were
also told that their chance of winning the iPad2 would be proportional to the number of
posts and responses on respective websites. Participants in the Twitter group were told
that their chances of winning the iPad2 would be proportional to the number of followers
and the frequency of their being listed. Experience with using Twitter was required. This
was because the reputation, matching and privacy features were designed based on the
key features of Twitter such as followers, following, hashtags and the general idea that
people could use Twitter to share personal needs and offers.
Initially, 150 current Twitter users responded to the recruitment advertisements
and signed up for the study. All participants were asked to consent to the study agreement
108
and complete a screening survey, which contained questions on participants’ previous
experience with Twitter and other location-based services, participants’ personal
innovativeness, openness and extraversion, and participants’ basic demographic
information. These personal and dispositional variables were measured as candidate
covariates that may have explained participants’ use of the respective systems in the
study. After the screening survey, 54 participants were randomly assigned to the
Voilah.com user group, 57 the Voilai.com user group, 39 the Twitter group. The initial
assignment was not equal across the groups, because it was expected that the attrition
rates would be higher for the Voilah.com and Voilai.com groups due to the additional
efforts demanded for signing up and using new services.
Participants in both the Voilah.com and Voilai.com groups were asked to sign up
for the respective websites and use them on a regular basis for about four weeks. In the
welcome email sent to participants in the Voilah.com group upon registration, it was
stated that Voilah.com was designed to help users share what they need in life and what
they want to give away or sell. For example, users can share their needs for used
textbooks, vehicles or simply advice. The same language was used for the emails to the
Voilai.com group, but supplemented with words like “near you”, “based on where you
are”, etc. A basic FAQ page was also set up on Voilah.com and Voilai.com for the
participants to understand the purpose of the websites. It was stated to the participants
that they were required and expected to make one or two posts or responses per week, but
there was no penalty if they did not contribute any content. Participants received no other
instructions or tutorials on the specific features of Voilah.com and Voilai.com, with the
109
expectation that they would explore and discover the use and purposes of these websites
in a more natural process. The Twitter group participants were told that they should
continue to use Twitter for personal purposes, and that they should turn off the
geotagging feature.
Two weeks after they completed the screening surveys, a reminder was emailed to
the participants, which contained the link to the “mid-term” web survey. Two weeks after
the participants completed the “mid-term” survey, another reminder was emailed for the
“final” survey. Most participants completed the surveys within two days after they
received the reminders. No more than 10 participants were resent the reminders two days
after the initial reminders, and took the surveys within no more than five days after the
second reminders. Participants were debriefed and thanked at the end of the final survey.
The study employed a naturalistic design in which participants received no
instructions on when, what and how they should use Voilah.com or Voilai.com. This
design intentionally prioritized external validity at the expense of control and internal
validity (Creswell, 2008). For this reason, the usage of Voilah.com and Voilai.com was
low in the first week of the study. To encourage usage and participation, a cash bonus
contest was set up to reward users with the highest reputation ranking based on the
number of their posts and responses. Users were notified about this contest after they
took the second survey. The cut-off date for the contest was set about two weeks after the
estimated end date of the study to avoid an explicit association with the final survey. The
same award was set up for Voilah, Voilai and Twitter groups in the same way, so that all
110
the participants were aware of the cash incentives. This ensured that there was no
difference in incentives and participation process across the groups.
The overall attrition rate for this study was high at 45% for all study participants.
But there was no discernible difference in the attrition patterns across the Voilah and
Voilai groups over time. Among the 54 initial participants in the Voilah group, six did
not sign up for the website (11%), 16 failed to complete the second survey on time
(30%), and one failed to complete the final survey online (2%). Among the 57 members
in the Voilai group, six did not sign up for the website (11%), 18 failed to complete the
second survey on time (32%), and three failed to responded to the invitation to the final
survey on time (5%). Among the 39 initial participants in the Twitter group, three did not
respond to the invitation for taking the screening survey (8%), and three did not complete
the second survey on time (8%).
Material and Measurement
Efforts were made to address the problem of common method variance (Podsakoff,
MacKenzie, Lee, & Podsakoff, 2003). Common method variance refers to the potential
bias from participants’ cognitive patterns or normative preferences during studies that
employ simultaneous measurements of constructs in similar ways. For example,
participants may respond to survey questions in a way that makes them appear coherent
and consistent (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). First, the variables
involved in the main effects models had different wording, framing language and scale
choices. It would have been ideal if the independent and dependent variables could be
measured at different times, so that participants’ dispositional or perceptual patterns
111
would not bias their responses in one setting. However, preliminary analysis indicated
that there were significant changes in several key independent variables about feature use
between the first and second survey (see Chapter Seven for details). Using the
independent variables from the first survey to predict dependent variables from the
second survey, therefore, would not accurately capture the potential change in
participants’ perceived use of the communities of interest over time. On balance, all
variables were measured in the final survey. Second, participants were randomly assigned
to different groups and exposed to the experiment conditions – Twitter, Voilah.com, and
Voilai.com – independently. This ensured that participants avoid the assignment bias,
framing perceptions and actions in consistence with the experiment condition.
The dependent variables of contribution quantity and quality were taken four
weeks after the participants started using the websites, which included both survey
measures and behavioral metrics. The quantity of contribution was measured as the
increase in the frequencies of posts and responses by the participants from the second
week into the study to the end of the study at the fourth week. For Voilah and Voilai
groups, the frequencies were calculated and recorded from the web server records. For
the Twitter group, the frequencies of sharing were recorded directly from the
participants’ Twitter page. As specified in the Informed Consent as part of the consent
process, participants’ activities were recorded on the web server in an unobtrusive way.
Only aggregated data, such as the total number of logins, posts or responses were
collected in the study. This ensured that participants would not be disturbed or influenced
during their use of Voilah.com, Voilai.com or Twitter, so that the threats to internal
112
validity could be minimized (Webb, Campbell, & Schwartz, 1999). The perceived quality
of contribution was assessed with a four-item, five-point reflective scale adapted from
Fulk et al (2004) (Cronbach’s alpha = 0.94). Items in this scale achieved satisfactory
convergent and discriminant validity, as with all other subsequent scales. A table is
provided in Appendix A, which lists the inter-item covariance for all measurement items.
The intermediate variables about social processes were measured in the final
survey taken approximately four weeks after the study started. The construct of
reciprocity was assessed with a five-item, five-point scale combined and adapted from
Kankanhalli et al (2005) and Wasko & Faraj (2005) (Cronbach’s alpha = 0.90). The
construct of social identification was assessed with a seven-item, five-point scale that
combines the cognitive (Tanis & Postmes, 2003), emotional (Wasko & Faraj, 2005) and
evaluative (Luhtanen & Crocker, 1992) dimensions of identification with a community
(Cronbach’s alpha = 0.95). The construct of interpersonal bonds in a community was
measured with a four-item, five-point scale adapted from the study by Prentice et al
(1994) (Cronbach’s alpha = 0.86). The constructs of bonding and bridging social capital
were assessed with an adapted version of Internet Social Capital Scales (ISCS) developed
by Williams (2006) (Bonding: Cronbach’s alpha = 0.92; Bridging: Cronbach’s alpha =
0.91).
The independent variables about the use of features were taken from the final
survey. Although the initial plan was to measure the actual usage of the reputation, social
matching and privacy features, the design of these features made it difficult to generate a
meaningful measure of usage based on the aggregated frequency of change in their
113
settings. For example, the total count of how often a user changed the reputation filter
was not a meaningful measure without knowing the actual change value or considering
the actual value of her reputation “threshold”. The frequency of a use’s configuration of
his privacy settings alone did not reflect his overall preference for communication
boundary. This was an unexpected design flaw that will be addressed in follow-up
studies.
In this study, instead, the use and awareness of various features was measured with
the formative and reflective survey items that assessed participants’ awareness of and
reliance on the design intention and actual functionality of those features. The use of the
reputation system was measured with a four-item, five-point scale that combines an
adapted scale in Ma & Agarwal (2007) for assessing the use of reputation metrics
(Cronbach’s alpha = 0.85). The use of social matching systems was measured with a
three-item, five-point, formative scale on the self-reported frequencies of using tags to
explore people, content and communities of similar interests, based on the propositions
from Terveen and McDonald (2005) (Cronbach’s alpha = 0.85). The use of privacy
systems was assessed with a three-item, five-point reflective scale about the awareness of
and reliance on privacy controls in the community (Cronbach’s alpha = 0.79).
To exclude potential influence of previous use experience on the actual quantity of
sharing, participants’ current activeness in using Twitter and their personal openness were
used as covariates. The current activeness was measured with a single question “On
average, how many tweets did you send in a day”. Because technology adoption studies
have widely supported the relationship between personal innovativeness and the
114
acceptance and use of technologies (Agarwal & Prasad, 1998; Bock, et al., 2005; Lu,
Yao, & Yu, 2005; Roy & Ghose, 2006), personal innovativeness in using information
technologies was also measured with a four-item, five-point scale (PIIT) proposed by
Agarwal and Prasad (1998) (Cronbach’s alpha = 0.87).
The measurement items for each of the latent independent variables are provided
in Appendix B. The descriptive statistics for all variables used in the analyses is also
provided in Appendix B.
Analysis
The hypotheses were tested with the statistical procedure of Partial Least Squares
(PLS). Developed as a variance-based approach to assessing complex structural
regression models, this procedure is widely used in research on the use and perception of
information technologies in the fields of business, marketing and management
information systems (Chin, Marcolin, & Newsted, 2003; Tenenhaus, Vinzi, Chatelin, &
Lauro, 2005). The PLS procedure has no distributional assumptions about data. It is also
tolerant of relatively small sample sizes that would have made it impossible to use
covariance-based structural modeling techniques such as Structural Equation Modeling
(SEM). The most important advantage of PLS for this study is that it allows for the
simultaneous test of both the structural model, which lays out the interrelationships
between the constructs, and the measurement model, which examines the reliability and
validity of the items used to measure the constructs. For the measurement model,
reliability, convergent validity and discriminant validity can be easily generated. For the
structural models, the coefficients and significance levels of multiple paths can be
115
simultaneously assessed. PLS is therefore an efficient and effective statistical tool for this
study. The PLS analysis was conducted with a widely used PLS software package called
SmartPLS (Ringle, Wende, & Will, 2005).
Four steps of analysis were conducted. First, the PLS algorithm procedure was run
to generate variance and covariance statistics for each latent variable for the combined
dataset (N = 94). Reliability of each construct was assessed with the scores of Cronbach’s
Alpha and Composite Reliability. Convergent validity was assessed by examining the
statistics in the cross-loadings matrix. Discriminant validity was examined by comparing
the Average Variance Extracted (AVE) score of each construct and the latent construct
correlations statistics. The criterion for acceptable discriminant validity was that the AVE
score of each construct should be higher than its correlations with all other constructs
(Chin, 1998). This procedure also generated the overall R
2
values for each endogenous
variable, which indicated the portion of variance in the endogenous variables explained
by the exogenous variables.
Second, to examine the main effects of the use of features on the associated
constructive processes as well as the main effects of these processes on participation
quantity and quality, the Bootstrapping procedure was performed to test the means,
standard errors and t statistics of the hypothesized direct paths between constructs in the
entire study sample (N = 94). The statistical significance of each path was tested by
checking its t-value at the significance level of p < .05 with the total sample size as the
degrees of freedom. For example, Hypothesis 1 stated that the use of reputation systems
has a positive influence on the perceived reciprocity in the community. This was tested
116
by examining the significance, direction and value of the standard coefficient for the path
from the latent variable of the use of reputation systems to the latent variable of the
perceived reciprocity.
Third, moderation effects were tested following the suggestions from Carte and
Russel (2003). Having running the main effect tests for the entire sample, Box’s test was
first performed to reject the null hypothesis that the two groups – the non-geolocation
group (pooled sample of Voilah and Twitter groups) and the geolocation group (the
Voilai user group) – shared equal variances for all variables. Then, path coefficients were
compared between the two groups following the guidelines from Chin (2004).
Furthermore, the statistical significance of the group difference in each path was assessed
with a standard t-test equation for comparing coefficients between groups of unequal
sample sizes (Kutner, Nachtscheim, Neter, & Li, 2004). This technique was used to test
the moderation effects laid out in chapter two, three and four. For example, Hypothesis 2
stated that the influence of the use of reputation systems on perceived generalized
reciprocity would be stronger when geolocation features are used in a community. This
was tested by comparing the value of the standardized coefficients for the path from the
latent variable of the use of reputation systems to the latent variable of reciprocity
between the group of participants who did not use geolocation features and the group that
used them.
Finally, to test hypotheses 16a and 16b, an independent-samples t-test was
performed. This test compared the means of the increases in the contribution quantities
and the perceived contribution quality between the group with no exposure to geolocation
117
features and the group that has been using such features. The research question was
assessed by examining the values and statistical significances of paths from the
constructive processes to the variables of contribution quantity and quality across the
groups of Twitter, Voilah.com, and Voilai.com.
A diagram is provided below to illustrate the structural model that combines the
inter-relationships proposed in Chapters Two through Five.

Figure 12: Theoretical Model
118
Chapter Seven: Results
Manipulation Check
Main Effect. It was assumed that the study participants used and relied on the
reputation, social matching and privacy features provided in the respective sharing
systems over time. This way, it was possible to establish that certain social processes
could be linked to these features. To assess this assumption, a paired-samples t-test was
performed to test if participants’ perceived use of different community features changed
over the course of study. For each survey, the values for each item in the scale for the use
of different features were summed up and used as the indicators for feature uses. The
results suggested that in the two weeks between the two surveys, there were statistically
significant increases in the perceived use of reputation systems (d = 0.13, p < .05) and a
significant decrease in the perceived use of privacy features (d = -0.18, p < .05). There
was no significant change in the perceived use of social matching systems (d = -0.07, p =
.34).
The overall usage of the study websites (Voilah.com and Voilai.com) was not very
high. In the four weeks of the study period, participants in the Voilah.com group posted a
total of 260 messages about their needs and offers of things like apartments, used
textbooks or general advice. Participants in the Voilai.com posted 191. However, a
paired-samples t-test showed that between the first and second surveys, there was a
significant increase in the number of postings for participants across the two groups (d =
3.26, p < .05). In comparison, participants in the Twitter group also showed a significant
increase in the number of postings in the same time period (d = 33.9, p < .01).
119
Overall, there was evidence that although the overall usage rate was low for the
customized community websites in the study, participants did increase their participation
across different systems over time during the course of the study. This increased
participation was also accompanied by their changing awareness and use of different
community features. This finding provided a practical basis for understanding and
interpreting the main and moderation effects specified in the second through fifth
chapters of the study.
Moderation Effect. There was a significant difference in the response to the
question “How often have you used the geotagging feature to filter content?” between the
geolocation group and the non-geolocation group, which comprised the pooled users of
Voilah.com and Twitter (mean for geolocation group = 2.62, mean for non-geolocation
group = 1.32, t (89) = 5.99, p <. 01). The manipulation of the location awareness stimulus
was therefore considered successful.
As the study also examined the moderation effect of geolocation awareness on the
relationships between the use of features and the perceptions of communities, it was
necessary to first establish that the geolocation and non-geolocation groups did not share
equal variances across all relevant variables. This was checked with a Box test of equality
of covariance matrices (Carte & Russel, 2003). The results rejected the null hypothesis
that the two groups shared equal variances for all latent variables, Box’s M = 99.67, F
(55, 11374) = 1.55, p <. 01. It was therefore possible to compare the path coefficients
between the two groups in order to discuss the moderation effect of geolocation
awareness.
120
Reputation Systems, Reciprocity and Location Awareness
H1 stated that the use of reputation systems is positively related to the perception
of reciprocity. This hypothesis was supported by a positive and significant path between
the latent variables of the use of reputation systems (URS) and perceived reciprocity
(GR) across all groups (path coefficient (ß) = 0.38, SD =0.09, p <. 01). R
2
for the
influence of URS on GR was 0.145, indicating that URS accounts for about 14.5% of the
variance in GR.
H2 stated that when geolocation features are used, the use of reputation systems
would have a stronger influence on perceived reciprocity. This hypothesis was supported
by the comparison of R
2
values and path coefficients between the Voilai group and the
pooled samples of Voilah and Twitter groups. In the Voilai group, R
2
for GR was 0.38
compared to 0.05 in the pooled group of Voilah and Twitter participants. The difference
in the path coefficients between these two groups was statistically significant (difference
(d) =0.38, t (92) = 1.97, p < .05). The results supported the idea that in online
communities, the use of reputation systems is indeed positively associated with the
perception of reciprocity. When people use geolocation features to tag and share
information based on their locations, the use of reputation systems seems to have an even
stronger effect on the perception of reciprocity.
Social Matching Systems, Social Identification, Interpersonal Bonds and Location
Awareness
H3 and H4 proposed that the use of social matching systems is positively related to
both perceived interpersonal bonds (H3) and social identities (H4) in online communities.
H3 was supported by a significant and positive path between the use of social matching
121
systems (USMS) and perceived interpersonal bonds (IB) (ß = 0.43, SD = 0.08, p < .01)
across all groups. H4 was also supported by a positive and significant path between
USMS and perceived social identity (SI) (ß = 0.34, SD = 0.09, p < .01). R
2
for IB and SI
was 0.18 and 0.11 respectively, indicating that the use of social matching systems had a
weak influence on both perceived social identities and perceived interpersonal bonds
across the user groups of Twitter, Voilah.com and Voilai.com.
H5 posited that community members’ perceived identification (SI) with the
community and perceived interpersonal bonds (IB) should be correlated. Because it was
impossible to test cyclic paths in SmartPLS, the correlation between interpersonal bonds
and social identification was assessed by checking the significance and value of the two
inter-paths separately. The results showed that IB had a positive and significant influence
on SI with a community (ß = 0.80, SD = 0.07, p < .01), and vice versa, SI had a
significant influence on IB (ß = 0.74, SD = 0.09, p < .01). H5 was therefore supported.
H6 and H7 stated that the use of social matching systems like tags and hashtags
has even a stronger influence on interpersonal bonds and social identities when
geolocation features are used. R
2
for IB was 0.26 in the geolocation group, whereas it is
0.06 in the non-geolocation group. The difference in the path coefficients from USMS to
IB between the two groups was statistically significant (d =0.28, t (92) = 2.21, p < .05).
In contrast, in the geolocation group, R
2
for SI was 0.088 whereas it was 0.14 in the non-
geolocation group. The difference in the path coefficients between these two groups was
statistically not significant (d =-0.08, t (92) = -0.64, p > .05). Therefore only H6 was
supported by the results. Overall, the results suggested that the use of matching systems
122
like tags is positively associated with both interpersonal bonds and social identification in
online communities. When location awareness is introduced, the use of discovery and
recommendation features has a stronger effect on the perception of interpersonal bonds,
indicating a stronger need for establishing personal connections based on geographic
proximity.
Privacy Controls, Social Capital and Location Awareness
H8 and H9 stated that the use of privacy systems (UPS) is positively related to
perceived bridging social capital (BRI) and bonding social capital (BON) respectively.
The results only provided support for H8. The path between UPS and BRI was positive
and significant (ß = 0.28, SD = 0.11, p < .01), whereas the path between UPS and BON
was not (ß = 0.15, SD = 0.21, p = 0.49). R
2
for BRI was 0.08, showing that the perceived
use of privacy systems had a weak influence on the perceived bridging social capital.
H10 stated that community members’ perceived bonding social capital (BON)
should be positively correlated with bridging social capital (BRI). Again, this was
assessed by examining the inter-paths between BON and BRI. The results showed that
BON had no significant influence on BRI (ß = 0.15, SD = 0.18, p = .39), nor did BRI
have a significant influence on BON (ß = 0.16, SD = 0.20, p = 0.41). Therefore H10 was
not supported by the data.
H11 and H12 stated that when location features are used, the influences of the use
of privacy systems on perceived bridging social capital and on bonding social capital will
be stronger. The path coefficient between UPS and BRI was significant and high (ß =
0.50, SD = 0.20, p < .01) in the geolocation group. However, the difference in this path
123
between the geolocation group and the non-geolocation group was not significant (d
=0.30, t (92) = 1.34, p = 0.18). Because the main effect stipulated was not statistically
significant for either the entire sample (ß = 0.15, SD = 0.21, p = 0.49) or the geolocation
(ß = 0.38, SD = 0.31, p = 0.22) and non-geolocation groups (ß = -0.02, SD = 0.22, p =
0.94) separately, the moderation effect as stipulated in H12 was not significant (d = 0.40,
t (92) = 1.034, p = 0.30). In general, the results supported the idea that the use of privacy
controls is positively related to the perception of the salience of bridging social capital in
online communities, indicating the importance of privacy features for exploring
alternative information and resources among diverse ties.
Information Contribution
Quality. H13a stated that the perceived quality of information contribution
(QUAL) is positively related to generalized reciprocity (GR) in the community. It was
supported by a positive and significant path between QUAL and GR (ß = 0.33, SD =
0.11, p < .01). H14a stated that QUAL is positively related to perceived interpersonal
bonds (IB) in the community, and H15a stipulated that QUAL is also positively related to
perceived social identity (SI). Only H15a was supported by the results, as the path
between QUAL and SI was significant (beta = 0.31, SD = 0.16, p < .05). H16a stated
that QUAL is positively associated with perceived bridging social capital (BRI) while
H17a states that QUAL is negatively related to bonding social capital (BON). H16a was
not supported (ß = 0.03, SD = 0.14, p = 0.82), whereas H17a was supported by a
negative and significant path coefficient (ß = -0.25, SD = 0.10, p < .05). In total, R
2
for
QUAL was 0.47, indicating a moderate influence from the constructs of reciprocity,
124
social identification and bonding social capital on perceived contribution quality. The
results, therefore, suggest that when people perceive a high level of reciprocity and have
a sense of identity within online communities, they will feel motivated to contribute
information of higher quality. Yet their motivation for contributing quality content is
negatively associated with the perceived bonding social capital in the community,
indicating an interesting trend for sharing practical and utilitarian information in
emerging communities without focusing on the experience in existing strong-tie social
circles.
Quantity. The covariate of current activeness in using Twitter had a negative and
statistically significant effect on the increase in information contribution over time (ß = -
0.10, SD = 0.04, p < .01). The other covariate, personal innovativeness (PIIT), had no
significant influence on the increase in information contribution over time (ß = 0.15, SD
= 0.11, p = 0.16).
H13b, H14b, H15b, H16b and H17b respectively stated that the increase in
information contribution over time (QUAN) is positively associated with perceived
reciprocity (GR), interpersonal bonds (IB), social identity (SI), bridging social capital
(BRI), and negatively associated with bonding social capital (BON). After excluding the
influence of the covariates, only H15b was supported with a positive and significant path
between SI and QUAN ((ß = 0.46, SD = 0.23, p < .05). R
2
for QUAN was 0.15,
indicating that social identification process has a small influence on the increase in
information contribution over time.

