tweeting to feel connected: a model for ... - wordpress.com · tweeting to feel connected: a model...
Post on 01-Jun-2020
5 Views
Preview:
TRANSCRIPT
Tweeting to Feel Connected: A Model for Social Connectedness in Online
Social Networks
Christoph Riedl, Felix Köbler, Suparna Goswami, and Helmut Krcmar
Christoph Riedl, Institute for Quantitative Social Science, Harvard University,
criedl@iq.harvard.edu
Felix Köbler, Information Systems, Technische Universität München, felix.koebler@in.tum.de
Suparna Goswami, Information Systems, Technische Universität München,
suparna.goswami@in.tum.de
Helmut Krcmar, Information Systems, Technische Universität München, krcmar@in.tum.de
Accepted for Publication at
International Journal of Human-Computer Interaction
(Original Research Article, regular issue)
Please address all correspondence on this paper to Christoph Riedl
Phone: (1) 857 204 9655; Email: criedl@iq.harvard.edu
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
1
Tweeting to Feel Connected: A Model for Social Connectedness in Online
Social Networks1
Abstract. Social connectedness is an indicator of the extent to which people can realize various
network benefits and is therefore a source of social capital. Using the case of Twitter, we develop and test
a theoretical model of social connectedness based on the functional and structural characteristics of
peoples’ communication behavior within a online social network. We investigate how social presence,
social awareness, and social connectedness influence each other, and when and for whom the effects of
social presence and social awareness are most strongly related to positive outcomes in social
connectedness. Specifically, we study the concurrent direct and moderating effect of two structural
constructs characterizing peoples’ online social network: network size and frequency of usage. We test
our research model using data (n = 121) collected from two sources: (1) an online survey of Twitter users
and (2) their usage data collected directly from Twitter. Our results indicate that social awareness, social
presence, and usage frequency have a direct effect on social connectedness, while network size has a
moderating effect. We find social presence partially mediating the relationship between social awareness
and social connectedness. We use the findings of our analysis to outline design implications for online
social networks from a human computer interaction perspective.
Keywords: Online social network, social capital, social presence, social connectedness, social awareness,
Twitter.
1 Acknowledgements. We wish to thank Dominikus Baur and four anonymous reviewers for their
helpful comments. Christoph Riedl acknowledges support from the German Research Foundation (DFG)
under grant code RI 2185/1-1. All mistakes remain the authors’ own.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
2
Introduction
Benefits accrue to individuals and groups through social interactions (Granovetter, 1973; Bandiera
et al., 2008). Such benefits are termed social capital and can serve as a strong motivation for establishing
and maintaining social connections (Bourdieu, 1986; Adler and Kwon, 2002a). In addition to real life or
offline social interactions for maintaining social connections, various modern means of communication
such as online social network sites (SNS) have evolved into significant means for developing,
maintaining, and strengthening social connections (e.g., Brandtzæg et al., 2010; Vasalou et al., 2010;
Humphreys, 2007; Ellison et al., 2006; Aral and Walker, 2011). Therefore, these online social networking
sites can serve as important sources of social capital and, consequently, the various benefits that can be
realized through enhanced social capital (Ellison et al., 2007; Zhao and Rosson, 2009; Ji et al., 2010).
While prior research has started establishing causal relationships between the use of SNS and social
capital, there is still a limited understanding when and for whom these effects are strongest.
The phenomenal growth of SNS over the last decade has resulted in increased research and
practical interest in them. Accordingly, they serve as a fertile ground for information systems (IS), human
computer interaction (HCI), and social science research (