125
Group differences. H18a and H18b stated that the perceived quality and the
increase in the quantity of contribution would be higher in a community when
geolocation features are used. These hypotheses were not supported by the data. Because
there were notable differences in the usability and focus for using Twitter and the custom
websites, the comparison in the increase in contribution quantities was also made
between the users of Voilah.com and Voilai.com. The average increase in contribution
quantity over time was 19.55 (SD =53.40) in the pooled non-geolocation group
(Voilah.com and Twitter), significantly higher than 2.20 (SD = 3.54) in the geolocation
group (Voilai.com) (d = 17.35, t(64) = 2.59, p < .05). But the average increase in
contribution quantity was only 4.29 (SD =15.76 ) in the Voilah.com group, no
significantly higher than that in the Voilai.com group (d = 2.09, t(33) = 0.72, p = 0.48).
The perceived contribution quality was 3.09 (out of a 5-point scale) (SD = 0.96) in
the non-geolocation group, lower than 3.26 (SD = 1.09) in the geolocation group, but the
difference was not statistically significant (d = -0.17, t(51) = -0.74, p = 0.46). Post-hoc
analysis, however, revealed that the change in the perceived contribution quality over
time was statistically significant (d = -0.40, t(55) = -1.97, p < .05), with an increase of
0.31 (SD =0.97) in the perceived contribution quality in the geolocation group and a
decrease of 0.09 (SD = 0.93) in the non-geolocation group.
RQ1 asked about the potential differences in the contribution patterns across
different systems. The main issue was what social processes were more salient in
different communities where the foci and the use of geolocation features were different.

126
Difference in the shared focus. This comparison was done between the Twitter
group and the combined groups of Voilah and Voilai. For the Twitter users, the perceived
quality of content contribution was only positively associated with bridging social capital
(BRI) (ß = 0.37, SD = 0.17, p < .05), and perceived social identity (SI) (ß = 0.52, SD =
0.17, p < .01). The increase in the quantity of contribution as measured by the number of
tweets was only positively related to the perceived bonding social capital (BON) (ß =
0.41, SD = 0.11, p < .01).
In contrast, in the Voilah and Voilai groups where users participated in a
customized, focused community for sharing personal needs and offers, the perceived
contribution quality was positively associated with generalized reciprocity (GR) (ß =
0.33, SD = 0.12, p < .01), but negatively related to bonding social capital (BON) (ß = -
0.24, SD = 0.12, p < .05).
Difference in the use of geolocation features. When the two communities
Voilah.com and Voilai.com shared the same code, features and foci, patterns differ in
content contribution when geolocation awareness was available. In the Voilah group,
perceived quality of contribution was only positively associated with reciprocity (GR) (ß
= 0.30, SD = 0.11, p < .01). In the Voilai group with geotagging features, in contrast, the
quality of contribution was negatively associated with bonding social capital (BON) (beta
= -0.29, SD = 0.09, p < .01). And the increase in the quantity of contribution over time
was positively associated with the perceived social identification (SI) (beta = 0.44, SD =
0.19, p < .05).
127
The results for the structural model are illustrated in the diagram below, and listed
in the table below.

Figure 13: Overall Structural Model Results
128
Table 1: Test Results for Supported Hypotheses
Hypothesis Coefficient
t-Statistic
(degrees of
freedom)
Standard
Error
Significance
H1: Community members’ use of reputation
systems that implement objective and relevant
measures of past behaviors will have a
positive influence on their perception of
reciprocity in the community
0.38 4.06 (93) 0.09 p < .01
H2: The influence of the use of reputation
systems on perceived generalized reciprocity
in an online community will be stronger when
geolocation features are used.
0.38 1.97 (92) 0.19 p < .05
H3: Community members’ use of social
matching systems with relevant tags has a
positive influence on their perception of
interpersonal bonds in the community.
0.43 5.33 (93) 0.08 p < .01
H4: Community members’ use of social
matching systems with relevant tags has a
positive influence on their perceived
identification with the community.
0.34 3.80(94) 0.08 p < .01
H5: Community members’ perceived
identification with the community is positively
correlated with their perceived interpersonal
bonds in the community.
0.74/0.80
13.9 (94) / 12.0
(94)
0.05/0.07 p < .01
H6: The influence of community members’
use of social matching systems on their
perceived interpersonal bonds in the
community will be stronger when location
features are used.
0.28 2.21(92) 0.13 p < .05
H8: Community members’ use of privacy
controls that configure the disclosure of
personal information to strangers or friends
will have a positive influence on their
perception of bridging social capital in the
community.
0.28 2.51 (94) 0.11 p < .01
H13a: The quality of individuals’ contribution
will be positively related to their perception of
generalized reciprocity in an online
community.
0.33 3.05 (94) 0.11 p < .01
H14a: The quantity of community members’
contribution will be positively related to their
perception of interpersonal bonds in the
community.
0.31 2.07 (94) 0.15 p < .05
H14b: The quantity of community members’
contribution will be positively related to their
perception of interpersonal bonds in the
community.
0.46 2.00 (94) 0.23 p < .05
H17a: The quality of individuals’ contribution
will be negatively related to their perception
of bonding social capital in a community.
-0.25 2.42 (94) 0.10 p < .05
129
Chapter Eight: Discussion
Summary
This study aims to produce a balanced account of how common features in online
communities influence some of the social processes in online communities, and how the
use of geolocation features like geotags further reinforce such relationships. The
assumption is that geolocation features need to be incorporated into the use of
fundamental features that configure the structure of mutual engagement, communication
and boundaries in a community. The use and awareness of such fundamental features is
essential for users to generate positive perceptions of the community and participate in
the sharing of relevant information with others. Geolocation features promote these
constructive social processes, because they configure the perception of geographic
proximity and influence the scope and intensity of communication. The results provide
mixed support for the hypotheses in the study. Overall, the associations between common
features in online communities and constructive social processes are shown to exist
among all groups. These associations appear to be stronger among the users of location-
aware systems than the other users. Reciprocity, social identification and interpersonal
bonds do seem to influence the perceived quality and actual increase in the quantity of
information contribution, but such influences tend to vary in different community
systems.
In specific, the use of reputation systems – systems that record, aggregate and
display past activities in a objective and systematic way – is positively associated with
perceived reciprocity in both general-purpose communities such as Twitter and the more
130
narrowly focused, emergent communities of Voilah and Voilai. This association means
that people who pay attention to and care about the reputation metrics in their own and
others’ profiles may have a stronger sense of reciprocity about their communities. In
other words, the users of reputation features tend to believe that reputation systems
reinforce reciprocation and recognition for sharing personal information. When
geolocation information is attached to users’ profiles, the use and awareness of reputation
information has an even stronger influence on perceived reciprocity. This reinforcement
effect of location awareness indicates that reputation systems have even a more important
role in shaping people’s perception of an equitable and reciprocal environment when
geotags are used to filter relevant content and identify users.
Recommendation and suggestion features based on tags and hashtags can help
users exploring content of common interests and identifying people who share such
interests. They are therefore considered as the basis of social matching systems in online
social networks (X. Li, et al., 2008). In the study, the use of such systems proves to be
positively related to perceived identities and interpersonal bonds across different
communities. This result suggests that by showing users the content they have interests in
and by connecting with people who share such interests, online communities can support
the development of both collective identities and interpersonal bonds (Ren, et al., 2007).
This observation is also supported by the mutual influences between the perceived
interpersonal bonds and identification. This indicates that these two fundamental
perceptions of the attachment to a community can mutual reinforce one another in
community members’ use of features for identifying common interests and agenda.
131
In a geolocation-aware community, furthermore, the link between the use of social
matching systems and the perceived interpersonal bonds is even stronger, showing that
interpersonal interaction is more important for users to grow attached to a location-aware
community. In contrast, geolocation-aware communities do not manifest a more salient
tendency for using tags to develop strong collective identities. In other words, geotagging
– or rather, geographic proximity given the shared locations for most of the participants
in the study –does not provide a supplemental cue for social identification above and
beyond the shared foci, objectives or interests that sustain a community.
The use of privacy systems is hypothesized to indicate a person’ perception of the
social benefits associated with the social interactions with different ties. In specific, the
awareness and use of privacy controls might be related to the perception of bonding
social capital, the bonding experience of emotional comfort and security, and bridging
social capital, the potential for accessing new opinions and alternative perspectives
(Williams, 2006). However, the hypotheses with regard to privacy and social capital
receive poor support from the results. The use of privacy controls is only positively
related to perceived bridging social capital, which indicates users’ perception of the
resources embedded across different circles of strong ties. This shows that the more often
users configure their social boundaries, the more likely they are aware of the potential
benefits of engaging with hidden or weak ties, such as new information or new
perspectives (Burt, 2002). However, there is no significant link between the control of
group boundaries with privacy features and bonding social capital, indicating that the
132
awareness and use of privacy settings do not influence users’ perception of bonding
social experiences in their strong-tie circles.
In general, the results support the idea that the perceived quality of information
contribution is driven by perceived reciprocity and social identity across different
community systems. This finding suggests that a systematic belief in generalized
reciprocity and a strong sense of collective identities can motivate individuals to
contribute quality content. In contrast, bonding social capital that is characterized by
emotional support from normally strong ties is negatively associated with perceived
contribution quality. This indicates that the motivation for sharing relevant information
about personal needs may be based on the desire to avoid interaction with strong ties. Put
in another way, contribution quality is negatively associated with social experiences of
the “echo chamber” perception, where interaction with existing close ties reinforces
sometimes wrong assumptions and ideas (Burt, 2007).
Yet only the perception of social identification in the community has a
significantly positive influence on the increase in the quantity of contribution across the
groups. This suggests that across different community systems that facilitate the sharing
of personal information, it is the gradual development of common purpose, foci, and
goals that promote constant participation over time. The social meaning of sharing and
communication about relevant information for emergent groups seems to lie in the
discovery of a common identity.
Furthermore, the contribution motivation appears to vary by the shared purpose of
a community. For example, participants in the Twitter group tend to think that the quality
133
of their contribution is driven by the access to bridging social capital in an open
community, but the actual increase in their sharing frequencies is predicted by the
expectation of getting bonding social capital with their strong ties. In more narrowly
focused communities like Voilah and Voilai, perceived contribution quality has more to
do with generalized reciprocity than with social capital, reflecting an emphasis on trustful
and rewarding experience in such emergent communities. The actual quantity of
contribution in such communities also has more to do with a stronger identity derived
from the shared focus, rather than interpersonal bonds.
When geolocation features are used, the contribution patterns also differ. In the
geolocation-aware community of Voilai.com, the perceived quality of contribution is
negatively related to bonding social capital. This suggests that the quality of users’
contribution is not so much driven by the preference for close and intimate social
interactions when geolocation features afford users the capacity for exploring new ties
and information in their neighborhoods. The actual quantity of contribution complies
with this trend, in that it is positively related to the perceived bridging social capital and
perceived interpersonal bonds based on shared interests.
Reputation Systems and Reciprocity
The meaning of reputation systems. The study attempts to extend the theory of
reputation and reputation systems in online communities that focus on information
sharing rather than e-commerce. Numerous studies have documented how the reputation
of a responsive and responsible seller or an appreciative buyer influences transaction
outcomes on e-commerce sites such as eBay (Bolton, et al., 2004; Resnick & Zeckhauser,
134
2002; Resnick, et al., 2000). Research on cooperation in early online communities also
reveals the impact of reputation on social interaction and online collaboration (Kollock,
1999b). The shared focus of these studies is the translation of the concept of reputation –
as what is generally believed or said about someone’s characteristics – in the online
environment. The concept of reputation systems reflects scholars’ efforts at explaining
the contextualized meaning of aggregated and personalized histories for bilateral
interactions in online communities (Zhou, et al., 2008). Reputation as the indicator of
trustworthiness, competence and credibility based on past behaviors can have a powerful
influence on transaction partners’ willingness to participate in future engagements
(Bolton, et al., 2004).
The idea of reputation as a “shadow of the future” can be extended into other types
of online communities, in which the focus might not be necessarily in commercial
transactions or cooperation that result in tangible outcomes. Yet another necessary shift is
the expansion of reputation metrics to include publicly available, objectively verified and
automatically collected activity statistics (Oulasvirta, et al., 2007). Evidence shows that
since people often unconsciously give away nonverbal cues in their use of
communication technologies, their activity patterns can serve as valid and effective
indicators for their attitudes, perceptions and intentions (Pentland, 2008). Therefore
feedback does not have to be explicitly solicited and created; it can be inferred from
aggregated behaviors. If following a person on Twitter indicates a form of endorsement
of relationships, recognition or affection, then the aggregated numbers of followers for
anyone should reflect what other people think about the relevant attributes of this person
135
as a Twitter user, such as credibility, humor, informedness, or popularity (Morris, et al.,
2012).
This is the logic behind the design and use of reputation systems in this study
based on activity metrics relevant to Twitter, Voilah.com and Voilai.com respectively.
Reputation systems must incorporate the systematic aggregation of people’s activities as
the implicit yet objective measure of one’s reputation (Farmer & Glass, 2010). Although
this concept has been widely adopted by communities like Twitter and startups like Klout
(Landman, 2011), academic research has yet to theorize and incorporate activity metrics
into the design and study of reputation systems in online social networks (c.f. Tong, et al.,
2008). This study implemented a rudimentary design of reputation systems based on this
concept and measured participants’ perceived use of these systems. The consistently high
reliability and validity values of the measurement across different community systems
shows the potential for extending the concept of reputation systems in more broad and
current online communities.
The meaning of generalized reciprocity. The technology that automatically
records and aggregates behavioral metrics at a community level makes it more efficient
and more effective to build a reputation system that is stronger against intentional
gaming, a criterion for successful reputation systems design (Zhou, et al., 2008). For
example, if the numbers of recommendations, forwarding and responses are considered
as part of a user’ s reputation, then the user would not be able to manipulate the system
by making many low-quality posts that would not easily get recommended, forwarded
and replied. As a result, the concept of reciprocity becomes even more relevant: the
136
exchange of benevolent acts cannot be hidden or lied about. Indeed, this study
demonstrates that the perceived use of reputation systems is positively associated with the
perceived degree of generalized reciprocity across different community systems.
Generalized reciprocity has been well researched in e-commerce and social
network research (Resnick, et al., 2000; Shumaker & Brownell, 1984). Yet there has been
relatively little effort at theorizing generalized reciprocity in online communities. This
might be because reciprocity was difficult to reinforce with the lack of persistent
identities (Kollock, 1999a). As a matter of fact, many normative behaviors in online
communities can be best described as the consequences of individuals perceiving and
believing in a certain degree of reciprocity in their communities (Kankanhalli, et al.,
2005). The perception of reciprocity is necessary for benevolent participation in
community actions to be visible, acknowledged and rewarded, and malign actions such as
disruption and free-riding to be exposed and reprimanded (Resnick, et al., 2000). This
study makes a unique contribution by revealing reciprocity as the key psychological
outcome for the use of community-wide reputation systems in online sharing
communities. Most notably, it uncovers the socio-technical practices around the use of
and reliance on systematic aggregation and display of non-obtrusive, objective and
comprehensive behavioral traces of community participants. More and more information-
sharing-oriented communities such as Quora and StackOverflow are implementing such
reputation systems as key features that stimulate participation (Kincaid, 2011; Rivilin,
2011). Generalized reciprocity at a community level fosters mutually accountable
behavior, and facilitates a reward structure that satisfies people’s intrinsic needs for
137
recognition and appreciation. This is a key insight for researchers and community
operators to develop effective participation and reward mechanisms that encourage
participation.
Building reputation systems online. As personal identification becomes
increasingly easy and prevalent in current online social networks, reputation systems no
longer simply serve as a security check for social interactions among users. Reputation
systems can be foundational to understanding people’s motivation for participation and
engagement with a given community. For example, effective reputation rewards and
virtual credit systems have proven essential for the success of emergent knowledge
sharing communities like Quora (Rivilin, 2011). While it is not necessarily the job of
researchers to design reputation systems, the knowledge of what constitutes valid
reputation information and how people think about reputation systems is essential for
understanding communication and collaboration online (Farmer & Glass, 2010).
The integration of objective and automatic aggregation of people’s past activities
with reputation systems presents researchers with new opportunities for understanding
the specific context and meaning of participation in online communities. The concept of
“honest signals” helps us understand the implication of this integration of passive and
active feedback given to people’s everyday activities. Honest signals refer to the
expressions, cues, and all kinds of traces people leave in their everyday social interaction
(Pentland, 2008). When such signals are aggregated into meaningful analytics, users can
construct vivid and reliable histories about one another. As a result, users will have a
strong motivation to contribute relevant content. At the same time, there will be efforts at
138
transparent and equitable distribution of recognition on the community operators’ side.
With such detailed activity histories, users’ focus on making relevant contribution is
increasingly driven by the perception of reciprocity at the community level. The activity
histories will become meaningful numbers that are associated with recognition and
rewards. The pursuit of such quantified reputation based on recognition and mutual
appreciation therefore can drive user participation. By thinking about the design and
meaning of different reputation systems online, communication researchers can therefore
gain a deeper understanding of what specific activities people do in online communities
and why.
Tags, Social Matching Systems and Community Attachment
Social matching systems in practice. This study presents a unique effort at
integrating the popular features of tags in online communities with the framework of
social matching systems (Terveen & McDonald, 2005). It shows that the use of
recommendation and suggestion features based on hashtags on Twitter has a positive
influence on the sense of collective identity and the perception of interpersonal bonds in
different communities. This finding has three implications.
First, the study combines the previous studies on recommendation systems and
social matching systems to explicate the socio-technical practices surrounding the use of
tags for identifying potentially relevant connections and content (X. Li, et al., 2008).
While tags were perceived to be one of the key technical characteristics for Web 2.0,
previous research mainly focused on their use in the visual and semantic categorization of
user-generated content (e.g. Breslin & Decker, 2007). Anecdotal evidence, however,
139
suggests that tags – particularly hashtags on Twitter – not only provide semantic meta
data about user-generated content; tags also become important tools for users to quickly
identify the current trends and foci in the community and explore social ties based on
shared interests (Levy, 2009). Tags therefore serve as the basis of social discovery
features that structure meta-data of content into meaningful indicators of interests and
preferences (X. Li, et al., 2008). In other words, tag-based recommendation and filtering
systems can have interesting social consequences for online communities, allowing for
users to identify relevant content and build interpersonal connections. Second, the study
reveals the significance of social matching features for the perception of collective
identification with online social networks. Previous research established that collective
identities are necessary for social groups to sustain a sense of shared purpose and develop
social cohesion among their members (Hogg & Turner, 1985). Early online communities,
because of the lack of communication richness and identification mechanisms,
particularly relied on the development of group identities to become meaningful and
engaging social environments (Postmes, 2006; Rogers & Lea, 2005). Yet there has been
little discussion of the specific technologies that support the recognition and sustaining of
social identities online beyond user profiles (Utz, 2008). This study is a preliminary
attempt at revealing the social process that foster collective identification. As processes
such as social categorization and task engagement usually promote identification, the use
of tags for recognizing the common interests and foci helps individuals develop
alignment with and attachment to what are known as topic-based groups (Prentice, et al.,
1994). Because the tags were always changing over time in the study, there were less
140
tangible boundaries in such tag-based topic groups in online communities (Papacharissi
& Olivera, 2012). In these groups, conversations may be more engaging and inclusive,
because tag-based filtering and recommendation features facilitate the discovery and
persistence of the focus of conversations. Openness, democratic communication, and
efficient organization of topic-centric conversations all contribute to the emphasis and
reinforcement of group identities. Discovering, joining and participating in these groups
become increasingly easy (Shirky, 2008). As a result, it might become more efficient for
individuals to engage with different interest groups at both behavioral and cognitive
levels.
Third, the study reveals that the use of social matching features facilitates the
exploration of social ties based on shared interests. The study confirms the general
awareness of this potential use of tags among online community users. It also illustrates
the positive effect of this awareness on the perceived salience of interpersonal
connections in different communities. Interpersonal bonds are important to emergent
social groups with fluid identities, because they indicate the level of mutual familiarity
and similarity (Prentice, et al., 1994). By using tags to search and identify shared
interests, individuals not only have the opportunity to activate latent ties, but also gain a
better understanding of the similarity or compatibility of interests in existing ties that are
otherwise hidden or hard to know (Haythornthwaite, 2002). In this sense, tags can
become contextual cues for emergent and ad hoc communication about shared interests.
For example, a Voilah.com user can use tags like “used cars” as her main interest. She
was then able to start conversations about potential transactions with both unconnected
141
ties and her Twitter followers. In online communities, such a broader application of tag-
based interest discovery can increase users’ sense of mutual engagement with others.
It is interesting to note that once the path from perceived interpersonal bonds to
perceived identification with the community is added, the influence of the use of social
matching systems on perceived identification becomes statistically non-significant. This
suggests, but not does not necessarily support, the possibility that perceived interpersonal
bonds has a mediating effect on the development of identification with a community
(MacKinnon, 2008). This makes sense because the fundamental affordance of a
recommendation feature is to first connect with like-minded individuals. It is likely that
people increase their identification with a community as they gradually develop personal
liking with more members and identify more common grounds, which makes the
community a more attractive social environment. More careful analysis can be conducted
to specify the detailed relationship between these two important mechanisms for building
communities with the discovery and communication features over time (Ren, .et al,
2007).
The development of online communities. This study assumes that social
networks like Twitter and Facebook are not only de facto online communities, but also
new social phenomena that entail a refreshed look at the classic definition and
operationalization of online communities. Across these groups, people can easily keep in
touch with friends they already know and discover new ties (A. Smith, 2011c). Because
users exchange a high volume of personal information, communication with diverse ties
over broad topics of common interests is a salient characteristic of such communities. For
142
this reason, interpersonal bonds and common identity both become necessary for building
and sustaining online communities. Indeed, results across the study sample suggest that
there is a common positive association between the attachment to an online community
and both perceived interpersonal bonds (r = .69, p < .01) and social identification (r =
.71, p < .01).
Interpersonal connection is the foundation of online social networks (Kumar, et al.,
2010). Compared to a decade ago, the increasing capacity of individuals to connect with a
wider range of people in constant, synchronous and rich communication is arguably the
most obvious affordance of services like Facebook and Twitter (Haythornthwaite, 2010).
It is the aggregation of millions of active lateral exchanges of all kinds of information
that shapes the structure of such massive hubs of personal networks (Golbeck, 2007). Yet
in Twitter and other social networks, shared identities also emerge from the meaningful
aggregation of personal ties, because users gradually discover shared purposes beyond
the simple affordance of sharing personal trivia between friends. Facebook brands itself
as a personal tool for openly sharing information with people that matter (Sengupta &
Miller, 2012). Twitter started out as a tool for keeping friends update with personal
events, but gradually turned itself into a multi-purpose platform for publishing breaking
news to sharing local deals (Papacharissi & Olivera, 2012). LinkedIn is about building
professional relationships, yet is also a tool that implements the founders’ philosophy
about life, work and society (Rowan, 2012). The shared foci and purposes in these
networks may vary in scale, substance and consequence. But the capacity for these
143
networks to accommodate the operation of different social groups is indeed what turns
these social networks into great communities (Shirky, 2008).
Online communities have always been built for exploring and engaging shared
interests (Kollock & Smith, 1999). But communities like Twitter differ from
communities in the early stage of the commercial Internet mainly because of their
capacity for developing parallel micro-networks based on different kinds of foci and
purposes, and their affordance of richer, enduring and inclusive interpersonal connections
(Shirky, 2008). A clear group boundary is usually considered a necessary condition for
collective actions in self-organized communities (Ostrom, 1990). This may not
necessarily be the case for current online communities like Twitter, or at least not the
case for developing social identification with Twitter as a community. The many small
sub-groups formed in online communities depend a great deal on the creation and
sustaining of interdependence among the ad hoc organization of shared interests. The
sense of mutual engagement must come from concerted efforts at creating open
conversations, mandating equitable distribution of resources, facilitating tangible actions
and sharing relevant rewards. Again, such efforts require each sub-group to embrace
inclusive and democratic communication, which is the key to developing interpersonal
affinity and connection, and ultimately a sense of interdependence (Wasko & Faraj,
2005). The findings in this study confirm that both interpersonal bonds and social
identification are associated with an attachment to online communities, and that
interpersonal bonds might be an essential basis for developing community identification.
Understanding this new pattern of community development online is therefore necessary
144
for researchers to uncover the socio-technical practices surrounding the new features
pushed to users with the expectation of increasing the communities’ “stickiness”. The
social glue in these communities, in this sense, consists of both a strong identity based on
shared purposes and interdependence, and interpersonal communication that is rich,
engaging and open.
Privacy and Social Capital
The meaning of privacy online. This study makes an exploratory claim that rather
than a binary choice between showing and hiding one’s information online, privacy is a
dynamic perception that reflects people’s preference for social benefits associated with
their engagement with different social groups. Use and awareness of privacy controls
indicates that a person is aware of the configuration of the boundaries for his or her
communication. Active configuration of these boundaries enables him or her to access
different forms of social capital. Unfortunately, the results fail to fully support this claim.
The use of privacy controls is only positively associated with bridging social capital.
On one hand, the significant path between the use of privacy features and bridging
social capital does suggest that individuals are aware of the broad social meaning of
privacy. Privacy is not simply about safeguarding against unauthorized and unwelcome
access and abuse of personal information (Luo, 2002). Rather, it is also about the
potential benefits of engagement with different social ties. From an ethics perspective,
privacy is indeed about individuals’ active control of benefits as the result of the
disclosure of personal information (Laufer & Wolfe, 1977). People have a sense of
control over the use of their private information. They therefore have a better idea of the
145
pros and cons of disclosing it within or outside their strong ties. The concept of bridging
social capital helps us interpret this socio-technical practice in the study of privacy in
online communities. In interpersonal interactions, the disclosure of private information
becomes a purely personal choice based on the assessment of the likelihood and value for
accessing resources in more diverse social interactions.
On the other hand, the lack of significant relationships between the use of privacy
features and bonding social capital suggests that the meaning of communication
boundaries for strong tie networks needs more careful consideration. The advantage of
accessing bridging social capital is often utilitarian, informational and practical.
Interacting outside structural enclosures of strong ties makes it possible for one to span
structural holes and access alternative perspectives and unique information (Burt, 2007).
In comparison, the advantage of accessing bonding social capital is more intangible and
figurative. The social interaction between strong ties normally provides a sense of
certainty, security, comfort and emotional support (Burt, 2000; Williams, 2006). At the
same time, the sources of bonding social capital may also be more diverse and
complicated. It is therefore likely that the study participants were unable to associate a
specific dimension of bonding social experience with their awareness of configuring
strong-tie group boundaries. For example, the choice between sharing one’s personal
needs with only close friends and sharing them with the whole world probably would not
make the intimate friendship stronger or worse. Yet this choice may have an impact on
whether these needs could be solved by people who are not one’s close friends. In
146
general, this alludes to the fact that communication between existing strong ties tends to
be less vulnerable to a change in communication media (Haythornthwaite, 2002).
There is reason to believe that privacy is related to the boundary condition for
people’s communication. However, this logic suggests that privacy control may only
affect the “filtering out” part of one’s social experience: with privacy control we know
who are not our friends or followers, and consequently who should not be included in our
social interactions. In comparison, we may be less clear about the “filtering in” part of
our social experience in the use of privacy controls. Using privacy controls to keep our
personal information only available to our strong ties is comforting. Yet it is not entirely
clear how our friends, followers and colleagues, who may all fall within the broad social
media terms of “friends” or “followers”, contribute to different dimensions of bonding
experiences. This is in part due to the complicated nature of social bonds in different
strong-tie networks. But it also reflects the complexity of designing and implementing
privacy features: these features must effectively accommodate users’ preferences for
different types of social engagement with their close ties. The lack of correlation between
bonding and bridging social capital in this study suggests that these two types of social
experiences may involve very different processes surrounding the configuration of
communication boundaries. Privacy online entails a much more careful explication and
dissection of the specific social benefits particularly associated with the bonding social
experiences with strong ties.
The meaning of social capital online. This study is one of the first efforts to
integrate the study of online privacy with the concept of social capital (c.f. Ellison, et al.,
147
2011). Although the results only weakly supported its hypotheses, the study uncovers two
important dimensions of the concept of social capital.
First, social capital is embedded in the use of a specific technology that configures
group boundaries, communication patterns and/or resource distribution (Resnick, 2001).
The framework of socio-technical capital stipulates that because human communication
processes are often mediated by technologies, social capital is by default a product of
various socio-technical practices (Resnick, 2001). People often manage personal
information in online social networks as a conscious and intentional choice (Madden,
2012). Therefore, their access and perception of different social capital will be closely
associated with the awareness and configuration of a specific feature that affects the scale
and scope of their activities. Social capital is therefore not just a theoretical and
metaphorical concept that is devoid of concrete, technical specifications. It is the
relational, psychological and informational outcome of the social use of a selected
communication channel, or an intentional choice for adding additional communication
cues. An obvious link therefore exists between the boundary condition for accessing
bonding social capital with strong ties and bridging social capital with weak ties, and
privacy features that control the boundaries for the disclosure of private information with
strong or weak ties. There may be more features that may also affect the group
boundaries for social capital. The discussion of social capital therefore must include its
technical basis and possible feature, be it the use of Twitter, a mobile application or an
online game.
148
Second, social capital is a product of the calculus of benefits in social interactions
across structural clusters. Theorists from Burt to Nahapiet have all argued that the overall
structural pattern of various clusters has a significant impact on the flow of information
and other resources within and across clusters (Burt, 2002; Nahapiet & Ghoshal, 1998).
Structured social interactions are often fluid and diverse. Different forms of social capital
may be available and possible from the same types of ties (Aral & Van Alstyne, 2011).
Still, there is always the issue of awareness and preference for social capital associated
with specific ties. If people intentionally extend their social circles and interact with new
ties, the chance of getting new information may be higher. Similarly, if there is more
intensive communication among existing strong ties, the chance for reinforcing the
bonding social experience will also be higher. The real question is the extent to which
people maintain the balance in time and effort for their social activities. The structural
configuration in this balancing act is vital to understand social capital. Therefore,
formation of social capital is a dynamic social process in which different forms of
benefits are balanced in people’s structured social interactions with strong ties and weak
ties. The discussion of social capital must include the specific structural configurations of
bonding and bridging social experiences.
In the current communication landscape, the dynamic balance between the
widespread concern with privacy and the overarching hype of openly sharing information
is a legal and business challenge for many online social networks (Biersdorfer, 2012;
Sengupta & Miller, 2012). Yet the users also engage in such balancing acts on a daily
basis. It is the outcomes of their acts that perhaps would hold more interest for
149
researchers of privacy and social capital. Researchers should employ a more flexible
perspective on privacy as a dynamic choice between different social benefits, and on
social capital as a structural and technical process. Such a view could potentially be more
effective at explaining the generational or cultural divide in online privacy (Debatin, et
al., 2009; Ellison, et al., 2011; Luo, 2002). For example, the real difference between
teenagers and older individuals in their attitudes towards online privacy may derive from
their differential perception of the benefits of bridging social capital and from their
capability in configuring privacy boundaries with the appropriate features (Madden,
2012; A. Smith, 2011c). The awareness of privacy and social capital as dynamic socio-
technical processes, as shown in this study, may prove useful for generating new
hypotheses and results on this topic.
The Social Meaning of Geocoding
Geocoding as a social “add-on”. As with other new technologies, the use of
geolocation features, or more specifically, geocoding, is configured by the affordances
and constraints of existing social structures (Plant, 2004). For an online community, these
social structures are produced and reproduced in the use practices of features that
configure communication, connection and boundaries. This is a key assumption for
studying geocoding: precisely because this technology does not seem to sustain
meaningful structures and communities, we need to understand what features and
practices do. Therefore geolocation technologies are integrated with other reputation,
recommendation and privacy features. They reinforce some of the constructive social
processes that facilitate information sharing in online communities. This assumption
150
means that geolocation awareness by itself does not necessarily create reciprocity,
connections and social capital. Rather, it increases the effectiveness of reputation systems
in promoting reciprocity, social matching systems in fostering social identification and
interpersonal bonds, and privacy controls in promoting social capital. The results provide
partial support for this assumption.
At the end of the study, participants in the geolocation-aware group did not
necessarily share more information (difference = -2.02, p = .55) or consider their
contribution to be of higher quality than those who did have access to the geolocation
features (difference = 0.17, p = .46). However, in the geolocation-aware group, the use of
reputation systems had a stronger influence on the sense of reciprocity in the community.
Similarly, the use of tags for exploring shared interests also had a stronger effect on the
interpersonal bonds in the geolocation-aware group. This in general supports the idea that
geographic proximity, albeit a regular and remarkable influence on social affinity, must
be incorporated into people’s communication practices in order to become a more
meaningful social experience (Latane & Liu, 1996; Latane, et al., 1995). In other words,
geocoding is the measure of proximity that improves connections and communication
already facilitated by reputation and matching features.
First, geolocation awareness may make reputation systems and social matching
systems more effective at turning a community into an accountable and engaging social
environment. Because location information is attached to each user’s content, it is easier
to verify and confirm the authenticity and relevance of user-created content. For example,
user “Freddie” (pseudonym) on Voilai.com has posted many offers of used textbooks
151
specifically for classes at USC. As a result, this user has accumulated a moderately high
“reputation” for being able to offer relevant content. The location tags show that this
user’s posts have all been made within one mile of the university. The reputation
information for this user therefore becomes more transparent and relevant. Furthermore,
because users can use geolocation features to narrow down the content of interest only at
places they are interested in, it would make their use of the community more efficient.
Social interaction based on this refined commonality of interests will be more endearing
and engaging, because the discovery of common interests is more accurate and relevant.
Co-location as a form of geographic homophily gives like minds more opportunities to
start conversations and build relationships. Simply put, geolocation awareness makes the
activity metrics in reputation systems more authentic and relevant. Proximity, in this
sense, is a unique contextual cue that validates and filters information in people’s social
interactions in a community.
Second, geolocation awareness may make it even more important for a community
to design and implement effective reputation and matching systems. It is relatively simple
to add geolocation features to a community, so that each user’s activity is supplemented
with location information. Yet geolocation information is an additional contextual cue
that needs to be processed and understood by users. Moreover, the disclosure of users’
location data also presents potential risks to their privacy and personal security. Because
it is now possible for users to configure geographic proximity in their social interactions,
users should expect to a more inviting, relevant and meaningful social environment. It is
therefore important for online communities to provide better ways to verify every
152
member’s histories and facilitate the pairing of like-minded members. With effective
reputation and discovery features, users can expect reward and recognition from their
participation, and engage with others about things and activities that matter to them. The
results suggest that the effective uses of reputation systems and social matching systems
account for a higher proportion of variance in the perceived reciprocity and interpersonal
bonds respectively. It is therefore reasonable to assume that in geolocation-aware
communities, there is higher reliance on effective reputation and matching systems.
These two interpretations are relevant to two types of communities. For
communities that have already implemented reputation systems or matching systems such
as eBay, GitHub, or Twitter, the addition of geolocation awareness will make these
existing features more effective and relevant. But geolocation may not make a successful
and meaningful standalone feature. For communities that are solely based on the sharing
of location data such as Foursquare, there must be effective reputation and matching
features. The implementation of such features may facilitate interpersonal connection to
develop over meaningful communication. These features may also foster a sense of
obligation and accountability over valid and reliable measures of users’ past
contributions. Geolocation awareness is a social “add-on” to other features that provide
the social bond and normative rules for a community.
It must also be emphasized that the results do not support this assumption for other
community processes. Geolocation awareness does not seem to affect the influence of the
use of social matching systems on social identification in a community. A possible
interpretation is that geolocation awareness does not provide additional cues for
153
strengthening or weakening social identities that are based on common interests in other
topics. That is, being at the same place together is not a sufficient condition for a social
group to have a particularly strong sense of identity. Furthermore, if this social group is
based on shared interests in some topic, co-location does not further increase the
influence of these interests This is somewhat inconsistent with the reasoning in this study,
which holds that location awareness theoretically should increase social presence for a
social group (Christopoulou, 2008). Yet it is not entirely unexpected. First, it is possible
that as most of the study participants are from the same university, co-location is deemed
as a taken-for-granted condition. Location awareness simply does not serve as an
additional social cue that supplements the interest in some topics. This explanation
implies an issue of external validity for the research method of this study (Brewer, 2000).
But more importantly, this finding suggests that the idea that proximity regularly
influences social affinity needs to be examined in the new environment of online
communities. While Social Impact Theory specifies the strong social influence of
proximity among existing ties, it is worth considering the new practices and processes
that develop affinity among newly formed ties in emergent communities (Latane & Liu,
1996).
Second, social identification is a complicated process that involves people’s
perception of a social group at cognitive, emotional and evaluative levels (Prentice, et al.,
1994). It is possible that as the geolocation-aware group develops its identity at different
levels, geolocation awareness fails to affect some of the identification process. For
example, users may not have developed an emotional attachment to the community that is
154
apparently constructed on common interests in sharing used textbooks. Therefore
geolocation awareness will not play a strong role in this identification process. This
explanation suggests that location awareness is a multi-dimensional experience that
involves geography, relationships and resources. Geographic proximity might influence
different phases and components of social identification in different ways, and configure
social affinity for different kinds of ties, tasks or situations (Paay & Kjeldskov, 2008). If
there are sufficient relational and informational cues to sustain people’s identification of
common identities and goals, geography does not necessarily add additional relevance or
meaning to this identification process. For example, an interest group of iPhone
developers on Facebook may not necessarily feel a stronger sense of identity by labeling
themselves or as residents of the Los Angeles area, unless there is a need for such
labeling to facilitate meet-ups at physical locations. Even so, this geographic labeling
may influence the emotional aspect of social identification - the emotional attachment to
an identity - differently than the evaluative aspect - the assessment of the pros and cons
for an identity.
From geocoding to geotagging. The findings suggest that rather than a disruptive
technological breakthrough, location awareness is more like a social “add-on” for the
community features that sustain reciprocal interactions and facilitate interpersonal
communication. Geolocation awareness is a technology that is appropriated in people’s
existing socio-technical experience in online communities (DeSanctis & Poole, 1994;
Rogoff, 1995). In technology appropriation, the novelty of a new technology quickly
dissipates when its aim shifts from changing existing practices to improving them (Wirth,
155
von Page, & Karnowski, 2008). Location is a highly social concept, involving the
geographic and physical surroundings and people’s social interaction and experiences in
particular places (Paay & Kjeldskov, 2008). Therefore the focus of geolocation
technologies is to add additional contextual information that is relevant to people’s social
engagement with their environments (geotagging), rather than provide static geographic
data that represents people’s whereabouts (geocoding). Location awareness improves our
social experience by annotating it, not recreating it. Similarly, geographic proximity is
not a contextual cue that automatically creates affinity among new and old ties in
emergent communities. It facilitates the development of mutual accountability and
interpersonal bonds at a community level (Latane & Liu, 1996). It must integrate with
features that configure the verification and matching of relevant attributes of community
members. The theory of social impact based on immediacy and proximity must be
considered along with the theories of identification, reciprocity and social capital.
This shift of focus is illustrated in the findings. Rather than directly increasing
participation, geolocation awareness makes reputation systems more effective at fostering
reciprocity, and social matching systems better at improving interpersonal bonds. This
shift is also partly demonstrated when the findings fail to support the assumptions that
geolocation awareness alone will improve social identification, or that geolocation
awareness will help privacy-savvy users access more bridging social capital. Location
awareness needs to be embedded within people’s use of and reliance on verification,
communication, discovery and structuring features that sustain interest and increase
mutual interdependence. It must be integrated with the constructive socio-technical
156
processes of reciprocity, identification and social capital building. The essential
significance of location awareness derives not from how the geographic coordinates of
every activity or every user can be captured and recorded. The social distance between
members of an online community does not derive from the perception of geographic
proximity alone. Rather, location awareness is interesting only when each element of
social interaction and information flow can be enhanced with geotags and made more
efficient with the dynamic configuration of physical proximity. Only when geotags make
reciprocation of contribution a more important process in a community, the verification
of identities through profiles can become more efficient and effective. Only when geotags
make interpersonal bonds more important for holding a community together, social
discovery systems can become more useful.
This observation coincides with the trend in the technology community that
geolocation awareness is delegated as a secondary support feature. For example,
Facebook discontinued its Places features and instead put geocoding behind all of its
major functions such as wall posting and photo sharing (Fiveash, 2011). After the initial
excitement about the geolocation technology subdued, many location-based services like
Loopt, Foursquare and Gowalla are struggling to find alternative ways to sustain lively
user communities (Van Grove, 2012). These companies have been experimenting with
engaging users in activities like merchant deals, business recommendations or story
telling. These efforts more or less all rely on users sharing personal information and
interacting with one another in meaningful ways. This trend suggests that as geolocation
awareness is a multi-dimensional social experience, geocoding by itself cannot fully
157
support a community service. In other words, meaningful social experience must be
created out of a community that shares useful and interesting resources about different
locations for its members.
To make such location-enhanced sharing an engaging community process,
designers and developers must conceive features that make contributors feel recognized
and rewarded and make interpersonal connections prominent, relevant and enduring.
When people commit time and effort to share something about a place, there needs to be
a community-wide mechanism that provides a sense of reciprocity. Generalized
reciprocity is important as the contributors not only feel that they will get something in
return from other contributors, but also believe in the community as a fair organization
that recognizes contribution. For this to work, Foursquare and other networks must devise
meaningful ways that turn participation into contribution of useful information about
places, and not simply check-ins at different places. To facilitate this the idea that such
contribution matters, community designers should implement reputation and reward
systems that satisfy both intrinsic and extrinsic needs for recognition. A combination of
symbolic rewards such as status badges and titles with practical rewards such as money
and merchandise could help making such systems an effective tool of reciprocation. For
example, designers at Foursquare can think of ways to increase reciprocation of
participation with the badges of mayors for their most active users. As a result, being a
mayor at some places not only brings perks like deals at restaurants, but also encourages
more recognition, respect and interaction from other users. This will help users
158
understand, evaluate and compare the extrinsic rewards of titles and badges in a
meaningful community context.
At the same time, community designers should be aware that interpersonal bonds
could be more important than shared identities in the use of geolocation features. The key
challenge for a location-aware community is to facilitate the discovery and
communication process among users who share interests. A fundamental feature is a
meaningful profile system that helps users express their interests and open themselves for
peer discovery. Popular location-based services like Foursquare have just begun to
implement a richer profile system, which corroborates the importance of such features
(Taylor, 2012). Yet a more pressing challenge is to design a communication and
messaging system that engage users in secure, open and meaningful conversations.
Communication that involves messaging, voice calls and video chat is already available
in some location-based services (Perez, 2012). Just showing “people who are near you”
in a mobile application will not make users instantly closer to others and ready to start
conversations. This is because geographic proximity does not guarantee mutual affinity
without implementing other verification, matching and communication features. The real
issue is how to help users develop stronger understanding about some subjects that
interest them, for example, local foods, deals or sightseeing spots. In other words,
location-based communication must provide meaning about some bonding experiences,
rather than just enable ad hoc conversations within random pairs. For example,
applications like Highlight can learn from Twitter and make most conversations public by
default. This way, communication about shared interests in some places can engage more
159
people and become the basis of more inclusive groups. By making otherwise private
conversations about recommendations for local restaurants public, for instance, location-
based communities can actually engage more interests in these recommendations and
improve interpersonal connections for more users.
The Social Meaning of Sharing
A key activity in online social networks like Twitter is the sharing of information
that has personal relevance and engages common interests (Israel, 2009). The
significance of sharing seemingly trivial information about oneself was ridiculed initially
as narcissistic self-promotion (Pemberton, 2009). It becomes increasingly clear that the
emphasis on connecting many people through constant and frequent sharing of all kinds
of information does turn Twitter from a group messaging application into a global online
community (C. Li & Bernoff, 2008). Information sharing is a community-building
process. It is a process that people collectively contribute personal resources, grow
attached to the shared goals of building public repositories of information, and engage in
accountable, meaningful and open communication with others. By implementing features
that facilitate users’ exploration of shared interests and discovery of meaningful
connections, therefore, social networks like Twitter foster constructive social processes
that build and sustain communities. Following this logic, we can examine what social
processes promote information sharing in order to have a better understanding of what
makes an online community work.
Quality and quantity of sharing. This study aims to show some of these
constructive social processes as well as the features that support these processes. The
160
results suggest that in general, the quality of information contribution is positively
predicted by reciprocity and social identification with a community. These findings are in
line with some previous studies that demonstrate the importance of community-wide
norms of reciprocation and strong social identification with the goals and purposes of the
community for information sharing effectiveness (Kankanhalli, et al., 2005; Wasko &
Faraj, 2005). The results from these groups all illustrate the importance of a strong social
identity that defines the scope of shared interests and purposes for participating in each
community. In other words, users of personal sharing systems can gradually derive a
sense of identification with their emergent communities, and consider this identification
process an important factor for their participation in information sharing.
Furthermore, the finding that bonding social capital is negatively associated with
the perceived quality of contribution is interesting. This suggests that for communities
that focus on sharing practical information, a weaker sense of bonding social experience
might actually motivate people to contribute more relevant and useful information
without being burdened with developing interpersonal liking and deeper emotional ties.
And a higher sense of bonding social capital might actually deter people from sharing
information that is useful and meaningful for a broader range of ties in the community.
This may result in a virtual segregation that reinforces redundant and constrained
information within the boundaries of existing strong ties (Gentzkow & Shapiro, 2011b).
In comparison, the actual increase in the quantity of information contribution is
only positively associated with social identification. It confirms that over time, the
development of a collective identity is helpful for online communities to sustain growth
161
and foster participation. This collective identity is based on the discovery and iteration of
shared goals, objectives and agenda that provide a sense of belongingness. Particularly
for emergent communities supported by niche sharing applications like Voilai.com, it is
important for a common identity - sharing personal needs with others around you - to
engagement members beyond bilateral conversations about personal trivia. Yet it should
be noted that other social processes like reciprocity and social capital are not found to be
associated with participation in online communities. This is an unfortunate result because
it is vital to know precisely what social processes predict the sustained participation in
online communities.
Reciprocity, community attachment and social capital might be the key socio-
technical processes that help users perceive the focus and purpose of a community. These
processes constitute the normative social influence that shapes users’ general perception
of their intrinsic needs of security, obligation and recognition (Fulk, et al., 1990). In
comparison, the actual increase in participation frequency over time might be driven by
more contextualized processes that vary across different communities. For example, a
Voilai.com user may claim that she has contributed relevant tweets because she believes
her contribution is recognized and appreciated by other users. Yet she only posted needs
or offers when she saw others posting similar content. The positive influence of
reciprocity on contribution quantity is only relevant when there are defined topics, foci
and purposes for engaging in reciprocal communication.
More specifically, perceived generalized reciprocity only influences quality and
not quantity of information contribution. When there is a stronger sense of equitable
162
exchange of contribution with other users, there is a stronger urge to share better content.
This suggests that the integration of interpersonal accountability and collective norms can
more effectively satisfy community members’ intrinsic needs for recognition and
appreciation, which in turn promotes the contribution of quality content. In other words,
the quality of contribution is influenced by the fulfillment of certain intrinsic needs such
as recognition and reputation rather than extrinsic needs, such as title, money or material
rewards (Roberts, et al., 2006).
Similarly, it is interesting that the perceived quality, and not quantity of
contribution is more effectively supported by the processes of social identification.
Across the groups in this study, the focus of the community systems is about sharing
personal information that might be of interest to another person, so that the shared
content is replied, forwarded or re-shared. There is no explicit mandate or description of
the identity and focus for each community system. Yet the results from these groups all
illustrate the importance of a strong social identity that lays out the shared interests and
purposes for participating in each community. In other words, users of personal sharing
systems can gradually derive a sense of identification with their emergent communities.
They also consider this identification process an important factor in their participation in
information sharing.
Sharing in different contexts of use. In the study, users in a more general-
purpose community like Twitter consider the access to diverse and new sources of
information and the sense of social identification to be the drivers for their contribution of
quality content. In comparison, users in the more narrow-focused Voilah and Voilai
163
groups still rely on the sense of reciprocity. This contrast suggests that the specificity of
focus in information sharing may have an influence on why people contribute quality
content. With a more specific focus on the content and category of information, people’s
motivation for sharing quality information and their actual volume of contribution may be
driven by the intrinsic needs for system-wide reciprocation of contribution (Jeppesen &
Frederiksen, 2006). When the focus is less specific, both extrinsic needs for new
perspectives and intrinsic needs for social identities drive the intention to contribute
quality information. This could lend insight to examining information sharing patterns at
different stages of community development, or between niche communities like
Foodspotting and broad-purpose communities like Twitter. When an online community
gradually narrows its focus over time, it may be necessary to make shifts so that users
become more aware of the normative processes that increase mutual accountability. This
way, the users can more effectively fulfill the needs for recognition and security through
the contribution of relevant content, rather than just focus on getting new information and
making new connections.
In addition to the different kinds of foci, the results do suggest that users of
different features have different patterns of community participation. In the Voilai group,
which has a more specific focus and geolocation features, social capital and interpersonal
bonds become salient motivations for contributing more and better information. While
the evaluation of contribution quality is negatively influenced by the access to bonding
social capital, the quantity of contribution is positively influenced by perceived
interpersonal bonds and bridging social capital. It is reasonable to assume that as
164
information-sharing activities are supplemented with location information, the
contribution of quality content is again driven by the needs for making new contacts
through communication with more diverse ties. This is in general in line with the latent
tie theory, which states that the addition of new communication technologies has the
greatest influence on activating latent ties and recasting weak ties (Haythornthwaite,
2002). Yet the finding also extends this theory to explain the impact of such access to
new ties on information sharing patterns and on community development. When a new
technology enables the formation of new ties and affords new ways to access new
information, a community may have a different focus on the constructive processes that
drive participation. Such processes may be about either interpersonal bonds or
generalized reciprocity. The choice of focus also depends on what fundamental
community features are affected by the technology.
In sum, the study provides a complicated if not vivid example for how different
contexts, foci and features in a community might influence the social processes
conducive to effective information sharing. The social meaning of sharing can be stated
this way: community building through the identification of shared purposes and interests
and reciprocal, inclusive communication that leads to meaningful relationships. First, the
focus on the development of collective identities is the universal motivation for both
sharing information of higher quality and sharing at a regular pace. This implication
enables us to think about online social networks like Twitter as a community that has
shared purposes and meanings, rather than as the aggregation of personal cliques.
Second, the perception of generalized reciprocity also motivates people to share
165
information of higher quality. This is a helpful idea as it frames information sharing on
social networks like Twitter and Voilai.com as a multi-directional exchange of
perspectives, news, and knowledge, rather than as purely narcissistic broadcast of one’s
daily trivia. Information sharing thus forms the basis for a community that encourages
equitable distribution of rewards and contribution. The meaning of reciprocal
relationships in information sharing is particularly relevant for social media applications
that aim to become more useful, either as portals of news or as repositories of knowledge.
Different communities define, develop and shift their focuses and communication
patterns at different stages in different ways at different stages. Therefore careful work is
needed if researchers want to have a better understanding of sharing as a community
building process that may foster reciprocity, interpersonal bonds and social identification
in different ways.
166
Chapter Nine: Conclusions
Summary
This study employed a field experiment to examine how the use of features such as
reputation systems, social matching systems and privacy controls influences several
community processes that in turn promote information sharing effectiveness. The
experiment also examines how geolocation awareness moderates these influences. The
results suggest that the use of reputation systems has a positive influence on the
reciprocity in a community, which is even stronger when geolocation information is
attached to user profiles and user-generated content. The use of social matching systems
such as tag-based recommendations of interests and connections is positively associated
with perceived interpersonal bonds and social identities in a community. Geolocation
awareness, again, reinforces the influence of the use of social matching systems on
perceived interpersonal bonds. The use of privacy controls is positively related to
bridging social capital in a community, which is not moderated by the presence of
geolocation information. Finally, generalized reciprocity and social identification
positively contribute to the perceived quality of information contribution.
With these findings, the study makes a preliminary effort at uncovering the social
meaning of sharing and geocoding, two interesting trends in current online social
networks. By showing the connections between features and social processes, the study
demonstrates that information sharing is a community building process in which
individuals actively employ communication, profile and privacy features to explore
shared interests, discover potential connections and access social capital. By revealing the
167
moderation effect of geolocation features on several key socio-technical processes, the
study also shows that geocoding is a social “add-on” that improves the effectiveness for
individuals to use reputation and discovery features to construct a reciprocal and bonding
social environment.
Theoretical Significance
This study makes three theoretical contributions. First, it shows that the
development of online communities depends on both social identification and
interpersonal bonds (Ren, et al., 2007). This finding advances the previous theory about
social identification in online communities such as SIDE by considering enhanced
interpersonal bonds as a result of the wide use of reputation, matching, and discovery
features (Postmes & Spears, 2000; Spears, Lea, Corneliussen, Postmes, & Ter Haar,
2002). It also extends the recent perspectives on interactivity, profile management, and
activity metrics in online communities to consider the impact of enhanced interpersonal
communication on the attachment to communities (Kalyanaraman & Sundar, 2008; Tong,
et al., 2008; Walther, Van der Heide, Kim, Westerman, & Tong, 2008). But more
importantly, it stresses that the investigation of different modes of online community
development must be grounded in the socio-technical processes surrounding the use
practices of specific features, and in the perceptual outcomes that are relevant to the
shared purposes of a community. This perspective is helpful for broadening the micro-
level focus on impression formation, relational affinity and social capital in classic CMC
research to consider the socio-cognitive processes of collective actions at the community
level.
168
Second, it provides a way to quantify the process in which new technology such as
geolocation awareness is appropriated to facilitate or constrain existing use practices. It is
widely accepted that rather than abruptly change behavior, new technology or features
are integrated into people’s daily practices of technology use for achieving established
goals and fulfilling existing preferences (Bijker & Pinch, 1984; MacKenzie & Wajcman,
1999; Mathieson, 1991). To avoid making social deterministic statements about the
overarching endurance of existing practices and preferences, one must first identify the
specific technical condition and context that shape people’s existing use of technology. It
would then be possible to discuss how new technology enriches or impoverishes this use
context. The study employs a moderation analysis to examine how geolocation awareness
facilitates the use of existing common features for improving mutual accountability and
enhance communication in online communities. By explicating the statistical
relationships captured in the socio-technical processes of new feature use, this exercise
provides useful lessons on how to study the introduction and appropriation of other new
technologies.
Third, this study provides an update for the Social Impact Theory (SIT) from the
perspective of community development. SIT is a useful framework for thinking about
social influence as a function of the number, immediacy and characteristics of social
actors (Latane, 1981). From this perspective, geolocation awareness is defined as the
perception and configuration of geographic proximity. Features like geotags facilitate
social interaction, because they afford users the capability to determine the relevance,
intensity and necessity for communication. So geolocation proximity becomes a vital
169
contextual cue for developing interpersonal bonds and identities. By examining the
integration between geolocation awareness and several constructive social processes, this
study aims to explicate the idea of “social distance” in the particular communication
environment of emerging communities (Latane & Liu, 1996). Social proximity, as this
study shows, indeed requires the development of reciprocity and interpersonal bonds and
the enhancement with geotags. The discussion of reciprocity, social identification,
interpersonal bonds, and social capital illuminates the potential factors or processes for
examining the “intersubjective” structures that configure people’s experience of
psychological proximity with others (Latane & Liu, 1996).
Practical Relevance
The study also provides an interesting case for rethinking strategies for studying
online communities in the age of Facebook and Twitter. It is normally a researcher’s job
to study retroactively the effects of a “new” technology (Kellermann, 1985). Yet the
availability of behavioral data from social media and the feasibility of building online
social networks make it possible - and indeed necessary - for researchers to take up a
more proactive role. Researchers can employ more grounded techniques to study online
communities and challenge existing assumptions. To advance both theory and methods,
researchers must acquire technical know-how to better understand how to build and study
an online community.
Because the use of online communities is a situated socio-technical process,
researchers must be able to have a better understanding of the social, communicative and
technical foundations of different features and the social-cognitive outcomes of the use of
170
such features. If code is the law for behavior in cyberspace, then there is value to know
how the code is written and used (Lessig, 1999). If features are the implementation and
interface of the code, then there is value to know how features are designed and used.
Based on the knowledge of this feature-based social process, researchers can build a
systematic understanding of different elements of the social experience in online
communities. With this knowledge, researchers can observe the evolution of
communication networks and the gradual development of community norms,
identification and bonds. The process of trying to build a working online community is
challenging, but provides many useful insights. For example, while it is relatively easy to
implement geolocation awareness, it is much more difficult to design an effective and
useful social matching systems that satisfy users’ needs for both content discovery and
relationship recommendation. Without the knowledge of how to design a reputation
system, it would never be possible to perceive how to implement effective reward
distribution and prevent the bi-polarization of user participation. Such experiences are
valuable, because they help researchers make relevant and realistic assumptions about the
technologies and communities they study.
This study also provides insights about geolocation technologies. Such insights
might be useful for designers who want to increase participation in online communities
and for users who wonder what to do with the disclosure of their personal information.
As the study shows, developers should overcome the temptation to put geocoding at the
center of community experiences. Rather, they should focus on building communication,
verification and discovery features that facilitate users’ engagement with peer users and
171
foster their identification with the shared purposes of the community. For them, the key
issue is how geolocation awareness improves the effectiveness of these features.
The findings of this study also show that users should stop thinking about privacy
as just a security feature. Rather, a more practical action is to focus on using privacy
controls to strategically configure their interactions to access different forms of social
capital. Users should also understand the tradeoffs in disclosing their location data and
start exploring the interest discovery and reputation features. In any case, meaningful
communities should, and indeed could be built based on the effective integration of a
hyperlocal focus and hyperactive sharing of personal information. This way, the promise
of location-based services could be fulfilled. With the help of geotagging technology,
effective communities can be built which help people find out interesting and relevant
information about many public places, and also afford people opportunities to make new
friends.
Limitations
The study has two limitations that may undermine the internal validity of the
results. First, the manipulation of the community features may not be effective. The study
aims to observe how people use and perceive several community features and how this
socio-technical process influences their perception of the respective community.
Significant efforts have been made to design these features and make them usable,
meaningful and relevant. However, there is no formal procedure that mandates the
participants to use all the features. Instead, participants are expected to explore and
discover different features. This is a deliberate effort at increasing the ecological validity
172
of the study by situating participants in a more naturalistic process of community use. Yet
this approach might threaten the content validity. Although post-hoc analysis showed a
significant change in the awareness and use of reputation and privacy features, there is no
guarantee that participants may not have actively used the designed features and acquired
sufficient knowledge about them. Although some of the hypothesized socio-technical
processes in the use of community features are observed in the study, the variances in
reciprocity, community attachment and social capital explained by the use of different
features are overall small. This raises the question of exactly how features that calculates
reputation, recommends content and connections, and configures group boundaries
matter to people’s social experience online. The usability of the websites may have also
been undermined by sporadic crashes and bugs, which further weakens the content
validity of the study.
Second, the use of community features is assessed with a mixture of formative and
reflexive measures that are taken in the same survey with the dependent variables of
reciprocity, community attachment and social capital. On one hand, this measurement of
the use of features may not truthfully reflect the actual use, therefore undermining the
construct validity. However, the assessment of technology use through formative – items
that ask about different aspects of use – and reflexive – items that ask about the
experience of use – measures is widely adopted and accepted in technology research
(Tung, Chang, & Chou, 2007; Venkatesh, Morris, Davis, & Davis, 2003; Yi, Jackson,
Park, & Probst, 2006). Therefore, a mixture of these measures will hopefully mitigate the
threat to construct validity (Diamantopoulos & Winklhofer, 2001; G. C. Moore &
173
Bendasat, 1991). On the other hand, as all variables are taken from one single survey,
there is the potential risk of common method bias (Podsakoff, et al., 2003). To address
this bias, efforts are made so that the survey measures have different styles of scales and
framing languages, and appear in different blocks that are separated by unrelated
questions. The goal of this arrangement is to minimize any conceptual connections
between the constructs of feature use, and the constructs of community processes.
Furthermore, the quantity of contribution is measured from users’ records on Twitter and
Voilah/Voilai server logs independently from the survey measures. Finally, the potential
bias has also been assessed to some extent with the assessment of discriminant validity
for each construct. As shown in the Appendix A, measurements for all constructs have
achieved satisfactory discriminant validity values, with their average variances extracted
(AVE) higher than the rules of thumb value of 0.5.
Future Research
This study makes claims about the positive influence of features such as reputation
systems, recommendation and matching systems and privacy controls on several
community processes. Underlying these claims is the assumption that people actively and
regularly pay attention to and care about these features in their use of online
communities. This is an assumption that may be shared by businesses, so it is worth
knowing whether or not people will use or even notice the features that developers spend
time and efforts producing. This question seems to demand answers from the tradition of
technology adoption studies (Davis, Bagozzi, & Warshaw, 1989; Venkatesh & Davis,
2000). Previous adoption research provides useful clues about how the attributes of new
174
technologies determine the likelihood for people to use them. Yet most of the attributes in
these studies such as ease of use, usefulness and trialability have little practical relevance,
when current technologies have these by design. More importantly, the cost of adopting
or not adopting a technology is in mostly small from a user’s perspective, requiring
perhaps only checking a box or clicking a button. This makes the actual prediction of
adoption with these variables often irrelevant or wasteful. As shown from anecdotes on
how technologies companies conduct field tests, new features are pushed to us for
immediate and sometimes mandatory use. It is the actual use process that is of actual
research interest (Christian, 2012).
Future research must find ways to untangle the socio-technical processes that are
central to the building and sustaining of communities, rather than just predict whether or
not individuals will accept and adopt new features. As these social processes often take
time to unwrap, it is necessary for researchers to conduct longitudinal field studies about
the development of social cognition that maximizes the perceived benefits and minimizes
the structural barriers to participation in online collective actions. Such an endeavor must
also be firmly grounded in the observed use practices surrounding the technical features,
because the specific user practices of technical artifacts determine the communication
patterns, interpersonal affinity and overall community structure (Orlikowski & Iacono,
2001).
There are two potential processes that are worth further investigation. The first is
the accumulation and access of different social capital in a community. Although the
hypotheses specifying the connections between privacy controls and social capital receive
175
the least support, the study does reveal that structural configurations and technological
mediation play a significant role in people’s perception of social capital. It is likely that
the structure of bonding and bridging social capital is more complicated than an on-off
switch of boundaries between strong tie closures and the dispersed sets of weak ties. A
more nuanced look at the overall centrality, density and other structural metrics in
individuals’ ego networks could reveal more about their actual access to bonding and
bridging resources (Burt, 2007). This also suggests that a detailed examination of more
technical features is necessary, because other features may influence communication or
organization processes that shape individuals’ ego networks.
Second, the social presence entailed in geolocation awareness is worth a more
nuanced look. This study successfully illustrates that geolocation awareness moderates or
reinforces the effectiveness of several constructive social processes in an online
community. Nevertheless, it is worth pointing out that location awareness is a rich social
experience (Aaltonen, et al., 2005). For example, over time, participants in the location-
aware group in this study reported a decreasing “radius of interest” – the geographic
range for them to get information from relative to where they live (d = -0.31, sd = 0.15,
t(25) = -1.99), whereas there were no significant changes in the radius of interest for
information in the non-geolocation group. Psychological engagement, behavioral
interdependence and communicative richness afforded by co-location are powerful social
experiences that serve as the basis for more enduring social groups (Scarpetta, 2008). It is
worthwhile to examine what specific content and activities can take advantage of these
positive social experiences. For example, mobile gaming might benefit from the social
176
presence afforded by location awareness whereas local news sharing might not. More
focused field experiments can manipulate the content or attributes of tasks for group
actions, and reveal the direct effect of social presence that is afforded by geolocation
technologies.
177
Bibliography
Aaltonen, A., Huuskonen, P., & Lehikoinen, J. (2005). Context awareness perspectives
for mobile personal media. Information Systems Management, 22(4), 43-55.
Amichai-Hamburger, Y., & McKenna, K. Y. A. (2008). The Contact hypothesis
reconsidered: Interacting via the Internet: Self presentation and impression
formation. Journal of Computer-Mediated Communication, 11, 825-843.
Aral, S., & Van Alstyne, M. W. (2011). Networks, information & brokerage: The
diversity-bandwidth tradeoff. American Journal of Sociology, 11(7), 90-171.
Ardichvili, A., Page, V., & Wentling, T. (2003). Motivation and barriers to participation
in virtual knowledge-sharing communities of practice. Journal of Knowledge
Management, 7(1), 64-77.
Aries, E. J., & Johnson, F. L. (1983). Close friendship in adulthood: Conversational
content between same-sex friends. Sex Roles, 9(12), 1183-1196.
Avrahami, D., Gergle, D., Hudson, S. E., & Kiesler, S. (2007). Improving the match
between callers and receivers: A study on the effect of contextual information on
cell phone interruptions. Behaviour & Information Technology, 26(3), 247-259.
Axelrod, R. (1985). The evolution of cooperation. New York, NY: Basic Books.
Ba, S., & Pavlou, P. A. (2002). Evidence of the effect of trust building technology in
electronic markets: Price premiums and buyer behavior. MIS Quarterly, 26(3),
243-266.
Bagozzi, R. P., & Dholakia, U. M. (2006). Open source software user communities: A
study of participation in Linux user groups. Management Science, 52(7), 1099-
1155.
Bakshy, E., Rosenn, I., Marlow, C. A., & Adamic, L. A. (2012, April 16-20). The role of
social networks in information diffusion. Paper presented at the ACM WWW
2012, Lyon, France.
Baltes, B. B., Dickson, M. W., Sherman, M. P., Bauer, C. C., & LaGanke, J. S. (2001).
Computer-mediated communication and group decision making: A meta-analysis.
Organizational Behavior and Human Decision Processes, 87(1), 156-179.
Bandura, A. (1985). Social foundations of thought and action. Englewood, NJ: Prentice-
Hall.
178
Banerjee, N., Chakraborty, D., Dasgupta, K., Mittal, S., Nagar, S., & Saguna (2009, May
18-20). R-U-In?: Exploiting rich presence and converged communications for
next-generation activity-oriented social networking. Paper presented at the Tenth
International Conference on Mobile Data Management: Systems, Services and
Middleware.
Barak, A., Boniel-Nissim, M., & Suler, J. (2008). Fostering empowerment in online
support groups. Computers in Human Behavior, 24(5), 1867-1883.
Barnes, S. B. (2006). A privacy paradox: Social networking in the United States. First
Monday, 11(9)
Barnes, S. B. (2008). Understanding social media from the media ecological perspective.
In E. A. Konijn, S. Utz, M. Tanis & S. B. Barnes (Eds.), Mediated interpersonal
communication (pp. 14-33). New York: Routledge.
Barnlund, D. C., & Harland, C. (1963). Propinquity and prestige as determinants of
communication networks. Sociometry, 26(4), 467-479.
Beale, R., & Lonsdale, P. (2004). Mobile context aware systems: The intelligence to
support tasks and effectively utilise resources. Paper presented at the Mobile
Human-Computer Interaction – MobileHCI 2004: 6th International Symposium,
Glasgow, UK.
Benevenuto, F., Duarte, F., Rodrigues, T., Almeida, V. A. F., Almeida, J. M., & Ross, K.
W. (2008). Understanding video interactions in YouTube. Paper presented at the
Proceeding of the 16th ACM international conference on Multimedia.
Bernhaupt, R., Mihalic, K., & Obrist, M. (2008). Usability evaluation methods for mobile
applications. In J. Lumsden (Ed.), Handbook of research on user interface design
and evaluation for mobile technology (Vol. 1-2, pp. 745-758). Hershey, PA:
Information Science Reference/IGI Global.
Biersdorfer, J. D. (2012, February 13). Q&A: When Facebook marks your spot. The New
York Times, Gadgetwise. Available at
http://gadgetwise.blogs.nytimes.com/2012/02/13/qa-when-facebook-marks-your-
spot/
Bijker, W. E. (1995). Of bicycles, bakelites, and bulbs: Toward a theory of sociotechnical
change. Boston, MA: MIT Press.
Bijker, W. E., Hughes, T. P., & Pinch, T. J. (Eds.). (1987). The social construction of
technological systems: New directions in the sociology and history of technology.
Cambridge, MA: MIT Press.
179
Bijker, W. E., & Pinch, T. J. (1984). The social construction of facts and artefacts: Or
how the sociology of science and the sociology of technology might benefit each
other. Social Studies of Science, 14, 399-441.
Bimber, B., Flanagin, A. J., & Stohl, C. (2005). Reconceptualizing collective action in
the contemporary media environment. Communication Theory, 15(4), 365-388.
Bock, G., Zmud, R. W., Kim, Y., & Lee, J. (2005). Behavioral intention formation in
knowledge sharing: Examining the roles of extrinsic motivators, social-
psychological forces and organizational climate. MIS Quarterly, 29(1), 87-111.
Boland, R. J., Tenkasi, R. V., & Te'eni, D. (1994). Designing IT to support distributed
cognition. Organization Science, 5(3), 456-475.
Bolton, G. E., Katok, E., & Ockefels, A. (2004). How effective are electronic reputation
mechanisms? An experimental investigation. Management Science, 50(11), 1587-
1502.
boyd, d. (2006). Friends, Friendsters, and MySpace Top 8: Writing community into being
on social network sites. First Monday, 11(12).
Bradley, N. A., & Dunlop, M. D. (2005). Toward a multidisciplinary model of context to
support context-aware computing. Human-Computer Interaction, 20, 403-446.
Brandon, D. P., & Hollingshead, A. B. (2004). Transactive memory systems in
organizations: Matching tasks, expertise and people. Organization Science, 15,
633-644.
Breslin, J., & Decker, S. (2007). The future of social networks on the Internet: The need
for semantics. Internet Computing, IEEE, 11(6), 86-90.
Brewer, M. B. (1979). In-group bias in the minimal intergroup situation: a cognitive-
motivational analysis. Psychological Bulletin, 86, 307-324.
Brewer, M. B. (2000). Research design and issues of validity. In H. Reis & C. Judd
(Eds.), Handbook of research methods in social and personality psychology. New
York, NY: Cambridge University Press.
Briggs, P., & Smyth, B. (2006). On the role of trust in collaborative Web search.
Artificial Intelligence Review, 25(1-2), 97-117.
Bristow, H. W., Baber, C., Cross, J., Knight, J. F., & Woolley, S. I. (2004). Defining and
evaluating context for wearable computing. International Journal of Human-
Computer Studies, 60(5-6), 798-819.
180
Burt, R. S. (2000). The network structure of social capital. In B. M. Staw & R. I. Sutton
(Eds.), Research in organizational behavior (pp. 345-423). New York, NY:
Elsevier Science JAI.
Burt, R. S. (2002). The social capital of structural holes. In M. F. Guillén (Ed.), The new
economic sociology: Developments in an emerging field (pp. 148-190). New
York, NY: Russell Sage Foundation.
Burt, R. S. (2007). Brokerage and closure: An introduction to social capital. New York,
NY: Oxford University Press.
Byrne, D., & Buehler, J. A. (1955). A note on the influence of propinquity upon
acquaintanceships. The Journal of Abnormal and Social Psychology, 51(2), 147-
148.
Cabrera, A., & Cabrera, E. G. (2002). Knowledge-sharing dilemmas. Organization
Studies, 23, 687-710.
Carte, A. T., & Russel, J. C. (2003). In pursuit of moderation: Nine common errors and
their Solutions. MIS Quarterly, 27(3), 479-501.
Castells, M. (2003). The Internet galaxy: Reflections on the Internet, business, and
society. New York, NY: Oxford University Press.
Castells, M., Fernandez-Ardevol, M., Qiu, J. L., & Sey, A. (2007). Mobile
communication and society: A global perspective. Cambridge, MA: MIT Press.
Chaiken, S. (1980). Heuristic versus systematic information processing and the use of
source versus message cues in persuasion. Journal of Personality and Social
Psychology, 39(5), 752-766.
Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information
processing within and beyond the persuasion context. In J. S. Uleman & J. A.
Bargh (Eds.), Unintended thought (pp. 212-252). New York, NY: Guilford Press.
Cheok, A. D., Teo, S. L., Cao, L., & Thang, L. N. (2005). Capture the flag: A multiplayer
online game for phone users. Paper presented at the Proceedings of the 2005
Ninth IEEE International Symposium on Wearable Computers (ISWC’05).
Chin, W. W. (1998). The Partial Least Squares approach to structural equation modeling.
In G. A. Marcoulides (Ed.), Modern methods for business research (pp. 295-336).
Marwah, NJ: Lawrence Erlbaum Associates.
181
Chin, W. W. (2004). Multi-group analysis with PLS: Frequently asked questions - Partial
Least Squares & PLS-graph. Available at http://disc-
nt.cba.uh.edu/chin/plsfaq.htm
Chin, W. W., Marcolin, B., & Newsted, P. R. (2003). A Partial Least Squares latent
variable modeling approach for measuring interaction effects: Results from a
Monte Carlo simulation study and an electronic-mail emotion/adoption study.
Information Systems Research, 14(2), 189-217.
Christian, B. (2012, April 25). The A/B test: Inside the technology that’s changing the
rules of business. Wired, 20(4).
Christopoulou, E. (2008). Context as a necessity in mobile applications. In J. Lumsden
(Ed.), Handbook of research on user interface design and evaluation for mobile
technology (Vol. 1-2, pp. 187-204). Hershey, PA: Information Science
Reference/IGI Global.
Clayton, N. (2011, November 21). Facebook sharing upsets users and shakes up news.
The Wall Street Journal, pp. 2A.
Coleman, J. S. (1988). Social capital in the creation of human capital. American Journal
of Sociology, 94(Supplement), S95-S120.
Connaughton, S. L., & Shuffler, M. (2007). Multinational and multicultural distributed
teams: A review and future agenda. Small Group Research, 38(3), 387-412.
Cook, M., & Campbell, D. T. (1979). Quasi-Experimentation: Design and analysis issues
for field settings. Boston, MA: Houghton Miffllin.
Counts, S. (2007). Group-based mobile messaging in support of the social side of leisure.
Computer Supported Cooperative Work (CSCW), 16(1-2), 75-97.
Cramton, C., & Orvis, K. (2003). Overcoming barriers to information sharing in virtual
teams. In C. Gibson & S. Cohen (Eds.), Virtual teams that work: Creating
conditions for effective virtual teams. New York, NY: Jossey Bass.
Creswell, J. W. (2008). Research Design: Qualitative, quantitative, and mixed methods
approaches. Thousand Oaks, CA: Sage Publications.
Culnan, M. J. R., & Bies, J. (2003). Consumer privacy: Balancing economic and justice
considerations. Journal of Social Issues, 59(2), 323-342.
Cuneo, A. Z. (2008). Mobile marketing based on place is finally making strides.
Advertising Age, 79(13), 24.
182
Daft, R. L., & Lengel, R. H. (1986). Organizational information requirements, media
richness, and structural design. Management Science, 32(5), 554-571.
Davis, F. D., Bagozzi, R. D., & Warshaw, P. R. (1989). User acceptance of computer
technology: A comparison of two theoretical models. Management Science, 35,
982-1002.
Debatin, B., Lovejoy, J. P., Horn, A.-K., Brittany, M. A., & Hughes, N. (2009). Facebook
and online privacy: Attitudes, behaviors, and unintended consequences. Journal
of Computer-Mediated Communication, 15(1), 83-108.
Dennis, A. R., & Garfield, M. J. (2003). The adoption and use of GSS in project teams:
Toward more participative processes and outcomes. MIS Quarterly, 27(2), 289-
323.
DeSanctis, G., & Poole, M. S. (1994). Capturing the complexity of advance technology
use: Adaptive Structuration Theory. Organization Science, 5(2), 121-147.
Dey, A. K., & Häkkilä, J. (2008). Context-awareness and mobile devices. In J. Lumsden
(Ed.), Handbook of research on user interface design and evaluation for mobile
technology (Vol. 1-2, pp. 205-217). Hershey, PA: Information Science
Reference/IGI Global.
Dey, A. K., & Mankoff, J. (2005). Designing mediation for context-aware applications.
ACM Transactions on Computer-Human Interaction, 12(1), 53-80.
Diamantopoulos, A., & Winklhofer, H. M. (2001). Index construction with formative
indicators: An alternative to scale development. Journal of Marketing Research,
38(2), 269-277.
Donald, B. (2011, March 16). Path social network announces more ways to share. The
Huffington Post. Available at http://www.huffingtonpost.com/2011/11/30/path-
social-network_n_1120424.html.
Donaldson, T., & Dunfee, T. W. (1994). Towards a unified conception of business ethics:
Integrative social contracts theory. Academy of Management Review, 19(2), 252-
284.
Donath, J. (2007). Signals in social supernets. Journal of Computer-Mediated
Communication, 13(1), article 12.
Drori, J. (2002, March 19). Location, location, location. Financial Times, p. 16,
183
Dutta-Bergman, M. J. (2004). Interpersonal communication after 9/11 via telephone and
internet: A theory of channel complementarity. New Media & Society, 6(5), 659-
673.
Eagle, N., Pentland, A., & Lazer, D. (2009). Inferring social network structure using
mobile phone data. Proceedings of the National Academy of Sciences, 106(36):
15274-15278.
Eagly, A. H., & Chaiken, S. (1993). The Psychology of Attitudes. New York, NY:
Hartcourt Brace College Publishers.
Ekeh, P. P. (1974). Social exchange theory: The two traditions. Cambridge, MA: Harvard
University Press.
Ellison, N., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook "friends:"
Social capital and college students’ use of online social network sites. Journal of
Computer-Mediated Communication, 12(4).
Ellison, N., Vitak, J., Steinfield, C., Gary, R., & Lampe, C. (2011). Negotiating privacy
concerns and social capital needs in a social media environment. In S. Trepte & L.
Reinecke (Eds.), Privacy online: Perspectives on privacy and self-disclosure in
the social web. New York, NY: Springer.
Emery, F. E., & Trist, E. L. (1960). Socio-technical systems. In C. W. Churchman & M.
Verhurst (Eds.), Management science, models and techniques (Vol. 2, pp. 83-97).
London, UK: Pergamon Press.
Fan, M., Tan, Y., & Whinston, A. (2005). Evaluation and design of online cooperative
feedback mechanisms for reputation management. IEEE transactions on
knowledge and data engineering, 17(2), 244-254.
Farmer, F. R., & Glass, B. (2010). Building web reputation systems. San Francisco, CA:
O'Reilly Media.
Ferran, C., & Watts, S. (2008). Videoconferencing in the field: A heuristic processing
model. Management Science, 54(9), 1565-1578.
Festinger, L., Schachter, S., & Back, K. (1950). The spatial ecology of group formation.
In L. Festinger, S. Schachter & K. Back (Eds.), Social pressure in informal
groups: A study of human factors in housing (pp. 141-161). Palo Alto, CA:
Stanford University Press.
Fiveash, K. (2011, August 24). Facebook ditches Places - but embiggens location
tracking. The Register, p. 33.
184
Flanagin, A., Stohl, C., & Bimber, B. (2006). Modeling the structure of collective action.
Communication Monographs, 73, 29-54.
Fogg, B. J. (2002). Persuasive technology: Using computers to change what we think and
do (interactive technologies). New York, NY: Morgan Kaufmann.
Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between
reviews and sales: The role of reviewer identity disclosure in electronic markets.
Information Systems Research, 19(3), 291-317.
Fulk, J., Heino, R., Flanagin, A., Monge, P., & Bar, F. (2004). A test of the individual
action model for organizational information commons. Organization Science, 15,
569-685.
Fulk, J., Schmitz, J., & Steinfeld, C. (1990). A social influence model of technology use.
In J. Fulk & C. Steinfeld (Eds.), Organizations and communication technology
(pp. 117-140). Thousand Oaks, CA: Sage.
Fuller, M. A., Serva, M. A., & Benamati, J. S. (2007). Seeing is believing: The transitory
influence of reputation information on e-commerce trust and decision making.
Decision Sciences, 38(4), 675-699.
Gentzkow, M., & Shapiro, J. M. (2011). Ideological segregation online and offline. The
Quarterly Journal of Economics, 126, 1799-1839.
Golbeck, J. (2007). The dynamics of Web-based social networks: Membership,
relationships, and change. First Monday, 12(11).
Goldberg, D. W. (2008). A geocoding best practices guide. Washington, DC: North
American Association of Central Cancer Registries, Inc.
Gosling, S. D., Rentfrow, P. J., & Swann Jr., W. B. (2003). A very brief measure of the
Big-Five personality domains. Journal of Research in Personality, 37, 504-528.
Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78,
1360-1380.
Guest, A. M., & Wierzbicki, S. K. (1999). Social ties at the neighborhood level. Urban
Affairs Review, 35(1), 92-111.
Gurak, L. J. (1999). The promise and peril of social action in cyberspace: ethos, delivery,
and the protests over MarketPlace and the Clipper chip. In M. A. Smith & P.
Kollock (Eds.), Communities in cyberspace (pp. 243-263). New York, NY:
Routledge.
185
Haddon, L. (2004). Information and communication technology in everyday life: A
concise introduction and research guide. New York, NY: Berg.
Hampton, K. (2007). Neighborhoods in the network society: The e-Neighbors study.
Information, Communication & Society, 10(5), 714-748.
Hampton, K., Goulet, L. S., Rainie, L., & Purcell, K. (2011). Social networking sites and
our lives. Pew Internet & American Life Project. Washington, DC: Pew Research
Center
Hampton, K., Livio, O., & Goulet, L. S. (2010). The social life of wireless urban spaces:
Internet use, social networks, and the public realm. Journal of Communication,
60(4), 701-722.
Hampton, K., & Wellman, B. (2001). Long distance community in the network society:
Contact and support beyond Netville. American Behavioral Scientist, 45(3), 436-
455.
Hardin, R. (1982). Collective Action. Washington, DC: RFF Press.
Haridakisa, P., & Hanson, G. (2009). Social interaction and co-viewing with YouTube:
Blending mass communication reception and social connection. Journal of
Broadcasting & Electronic Media, 53(2), 317-335.
Harkins, S. G., & Szymanski, K. (1989). Social loafing and group evaluation. Journal of
Personality and Social Psychology, 56, 934-941.
Harris, L., & Llanso, E. (2011). Facebook age and anonymity: civility vs. freedom of
speech. ABC News. Available at http://abcnews.go.com/Technology/facebook-
age-anonymity-civility-freedom-speech/story?id=13735349
Harrison, D. A., Mohammed, S., McGrath, J. E., Florey, A. T., & Vanderstoep, S. W.
(2003). Time matters in team performance: Effects of member familiarity,
entrainment, and task discontinuity on speed and quality. Personnel Psychology,
56(3), 633-645.
Haythornthwaite, C. (2002). Strong, weak, and latent ties and the impact of new media.
The Information Society, 18(5), 385-401.
Haythornthwaite, C. (2005). Social networks and internet connectivity effects.
Information, Communication & Society, 8(2), 125-147.
Haythornthwaite, C. (2006). Social networks and online community. In A. Joinson, K. Y.
A. McKenna, T. Postmes & U. Reips (Eds.), Oxford handbook of Internet
psychology (pp. 121-137). Oxford, UK: Oxford University Press.
186
Haythornthwaite, C. (2010). Social networks and information transfer. Encyclopedia of
Library and Information Sciences, 1(1), 4837-4847.
Healey, P. G. T., White, G., Eshghi, A., Reeves, A. J., & Light, A. (2007).
Communication spaces. Computer Supported Cooperative Work (CSCW), 17(2-
3), 169-193.
Heiferman, S. (2009, September 6). The pursuit of community. The New York Times, p.
BU7.
Heinz, M., & Rice, R. E. (2009). An integrated model of knowledge sharing in
contemporary communication environments. In C. Beck (Ed.), Communication
yearbook (Vol. 33, pp. 172-195). New York, NY: Routledge.
Hinds, P., Carley, K., Krackhardt, D., & Wholey, D. (2000). Balancing similarity,
competence, and familiarity: A study of how people choose workgroup members.
Organizational Behavior and Human Decision Processes, 81, 226-251
Hoadley, C. M., Xu, H., Lee, J. J., & Rosson, M. B. (2010). Privacy as information
access and illusory control: The case of the Facebook News Feed privacy outcry.
Electronic Commerce Research and Applications, 9(1), 50-60.
Hogg, M. A., & Turner, J. C. (1985). Interpersonal attraction, social identification and
psychological group formation. European Journal of Social Psychology, 15(1),
51-66.
Hollingshead, A. B., Fulk, J., & Monge, P. (2002). Fostering intranet knowledge sharing:
An integration of transactive memory and public goods approaches. In P. Hinds &
S. Kiesler (Eds.), Distributed Work (pp. 335-356). Cambridge, MA: The MIT
Press.
Hou, J. (2007). Price determinants in online auctions: A comparative study of eBay in
China and US. Journal of Electronic Commerce Research, 8(3), 172-183.
Humphreys, L., & Liao, T. (2011). Mobile geotagging: Reexamining our interactions
with urban space. Journal of Computer-Mediated Communication, 16(3), 407-
423.
Ingram, M. (2011, January 30). Hyper-local news: It's about the community or it fails.
GigaOm. http://gigaom.com/2011/03/04/hyper-local-news-its-about-the-
community-or-it-fails/.
Israel, S. (2009). Twitterville: How businesses can thrive in the new global
neighborhoods. New York, NY: Portfolio Hardcover.
187
Jarvenpaa, S. L., & Tractinsky, N. (1999). Consumer trust in an internet store: A cross-
cultural validation. Journal of Computer-Mediated Communication, 5(2).
Jenkins, H. (2006). Convergence Culture. New York, NY: New York University Press.
Jeppesen, L. B., & Frederiksen, L. (2006). Why do users contribute to firm-hosted user
communities?: The case of computer-controlled music instruments. Management
Science, 52(7), 1099-1115.
Jetten, J., Duck, J., Terry, D. J., & O'Brien, A. (2002). Being attuned to intergroup
differences in mergers: The role of aligned leaders for low-status groups.
Personality & Social Psychology Bulletin, 28(9), 1194-1201.
Jones, Q., Grandhi, S. A., Karam, S., Whittaker, S., Zhou, C., & Terveen, L. (2007).
Geographic ‘Place’ and ‘Community Information’ preferences Computer
Supported Cooperative Work (CSCW), 17(2-3), 137-167.
Jøsang, A., Ismail, R., & Boyd, C. (2007). A survey of trust and reputation systems for
online service provision. Decision Support Systems, 43(2), 618-644.
Julsrud, T. E., & Bakke, J. W. (2009). Trust, friendship, and expertise: The use of email,
mobile dialogues and SMS to develop and sustain relations in a distributed work
group. In R. Ling & S. W. Campbell (Eds.), The reconstruction of space and time:
Mobile communication practices (pp. 159-190). New Brunswick, NJ: Transaction
Publishers.
Junglas, I. A., & Watson, R. T. (2008). Location-based services: Evaluating user
perceptions of location-tracking and location-awareness services.
Communications of the ACM, 51(3), 65-69.
Kalyanaraman, S., & Sundar, S. S. (2008). Impression formation effects in online
mediated communication. In E. A. Konijn, S. Utz, M. Tanis & S. B. Barnes
(Eds.), Mediated interpersonal communication (pp. 217-233). New York, NY:
Routledge.
Kanfer, R. (1990). Motivation theory and industrial organizational psychology. In M. D.
Dunnette & L. Hough (Eds.), Handbook of industrial and organizational
psychology: Theory in industrial and organizational psychology (Vol. 1, pp. 75-
170). Palo Alto, CA: Consulting Psychologists Press.
Kankanhalli, A., Tan, B. C. Y., & Wei, K.-K. (2005). Contributing knowledge to
electronic knowledge repositories: An empirical investigation. MIS Quarterly,
29(1), 113-143.
188
Karau, S. J., & Williams, K. D. (1993). Social loafing: A meta-analytic review and
theoretical integration. Journal of Personality and Social Psychology, 65(4), 681-
706.
Katz, J. E., & Rice, R. (2002). Social consequences of internet use: Access, involvement,
and interaction. Cambridge, MA: MIT Press.
Katz, J. E., Rice, R. E., Acord, S., Dasgupta, K., & David, K. (2004). Personal mediated
communication and the concept of community in theory and practice. In P.
Kalbfleisch (Ed.), Communication and community, Communication Yearbook
(Vol. 28, pp. 315-371). Mahwah, NJ: Erlbaum.
Kellermann, K. (1985). Memory processes in media effects. Communication Research,
12(1), 83-131.
Kim, A. J. (2000). Community building on the web. New York, NY: Peachpit Press.
Kincaid, J. (2011, March 31). Facebook turns To StackOverflow to power developer
forums. TechCrunch. http://techcrunch.com/2011/08/24/facebook-turns-to-stack-
overflow-to-power-developer-forums/.
Kirkpatrick, M. (2011, January 10). Hyperlocal heartbreak: Why haven't neighborhood
news technologies worked out? ReadWriteWeb.
http://www.readwriteweb.com/archives/the_heartbreak_of_hyperlocal_news_aol_
scoops_up_ou.php.
Klimmt, C., & Hartmann, T. (2008). Mediated interpersonal communication in
multiplayer video games: implications for entertainment and relationship
management. In E. A. Konijn, S. Utz, M. Tanis & S. B. Barnes (Eds.), Mediated
interpersonal communication (pp. 309-330). New York, NY: Routledge.
Kollock, P. (1999a). The economics of online cooperation. In M. A. Smith & P. Kollock
(Eds.), Communities in cyberspace (pp. 220-239). New York, NY: Routledge.
Kollock, P. (1999b). The production of trust in online markets. Advances in Group
Processes, 16, 99-123.
Kollock, P., & Smith, M. A. (1999). Communities in cyberspace. In M. A. Smith & P.
Kollock (Eds.), Communities in cyberspace (pp. 3-28). New York, NY:
Routledge.
Koskela, T., Kostamo, N., Kassinen, O., Ohtonen, J., & Ylianttila, M. (2007, Nov. 4-9).
Towards context-aware mobile Web 2.0 service architecture. Paper presented at
the International Conference on Mobile Ubiquitous Computing, Systems, Services
and Technologies, UBICOMM '07.
189
Kudaravalli, S., & Faraj, S. (2008). The structure of collaboration in electronic networks.
Journal of the Association for Information Systems, 9(10/11), Article 1.
Kumar, R., Novak, J., & Tomkins, A. (2010). Structure and evolution of online social
networks. In P. S. S. Yu, J. Han & C. Faloutsos (Eds.), Link mining: Models,
algorithms, and applications (pp. 337-357). New York, NY: Springer
Kutner, M., Nachtscheim, C., Neter, J., & Li, W. (2004). Applied linear statistical
methods (5th ed.). New York, NY: McGraw-Hill/Irwin.
Lam, S. S. K., & Schaubroeck, J. (2000). Improving group decisions by better pooling
information: A comparative advantage of group decision support systems. Journal
of Applied Psychology, 85(4), 565-573.
Landman, B. (2011, November 20). Are you a V.I.P.? Check your Klout score. The New
York Times, p. ST13.
Latane, B. (1981). The psychology of social impact. American Psychologist, 36, 343-356.
Latane, B., & Liu, J. (1996). The intersubjective geometry of social space. Journal of
Communication, 46(4), 26-34.
Latane, B., Liu, J., Nowak, A., Bonevento, M., & Zheng, L. (1995). Distance matters:
Physical space and social impact. Personality & Social Psychology Bulletin, 21,
795-805.
Lee, H., Gurrin, C., Jones, G. J. F., & Smeaton, A. F. (2008). Interaction design for
personal photo management on a mobile device. In J. Lumsden (Ed.), Handbook
of research on user interface design and evaluation for mobile technology (Vol.
1-2, pp. 69-85). Hershey, PA: Information Science Reference/IGI Global.
Lenhart, A., & Madden, M. (2007). Teens, privacy and online social networks. Pew
Internet & American Life Project. Washington, DC: Pew Research Center
Lessig, L. (1999). Code and other laws of cyberspace. New York, NY: Basic Books.
Levy, S. (2009, November). Mob rule!: How users took over Twitter. Wired, 17(11).
Li, C., & Bernoff, J. (2008). Groundswell: Winning in a world transformed by social
technologies. Boston, MA: Harvard Business School Press.
Li, X., Guo, L., & Zhao, Y. E. (2008). Tag-based social interest discovery. Paper
presented at the 17th international conference on World Wide Web, New York,
NY.
190
Licoppe, C., & Smoreda, Z. (2006). Rhythms and ties: Toward a pragmatics of
technologically mediated sociability. In R. Kraut, M. Brynin & S. Kiesler (Eds.),
Computers, phones, and the internet: Domesticating information technology (pp.
296-314). New York, NY: Oxford University Press.
Ling, K., Beenen, G., Ludford, P., Wang, X., Chang, K., Li, X., et al. (2005). Using
social psychology to motivate contributions to online communities. Journal of
Computer-Mediated Communication, 10(4), Article 10.
Ling, R. (2008). New tech, new ties: How mobile communication is reshaping social
cohesion. Boston, MA: The MIT Press.
Ling, R., & Campbell, S. W. (Eds.). (2008). The reconstruction of space and time:
Mobile Communication practices. New York, NY: Transaction Publishers.
Lohr, S. (2011, March 26). In a new Web world, no application is an island. The New
York Times, p. BU3.
Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scale: Self-evaluation of
one's social identity. Personality and Social Psychology Bulletin, 18(3), 302-318.
Luo, X. (2002). Trust production and privacy concerns on the internet: A framework
based on relationship marketing and social exchange theory. Industrial Marketing
Management, 31(2), 111-118.
Ma, M., & Agarwal, R. (2007). Through a glass darkly: Information technology design,
identity verification, and knowledge contribution in online communities.
Information Systems Research, 18(1), 42-67.
MacKenzie, D., & Wajcman, J. (1999). The social shaping of technology (2nd ed.). New
York, NY: McGraw Hill Education.
MacKinnon, D. P. (2008). Introduction to statistical mediation analysis. Boca Raton, FL:
CRC Press.
Madden, M. (2012). Privacy management on social media sites. Pew Internet &
American Life Project. Washington, DC: Pew Research Center.
Markus, M. L. (2001). Toward a theory of knowledge reuse. Journal of Management
Information Systems, 18(1), 57-93.
Mathieson, K. (1991). Predicting user intentions: Comparing the Technology Acceptance
Model with the Theory of Planned Behavior. Information Systems Research, 2(3),
173-191.
191
McGrath, J. E. (1984). Groups: Performance and interaction. Englewood Cliffs, NJ:
Prentice Hall.
McMillan, R. (2012, February 21). Lord of the files: How GitHub tamed free software.
Wired, 20(2).
Metzger, M., Flanagin, A., & Medders, R. B. (2010). Social and heuristic approaches to
credibility evaluation online. Journal of Communication, 60, 413-439.
Meyrowitz, J. (1985). No sense of place: the impact of electronic media on social
behavior. New York, NY: Oxford University Press.
Miller, C. C. (2011, June 29). Another try by Google to take on Facebook. The New York
Times, p. B1.
Miller, C. C., & Stone, B. (2009, April 13). ‘Hyperlocal’ web sites deliver news without
newspapers. The New York Times, p. B1.
Miller, N., Resnick, P., & Zeckhauser, R. (2005). Eliciting informative feedback: The
peer-prediction method. Management Science, 51(9), 1359-1373.
Minch, R. P. (2004, January 5-8). Privacy issues in location-aware mobile devices. Paper
presented at the System Sciences, 2004. Proceedings of the 37th Annual Hawaii
International Conference Hawaii, USA.
Mischel, W. (1968). Personality and assessment. New York, NY: Wiley.
Mischel, W. (1973). Toward a cognitive social learning reconceptualization of
personality. Psychological Review, 80(4), 252-283.
Mislove, A., Viswanath, B., Gummadi, K. P., & Druschel, P. (2010). You are who you
know: Inferring user profiles in online social networks. Paper presented at the
Proceedings of the third ACM international conference on Web search and data
mining.
Mitch, W. (2008). Birth Of Location-Based Spam. InformationWeek, 1195, 22.
Mok, D., Wellman, B., & Basu, R. (2007). Did distance matter before the Internet?
Interpersonal contact and support in the 1970s. Social Networks, 29, 430-461.
Moon, J. Y., & Sproull, L. (2008). The role of feedback in managing the Internet-based
volunteer work force. Information Systems Research, 19(4), 494-515.
192
Moore, G. C., & Bendasat, I. (1991). Development of an instrument to measure the
perceptions of adopting an information technology innovation. Information
Systems Research, 2(3), 192-223.
Moore, K. (2011). 71% of online adults now use video-sharing sites. Pew Internet &
American Life Project. Washington, DC: Pew Research Center.
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship
marketing. Journal of Marketing, 58(3), 20-38.
Morris, M., Counts, S., Roseway, A., Hoff, A., & Schwarz, J. (2012). Tweeting is
believing? Understanding microblog credibility perceptions. Paper presented at
the CSCW 2012, Seattle, Washington.
Nahapiet, J., & Ghoshal, S. (1998). Social capital, intellectual capital and the
organizational advantage. Academy of Management Review, 23(2), 242-266.
Nakashima, E. (2008, July 12). When the phone goes with you, everyone else can tag
along. The Washington Post.
Nardi, B. (2005). Beyond bandwidth: Dimensions of connection in interpersonal
communication. Computer Supported Cooperative Work (CSCW), 14(1), 91-130.
Nie, N. H., Hillygus, D. S., & Erbring, L. (2002). Internet use, interpersonal relations,
and sociability: A time diary study. In B. Wellman & C. Haythornthwaite (Eds.),
The Internet in everyday life: Blackwell.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization
Science, 5(1), 14-37.
Oldenberg, R. (1999). The great good place: Cafes, coffee shops, bookstores, bars, hair
salons, and other hangouts at the heart of a community (3rd ed.). New York, NY:
Marlowe & Company.
Olivera, F., Goodman, P. S., & Tan, S. S.-L. (2008). Contribution behaviors in distributed
environments. MIS Quarterly, 32(1), 23-42.
Olson, M. (1965). The logic of collective action: Public goods and the theory of groups.
Boston, MA: Harvard University Press.
Orlikowski, W. J., & Iacono, C. S. (2001). Desperately seeking the "IT" in IT research: A
call to theorizing the IT artifact. Information Systems Research, 12(2), 121-134.
Ostrom, E. (1990). Governing the commons: The evolution of institutions for collective
action. New York, NY: Cambridge University Press.
193
Oulasvirta, A., Petit, R., Raento, M., & Tiitta, S. (2007). Interpreting and acting on
mobile awareness cues. Human-Computer Interaction. Special Issue: Awareness
systems design, 22(1-2), 97-135.
Paay, J., & Kjeldskov, J. (2008). Understanding situated social interactions: A case study
of public places in the city. Computer Supported Cooperative Work (CSCW),
17(2-3), 275-290.
Palan, E. (2012, March 12). How well do you know your neighbors? Philly Magazine.
Available at http://blogs.phillymag.com/the_philly_post/2012/02/27/philadelphia-
how-well-do-you-know-your-neighbors/.
Papacharissi, Z., & Olivera, M. d. F. (2012). Affective news and networked publics: The
rhythms of news storytelling on #Egypt. Journal of Communication, 66(2), 266-
282.
Pattison, K. (2010, March 25). Talking to the chief of Yelp, the site That businesses love
to hate. The New York Times, p. B9.
Pavlou, P. A., Liang, H., & Xue, Y. (2007). Understanding and mitigating uncertainty in
online exchange relationships: A principal-agent perspective. MIS Quarterly,
31(1), 105-136.
Pemberton, A. (2009, February 22). A load of Twitter. The Sunday Times.
Pentland, A. (2008). Honest signals: How they shape our world. Boston, MA: The MIT
Press.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. (2003). Common
method biases in behavioral research: A critical review of the literature and
recommended remedies. Journal of Applied Psychology, 88(5), 879-903.
Pool, I. d. S. (1983). Technologies of freedom. Boston, MA: Harvard University Press.
Postmes, T. (2006). The psychological dimensions of collective action, online. In A.
Joinson, K. Y. A. McKenna, T. Postmes & U. Reips (Eds.), Oxford handbook of
Internet psychology (pp. 165-184). Oxford, UK: Oxford University Press.
Postmes, T., & Spears, R. (2000). Refining the cognitive redefinition of the group:
Deindividuation effects in common bond vs. Common identity groups. In T.
Postmes, R. Spears, M. Lea & S. Reicher (Eds.), Side effects centre stage: Recent
developments in studies of deindividuation in groups (pp. 63-78). Amsterdam, the
Netherlands: KNAW.
194
Postmes, T., Spears, R., & Lea, M. (1998). Breaching or building social boundaries?
SIDE-effects of computer-mediated communication. Communication Research,
25, 689-715.
Prentice, D. A., Miller, D. T., & Lightdale, J. R. (1994). Asymmetries in attachments to
groups and to their members: Distinguishing between common identity and
common-bond groups. Personality and Social Psychology Bulletin, 20(5), 484-
493.
Putnam, R. (2000). Bowling alone: The collapse and revival of American community.
New York, NY: Simon & Schuster.
Quan-Hasse, A., Wellman, B., Witte, J. C., & Hampton, K. (2002). Capitalizing on the
Net: Social contact, civic engagement and sense of community. In B. Wellman &
C. Haythornthwaite (Eds.), The Internet in everyday life. Malden, MA: Blackwell
Publishing.
Raento, M., & Oulasvirta, A. (2008). Designing for privacy and self-presentation in
social awareness. Personal and Ubiquitous Computing, 12(7), 527-542.
Rainie, L., Fox, S., Lenhart, A., & Spooner, T. (2000). Tracking online life: How women
use the Internet to cultivate relationships with family and friends. Washington,
DC: Pew Internet & American Life Project.
Ravaja, N., Saari, T., Turpeinen, M., Laarni, J., Salminen, M., & Kivikangas, M. (2006).
Spatial presence and emotions during video game playing: Does it matter with
whom you play? Presence: Teleoperators and Virtual Environments, 15(4), 381-
392.
Rawassizadeh, R. (2011). Towards sharing life-log information with society. Behavior &
Information Technology, 1, 1-11.
Ren, Y., Kraut, R., & Kiesler, S. (2007). Applying common identity and bond theory to
design of online communities. Organization Studies, 28(3), 377-408.
Resnick, P. (2001). Beyond bowling together: SocioTechnical capital. In J. M. Carroll
(Ed.), HCI in the new millennium (pp. 247-272). Boston, MA: Addison-Wesley.
Resnick, P., & Zeckhauser, R. (2002). Trust among strangers in Internet transactions:
Empirical analysis of eBay’s reputation system. In M. R. Baye (Ed.), The
economics of the Internet and E-commerce: Volume 11 of advances in applied
microeconomics. New York, NY: Elsevier Science.
Resnick, P., Zeckhauser, R., Friedman, E., & Kuwabara, K. (2000). Reputation systems.
Communications of the ACM, 43(12), 45-48.
195
Resnick, P., Zeckhauser, R., Swanson, J., & Lockwood, K. (2006). The value of
reputation on eBay: A controlled experiment. Experimental Economics, 9(79-
101), 79.
Rheingold, H. (2002). The smart mobs: the next social revolution. Cambridge, MA:
Perseus.
Ringle, C. M., Wende, S., & Will, A. (2005). SmartPLS (Version 2.0 (beta)). Hamburg,
Germany: SmartPLS.
Rivilin, G. (2011, May). Does Quora really have all the answers? Wired, 19(5).
Robert, L. P., Dennis, A. R., & Ahuja, M. K. (2008). Social capital and knowledge
integration in digitally enabled teams. Information Systems Research, 19(3), 314-
334.
Roberts, J., Hann, I.-H., & Slaughter, S. (2006). Understanding the motivations,
participation, and performance of open source software developers: A
longitudinal study of the apache projects. Management Science, 52(7), 984-999.
Rogers, P., & Lea, M. (2005). Social presence in distributed group environments: The
role of social identity. Behaviour & Information Technology, 24(1), 151-158.
Rogoff, B. (1995). Observing sociocultural activity on three planes: Participatory
appropriation, guided participation, and apprenticeship. In J. V. Wertsch, P. D.
Rio & A. Alvarez (Eds.), Sociocultural studies of mind. New York, NY:
Cambridge University Press.
Rowan, D. (2012, March 20). For LinkedIn founder Reid Hoffman, relationships rule the
world. Wired, 20(3).
Sabherwal, R., & Robey, D. (1995). Reconciling variance and process strategies for
studying ISD. Information Systems Research, 6(4), 303-327.
Scarpetta, F. (2008). Practices to display social presence: A study in a shared mediated
environment. PsychNology Journal. Special Issue: Presence. Where do you feel
you are today?, 6(1), 27-59.
Schifanella, R., Barrat, A., Cattuto, C., Markines, B., & Menczer, F. (2010). Folks in
folksonomies: Social link prediction from shared metadata. Paper presented at the
Proceedings of the 3rd ACM international conference on Web search and data
mining.
Schmandt, C., & Marmasse, N. (2004). User-centered location awareness. Computer,
37(10), 110-111.
196
Schmitz, J., & Fulk, J. (1991). Organizational colleagues, media richness and electronic
mail. Communication Research, 18(4), 487-523.
Sengupta, S., & Miller, C. C. (2012, February 2). 'Social Mission’ Vision Meets Wall
Street. The New York Times, p. B1.
Shaw, B., Scheufele, D. A., & Catalano, S. (2007). The role of presence awareness in
organizational communication: An exploratory field experiment. Behaviour &
Information Technology, 26(5), 377-384.
Shirky, C. (2008). Here comes everybody: The power of organizing without
organizations. New York, NY: Penguin Press HC.
Shirky, C. (2010). Cognitive surplus: Creativity and generosity in a connected age. New
York, NY: The Penguin Press HC.
Shirky, C. (2011). The political power of social media. Foreign Affairs,
January/February.
Shumaker, S., & Brownell, A. (1984). Toward a theory of social support: Closing
conceptual gaps. Journal of Social Issues, 40(4), 11-36.
Sieloff, C. G. (1997). “If only HP knew what HP knows”: the roots of knowledge
management at Hewlett-Packard. Journal of Knowledge Management, 3(1), 47-
53.
Simpson, J., & Weiner, E. (Eds.). (1989). Oxford English dictionary. Oxford, UK.
Smith, A. (2011a). Smartphone adoption and usage. Pew Internet & American Life
Project. Washington, DC: Pew Research Center
Smith, A. (2011b). Twitter update 2011. Pew Internet & American Life Project.
Washington, DC: Pew Research Center
Smith, A. (2011c). Why Americans use social media. Pew Internet & American Life
Project. Washington, DC: Pew Research Center
Smith, E., Consolvo, S., & Lamarca, A. (2005). The Drop: Pragmatic problems in the
design of a compelling, pervasive Game. ACM Computers in Entertainment, 3(3),
4C.
Spagnolli, A., & Gamberini, L. (2005). A place for presence: Understanding the human
involvement in mediated interactive environments. PsychNology Journal: Special
Issue on Space, Place & Technology, 3(1), 6-15.
197
Spears, R., Lea, M., Corneliussen, R. A., Postmes, T., & Ter Haar, W. (2002). Computer-
mediated communication as a channel for social resistance: The strategic side of
SIDE. Small Group Research, 33, 555-574.
Spears, R., Lea, M., & Postmes, T. (2006). Computer-mediated communication and
social identity. In A. Joinson, K. Y. A. McKenna, T. Postmes & U. Reips (Eds.),
Oxford handbook of Internet psychology (pp. 253-269). Oxford, UK: Oxford
University Press.
Spoor, J. R., & Kelly, J. R. (2004). The evolutionary significance of affect in groups:
Communication and group bonding. Group Processes Intergroup Relations, 7(4),
398-412.
Streefkerk, J. W., van Esch-Bussemakers, M. P., Neerincx, M. A., & Looije, R. (2008).
Evaluating context-aware mobile interfaces for professionals. In J. Lumsden
(Ed.), Handbook of research on user interface design and evaluation for mobile
technology (Vol. 1-2, pp. 759-779). Hershey, PA: Information Science
Reference/IGI Global.
Surowiecki, J. (2005). The wisdom of crowds. New York, NY: Anchor.
Szmigin, I., & Reppel, A. E. (2004). Internet community bonding: the case of
macnews.de. European Journal of Marketing, 38(5/6), 626-640.
Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S.
Worchel & W. Austin (Eds.), Psychology of intergroup relations (pp. 7-24).
Chicago, IL: Nelson-Hall.
Takhteyev, Y., Gruzd, A., & Wellman, B. (2012). Geography of Twitter networks. Social
Networks, 34, 73-81.
Tanis, M. (2008). What makes the internet a place to seek social support? In E. A.
Konijn, S. Utz, M. Tanis & S. B. Barnes (Eds.), Mediated interpersonal
communication (pp. 290-308). New York: Routledge.
Tanis, M., & Postmes, T. (2003). Social cues and impression formation in CMC. Journal
of Communication, 53, 676-693.
Tapscott, D., & Williams, A. D. (2008). Wikinomics: How mass collaboration changes
everything (Expanded ed.). New York, NY: Portfolio.
Tenenhaus, M., Vinzi, V. E., Chatelin, Y.-M., & Lauro, C. (2005). PLS path modeling.
Computational statistics & data analysis, 48(1), 159-205.
198
Terveen, L., & McDonald, D. W. (2005). Social matching: A framework and research
agenda. ACM Transactions on Computer-Human Interaction, 12(3), 401-434.
Tiwana, A., & McLean, E. R. (2005). Expertise integration and creativity in information
systems development. Journal of Management Information Systems, 22(1), 13-43.
Tong, S. T., Heide, B. V. D., Langwell, L., & Walther, J. (2008). Too much of a good
thing?: The relationship between number of friends and interpersonal impressions
on Facebook. Journal of Computer-Mediated Communication, 13, 531-549.
Tung, F.-C., Chang, S.-C., & Chou, C.-M. (2008). An extension of trust and TAM model
with IDT in the adoption of the electronic logistics information system in HIS in
the medical industry. International Journal of Medical Informatics, 77(5), 324-
335.
Utz, S. (2008). Social identification within virtual communities. In E. A. Konijn, S. Utz,
M. Tanis & S. B. Barnes (Eds.), Mediated interpersonal communication (pp. 252-
270). New York, NY: Routledge.
Valenti, M. A., & Rockett, T. (2008). The effects of demographic differences on forming
intragroup relationships. Small Group Research, 39(2), 179-202.
Van Grove, J. (2012, March 9). Loopt was a lemon, dropping to under a million daily
active users prior to sale. VentureBeat. Available at
http://venturebeat.com/2012/03/09/loopt-dau/
van Knippenberg, D., De Dreu, C. K. W., & Homan, A. C. (2004). Work group diversity
and group performance: an integrative model and research agenda. Journal of
Applied Psychology, 89, 1008-1022.
Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the Technology
Acceptance Model: Four longitudinal field studies. Management Science, 46(2),
186-204.
Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, D. M. (2003). User acceptance of
information technology: Toward a unified view. Management Information
Systems Quarterly, 27(3), 425-478.
Vogiazou, Y., Raijmakers, B., Geelhoed, E., Reid, J., & Eisenstadt, M. (2007). Design for
emergence: experiments with a mixed reality urban playground game. Personal
and Ubiquitous Computing, 11(1), 45-58.
Walther, J. (1996). Computer-mediated communication: Impersonal, interpersonal, and
hyperpersonal interaction. Communication Research, 23(1), 3-43.
199
Walther, J., & Parks, M. R. (2002). Cues filtered out, cues filtered in: Computer-mediated
communication and relationships. In M. L. Knapp & J. A. Daly (Eds.), Handbook
of interpersonal communication (3rd ed., pp. 529-563). Thousand Oaks, CA:
Sage.
Walther, J., Van der Heide, B., Hamel, L. M., & Shulman, H. C. (2009). Self-generated
versus other-generated statements and impressions in computer-mediated
communication: A test of Warranting theory using Facebook. Communication
Research, 36(2), 229-253.
Walther, J., Van der Heide, B., Kim, S., Westerman, D., & Tong, S. T. (2008). The role
of friends' appearance and behavior on evaluations of individuals on Facebook:
Are we known by the company we keep? Human Communication Research,
34(1), 28-49.
Wasko, M. M., & Faraj, S. (2005). Why should I share? Examining social capital and
knowledge contribution in electronic networks of practice. MIS Quarterly, 29(1),
35-57.
Wasko, M. M., & Teigland, R. (2002). The provision of online public goods: Examining
social structure in a network of practice. Paper presented at the 23rd Annual
International Conference on Information Systems, Barcelona, Spain.
Webb, E. J., Campbell, D. T., & Schwartz, R. D. (1999). Unobtrusive measures (Revised
ed.). Thousand Oaks, CA: Sage.
Wellman, B. (1997). An electronic group is virtually a social network. In S. Kiesler (Ed.),
Cultures of the Internet (pp. 179-205). Marwah, NJ: Lawrence Erlbaum.
Wellman, B., & Gulia, M. (1999). Virtual communities as communities: Net surfers don't
ride alone. In M. A. Smith & P. Kollock (Eds.), Communities in cyberspace (pp.
167-194). New York, NY: Routledge.
Wellman, B., Quan-Hasse, A., Boase, J., Chen, W., Hampton, K., & Isla de Diaz, I.
(2003). The social affordances of the Internet for networked individualism.
Journal of Computer-Mediated Communication, 8(3).
Westin, A. F. (1967). Privacy and freedom. New York, NY: Athenaum.
Whitmeyer, J. M. (2000). Effect of positive reputation systems. Social Science Research,
29(1), 188-207.
Williams, D. (2006). On and off the 'net: Scales for social capital in an online era.
Journal of Computer-Mediated Communication, 11(2), Article 11.
200
Wortham, J. (2009, October 19). Face-to-face socializing starts with a mobile post. The
New York Times, p. B1.
Xu, H., Teo, H. H., Tan, B. C. Y., & Agarwal, R. (2010). The role of push-pull
technology in privacy calculus: The case of Location-Based Services. Journal of
Management Information Systems, 26(3), 137-176.
Yi, M. Y., Jackson, J. D., Park, J. S., & Probst, J. C. (2006). Understanding information
technology acceptance by individual professionals: Toward an integrative view.
Information & Management, 43, 350-363.
Yu, B., & Fei, H. (2009). Modeling social cascade in the Flickr social network. Paper
presented at the Proceedings of the 6th international conference on Fuzzy systems
and knowledge discovery.
Zammuto, R. F., Griffith, T. L., Majchrzak, A., Doughetry, D. J., & Faraj, S. (2007).
Information technology and the changing fabric of organization. Organization
Science, 18(5), 749-762.
Zhang, J., Ghorbani, A. A., & Cohen, R. (2007). A familiarity-based trust model for
effective selection of sellers in multiagent e-commerce systems. International
Journal of Information Security, 6, 333-344.
Zhou, M., Dresner, M., & Windle, R. J. (2008). Online reputation systems: Design and
strategic practices. Decision Support Systems, 44(4), 785.
Zhu, Y. (2010). Measurement and analysis of an online content voting network: a case
study of Digg. Paper presented at the Proceedings of the 19th international
conference on World Wide Web.
Zickuhr, K., & Smith, A. (2011). 28% of American adults use mobile and social location-
based services. Pew Internet & American Life Project. Washington, DC: Pew
Research Center
201
Appendices
Appendix A: Reliability, Convergent Validity and Discriminant Validity for the
Measurement Model
Table 2: Reliability

Cronbach’s Alpha Composite Reliability
Use of Reputation Systems (URS)
0.8548 0.8548
Use of Social Matching Systems (USMS)
0.8454 0.9676
Use of Privacy Systems (UPS)
0.9676 0.7413
Generalized Reciprocity (GR)
0.9023 0.9157
Interpersonal Bonds (IB)
0.8601 0.9213
Social Identification (SI)
0.9462 0.9760
Bonding Social Capital (BON)
0.9245 0.9325
Bridging Social Capital (BRI)
0.9071 0.9125
Personal Innovativeness in IT (PIIT)
0.8711 0.8307
Perceived Quality of Sharing (QUAL)
0.9334 0.9947

Table 3: Convergent Validity

Use of
Reputation
Systems
(URS)
Use of
Social
Matching
Systems
(USMS)
Use of
Privacy
Systems
(UPS)
Generalized
Reciprocity
(GR)
Inter-
personal
Bonds
(IB)
Social
Identification
(SI)
Bonding
Social
Capital
(BON)
Bridging
Social
Capital
(BRI)
Personal
Innovative-
ness in IT
(PIIT)
Perceived
Quality of
Sharing
(QUAL)
URS1 0.8026 0.2266 -0.0357 0.3897 0.4700 0.2661 -0.1534 0.3341 0.0123 0.3364
URS2 0.7937 0.3794 0.2887 0.3119 0.4248 0.2758 -0.1960 0.4435 0.0331 0.3530
URS3 0.7825 0.4086 0.1634 0.2749 0.2333 0.1020 -0.0701 0.4141 0.1728 0.1683
URS4 0.7644 0.5140 -0.0360 0.1318 0.3686 0.3508 0.0000 0.5346 -0.1241 0.1838
USMS1 0.3559 0.8421 0.2880 0.4208 0.3028 0.1867 0.0185 0.2557 0.1413 0.2083
USMS2 0.5145 0.9914 0.1976 0.4498 0.5177 0.3884 -0.0119 0.3177 0.0500 0.3019
USMS3 0.1618 0.8143 0.2502 0.1686 0.1849 0.2368 0.0559 0.2436 -0.0774 0.3788
USMS4 0.4858 1.0853 0.4018 0.4733 0.4710 0.3779 0.0015 0.4218 0.0332 0.3576
UPS1 -0.0342 0.1626 0.6388 0.1187 0.0365 0.0257 0.2371 0.1010 -0.1314 0.0555
UPS2 0.2075 0.2530 0.8770 0.3600 0.1900 0.1072 0.0442 0.3124 0.0761 0.2158
UPS3 0.0119 0.2361 0.5620 0.0375 0.0190 0.0781 0.0541 0.1016 -0.0320 -0.0209
GR1 0.3335 0.4842 0.2773 0.8702 0.4870 0.4666 0.0503 0.4617 0.2617 0.4716
GR2 0.3660 0.4654 0.2939 0.8389 0.6421 0.5439 0.0583 0.5140 0.1812 0.4394
GR3 0.3070 0.3021 0.2035 0.8759 0.5316 0.4336 -0.0878 0.3657 0.1650 0.5143
GR4 0.3050 0.2162 0.1492 0.8556 0.5090 0.3238 -0.2366 0.2961 0.0878 0.4189
GR5 0.2575 0.2495 0.3071 0.6872 0.4198 0.3311 -0.0773 0.3112 0.2158 0.5072
IB1 0.4185 0.2715 0.1785 0.6624 0.9495 0.6661 -0.1616 0.4202 0.0820 0.5543
202
Table 3: Convergent Validity (Continued)

Use of
Reputation
Systems
(URS)
Use of
Social
Matching
Systems
(USMS)
Use of
Privacy
Systems
(UPS)
Generalized
Reciprocity
(GR)
Inter-
personal
Bonds
(IB)
Social
Identification
(SI)
Bonding
Social
Capital
(BON)
Bridging
Social
Capital
(BRI)
Personal
Innovative-
ness in IT
(PIIT)
Perceived
Quality of
Sharing
(QUAL)
IB2 0.4577 0.2922 0.1595 0.6707 0.9187 0.6365 -0.2015 0.4301 -0.0022 0.4785
IB3 0.5375 0.6042 0.1237 0.3205 0.8298 0.6595 -0.0612 0.4570 -0.0194 0.4500
IB4 0.2373 0.2386 0.0369 0.5592 0.8442 0.5749 -0.0891 0.2751 0.1167 0.5294
SI1 0.3129 0.3876 0.1761 0.6067 0.6219 0.9895 0.1201 0.5408 0.2814 0.6180
SI2 0.2217 0.3197 0.1612 0.5079 0.6312 0.9865 0.0593 0.3999 0.2108 0.6578
SI3 0.3366 0.1928 -0.0419 0.4349 0.6239 1.0106 0.0093 0.4858 0.1549 0.4598
SI4 0.2963 0.1824 -0.0795 0.4553 0.6841 1.0472 -0.0265 0.5819 0.1043 0.5045
SI5 0.3150 0.3119 0.0265 0.3976 0.6156 0.8307 0.0952 0.4484 0.0926 0.3906
SI6 0.1887 0.3199 0.1581 0.4031 0.5690 0.7663 0.1289 0.4575 0.1875 0.4052
SI7 0.3048 0.4257 0.1882 0.4501 0.6809 0.7972 0.1106 0.4603 -0.0113 0.3146
BON1 -0.1736 0.0449 0.0062 0.0000 -0.1948 0.0730 0.7890 0.0647 0.0338 -0.2251
BON2 -0.0573 0.1354 0.1611 0.0598 -0.0028 0.1149 0.7566 0.1880 -0.0653 -0.0768
BON3 -0.1933 0.0386 0.2053 0.0682 -0.1140 0.0334 0.6972 0.0712 -0.0550 -0.1126
BON4 -0.1614 0.1058 0.0705 -0.0081 -0.0857 0.0065 0.6164 -0.0202 -0.0347 -0.0291
BON5 -0.2753 -0.0193 0.0667 -0.2042 -0.1935 0.0946 1.0412 0.1245 0.0157 -0.2435
BON6 -0.1262 -0.0197 0.0968 -0.0633 -0.0213 0.1746 0.9644 0.1227 0.0082 -0.1511
BON7 0.0837 -0.0651 0.2196 0.0044 -0.1544 0.0282 0.6850 0.2937 0.1576 -0.0756
BON8 -0.1336 -0.0276 0.1418 -0.1160 -0.1260 0.0223 1.0001 0.1413 -0.0655 -0.3950
BRI1 0.5078 0.3076 0.2047 0.2743 0.4061 0.5171 0.0266 0.8349 -0.0538 0.3684
BRI2 0.3390 0.3153 0.2095 0.4263 0.2927 0.3761 0.1954 0.7268 0.0330 0.1855
BRI3 0.4576 0.1900 0.2589 0.4418 0.4141 0.4134 0.1780 0.8470 0.1439 0.1842
BRI4 0.4083 0.3527 0.2573 0.4274 0.4121 0.3566 0.1844 0.7109 0.1643 0.2780
BRI5 0.3306 0.2366 0.1369 0.4465 0.3937 0.4635 0.0151 0.8051 -0.0501 0.4488
BRI6 0.4189 0.2311 0.2540 0.2259 0.2854 0.3125 0.1229 0.7984 0.2571 0.1686
BRI7 -0.1736 0.0449 0.0062 0.0000 -0.1948 0.0730 0.0266 0.8349 0.0338 -0.2251
PIIT1 0.1142 0.0213 -0.0144 0.2452 0.1045 0.1709 0.0032 0.0684 0.8442 0.2515
PIIT2 -0.0446 0.0625 0.0329 0.0944 -0.1487 0.0049 -0.0225 0.0720 0.8475 0.0612
PIIT3 0.0570 0.1925 0.0415 0.1398 0.0051 0.1171 0.1956 0.0948 0.6396 0.0409
PIIT4 0.0599 0.0187 -0.0581 0.1452 0.1225 0.2013 -0.0093 0.0975 0.6201 0.1953
QUAL1 0.3028 0.3626 0.2251 0.5181 0.5185 0.4521 -0.1905 0.3225 0.1714 1.0224
QUAL2 0.3070 0.2997 0.1688 0.4918 0.4829 0.4434 -0.2787 0.3233 0.1915 0.9092
QUAL3 0.3442 0.2670 0.1551 0.5716 0.5649 0.5273 -0.2789 0.3715 0.2604 1.0885
QUAL4 0.4438 0.3562 0.1009 0.6366 0.7032 0.5697 -0.2435 0.3458 0.2700 0.9275

203
Table 4: Discriminant Validity
AVE
Use of
Reputation
Systems
(URS)
Use of
Social
Matching
Systems
(USMS)
Use of
Privacy
Systems
(UPS)
Generalized
Reciprocity
(GR)
Inter-
personal
Bonds
(IB)
Social
Identification
(SI)
Bonding
Social
Capital
(BON)
Bridging
Social
Capital
(BRI)
Personal
Innovative
-ness in IT
(PIIT)
Perceived
Quality of
Sharing
(QUAL)
URS 0.6060
USMS 0.4363 0.8832
UPS 0.1234 0.2976 0.5177
GR 0.3810 0.2976 0.2976 0.6865
IB 0.4886 0.4259 0.1452 0.6286 0.7470
SI 0.3019 0.3355 0.1014 0.5133 0.6951 0.8547
BON -0.1626 0.0088 0.1480 0.0633 -0.1448 0.0786 0.6468
BRI 0.5226 0.3394 0.2749 0.4766 0.4646 0.5188 0.1587 0.5999
PIIT 0.0682 0.0435 -0.0168 0.2218 0.0477 0.1699 0.0074 0.1003 0.5562
QUAL 0.3592 0.3250 0.1604 0.5671 0.5825 0.5370 -0.2519 0.3466 0.2305 0.9793

204
Appendix B: Measurement Items
Use of Reputation Systems (URS)*
1. I think that other people considered my ranking (reputation) when they interacted
with me.
2. I think that other people browsed my profile to find out more about me
3. I think that other people checked my past history when they interact with me.
4. On Twitter/Voilah/Voilai, people rely on reputation information in profiles for their
social interaction.
Use of Social Matching Systems (USMS)**
1. How often did you use hashtags/tags to find about topics that are relevant to your
interests in the past two weeks?
2. How often did you use hashtags/tags to find people who might share similar interests
with you in the past two weeks?
3. How useful was the feature "Trending Topics" (Twitter)/”Interests”(Voilah/Voilai)
for you to see what was popular now on Twitter/Voilah/Voilai in the past two weeks?
4. How often did you tag your posts to make sure they match other people’s interests in
the past 2 weeks?
Use of Privacy Systems (UPS)*
1. Privacy control helped me control who can access my personal information
2. I changed my privacy settings in online social networks.
3. Privacy settings had an influence on what information I can get and what people I
meet online.
205
Generalized Reciprocity (GR)*
1. When I shared some information on Twitter/Voilah/Voilai, I expected somebody to
respond when I'm in need of other information.
2. When I shared something useful on Twitter/Voilah/Voilai, I expected to get some
useful information back when I need it.
3. When I shared some information on Twitter/Voilah/Voilai, I believed that my queries
for other information will be answered in future.
4. I knew that other Twitter/Voilah/Voilai users will help me with useful information, so
it's only fair to help other members by sharing something I know.
5. I trusted that Twitter/Voilah/Voilai has a system that fairly records and reflects my
good behavior in this community.
Interpersonal Bonds (IB)*
1. I felt close to the other users in the Twitter/Voilah/Voilai community.
2. Many users in the Twitter/Voilah/Voilai community have influenced my thoughts and
behaviors.
3. Many of my friends were also users in the Twitter/Voilah/Voilai community.
4. In many ways, I felt I am similar to the other users in the Twitter/Voilah/Voilai
community.
Social Identification (SI)*
1. I identified myself with the users of the Twitter/Voilah/Voilai community.
2. I saw myself as a member of the Twitter/Voilah/Voilai community.
3. I would have felt a loss if this community were no longer available.
206
4. I really cared about the fate of this community.
5. I felt a great deal of loyalty to this community.
6. The things I did and said in the Twitter/Voilah/Voilai community are an important
reflection of who I am.
7. In general, belonging to Twitter/Voilah/Voilai has been an important part of my self-
image.
Bonding Social Capital (BON)*
1. There were several people in my strong-tie networks I trust to help solve my
problems.
2. There was someone in my strong-tie networks I could turn to for advice about making
very important decisions.
3. When I felt lonely, there were several people in my strong-tie networks I can talk to.
4. If I needed an emergency loan of $500, I knew someone in my strong-tie networks I
can turn to.
5. The people I interacted with in my strong-tie networks would put their reputation on
the line for me.
6. The people I interacted with in my strong-tie networks would be good job references
for me.
7. The people I interacted with in my strong tie networks would share their last dollar
with me.
8. The people I interacted with in my strong-tie networks would help me fight an
injustice.
207
Bridging Social Capital (BRI)*
1. Interacting with people outside my strong-tie networks made me interested in things
that happened outside of my town.
2. Interacting with people outside my strong-tie networks made me want to try new
things.
3. Interacting with people outside my strong-tie networks made me interested in what
people unlike me are thinking.
4. Interacting with people outside my strong-tie networks made me feel like part of a
larger community.
5. Interacting with people outside my strong-tie networks made me feel connected to the
bigger picture.
6. Interacting with people outside my strong-tie networks gave me new people to talk to.
7. I came in contact with new people all the time.
Personal Innovativeness in IT (PIIT)*
1. If I heard about a new technology, I would look for ways to experiment with it.
2. Among my peers, I am usually the first to try out new information technologies.
3. In general, I am hesitant to try out new technologies (reverse-coded).
4. I like to experiment with new information technologies.
Quality of Information Contribution (QUAL)*
1. I helped other people on Twitter/Voilah/Voilai who need help and information from
other users.
2. I took an active part in making Twitter/Voilah/Voilai a vibrant community
208
3. I have contributed useful information on Twitter/Voilah/Voilai.
4. I have contributed information to other Twitter/Voilah/Voilai users that resulted in
their development of new ideas or perspectives.

* On a scale of five, one is “Strongly Disagree”, two is “Disagree”, three is “Neither
Disagree nor agree”, four is “Agree” and five is “Strongly Agree”
** On a scale of five, one is “Never”, two is “Rarely”, three is “Sometimes”, four is
“Often” and five is “Frequently”
209
Appendix C: Un-hypothesized Paths
Table 5: Test Results for Un-hypothesized Paths
Paths Coefficient
Standard
Error
t-Statistic
(degrees of
freedom)
Significance
Use of Reputation Systems – Interpersonal
Bonds **
0.49 0.07 6.92 (94) p < .01
Use of Reputation Systems – Social
Identification *
0.33 0.13 2.56 (92) p < .05
Use of Reputation Systems – Bonding Social
Capital
-0.29 0.08 1.91 (94) p = 0.06
Use of Reputation Systems – Bridging Social
Capital **
0.50 0.12 4.15(94) p < .01
Use of Reputation Systems – Quality of
Contribution
0.24 0.13 1.80 (94) p = 0.07
Use of Reputation Systems – Increase in the
Quantity of Contribution
0.10 0.11 1.45 (94) p = 0.15
Use of Social Matching Systems –
Generalized Reciprocity **
0.36 0.08 4.26 (94) p < .01
Use of Social Matching Systems – Bonding
Social Capital
0.19 0.16 1.17 (94) p = 0.25
Use of Social Matching Systems – Bridging
Social Capital
0.11 0.09 1.22 (94) p = 0.23
Use of Social Matching Systems – Quality of
Contribution
0,20 0.13 1.55 (94) p =0.13
Use of Social Matching Systems – Increase in
the Quantity of Contribution
-0.22 0.12 1.84 (94) p = 0.07
Use of Privacy Systems – Generalized
Reciprocity
0.25 0.15 1.69 (94) p = 0.09
Use of Privacy Systems – Interpersonal Bonds 0.11 0.08 1.39 (94) p = 0.17
Use of Privacy Systems – Social Identification 0.05 0.14 0.33 (94) p = 0.74
Use of Privacy Systems – Quality of
Contribution
0.11 0.11 1.02 (94) p = 0.31
Use of Privacy Systems – Increase in the
Quantity of Contribution
0.03 0.13 0.25 (94) p = 0.80

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Building personal wellness communities: meaningful play in the everyday life of a network society

PDF

Social media and health: social support and social capital on pregnancy-related social networking sites

PDF

Doing family on Facebook: connecting communicative affordances of mobile media, parental practices, and social capital

PDF

Relationship formation and information sharing to promote risky health behavior on social media

PDF

Social motivation and credibility in crowdfunding

PDF

Identity, trust, and credibility online: evaluating contradictory user-generated information via the warranting principle

PDF

Benefits of social networking in online social support groups

PDF

The patterns, effects and evolution of player social networks in online gaming communities

PDF

Video game live streaming and the perception of female gamers

PDF

The formation and influence of online health social networks on social support, self-tracking behavior and weight loss outcomes

PDF

Democrats vs. Republicans: how Twitter users share and react to political fact-checking messages

PDF

The social groups approach to quitting smoking: An examination of smoking cessation online and offline through the influence of social norms, social identification, social capital and social support

PDF

Motivation for participating in virtual religious communities: developing social capital

PDF

Creating community online: the effects of online social networking communities on college students' experiences

PDF

The effect of Twitter social support on health outcomes and its mediators: a randomized controlled trial of a social support intervention on Twitter for patients affected by cancer

PDF

Crisis and stasis: understanding change in online communities

PDF

The evolution of multilevel organizational networks in an online gaming community

PDF

The role of social control in promoting healthy eating behavior among Chinese immigrants: an ecological approach

PDF

Gamification + HCI + CMC: effects of persuasive video games on consumers’ mental and physical health

PDF

Ecology and network evolution in online innovation contest crowdsourcing

Asset Metadata

Creator Xiong, Li (author)

Core Title The social meaning of sharing and geocoding: features and social processes in online communities

School Annenberg School for Communication

Degree Doctor of Philosophy

Degree Program Communication

Publication Date 11/05/2012

Defense Date 06/22/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag field experiment,geographic proximity,geolocation,information sharing,OAI-PMH Harvest,online communities,privacy,reputation systems,social capital

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Williams, Dmitri (committee chair), Hollingshead, Andrea B. (committee member), Majchrzak, Ann (committee member), McLaughlin, Margaret L. (committee member)

Creator Email leoxbear@gmail.com,lixiong@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-108358

Unique identifier UC11289518

Identifier usctheses-c3-108358 (legacy record id)

Legacy Identifier etd-XiongLi-1268.pdf

Dmrecord 108358

Document Type Dissertation

Rights Xiong, Li

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA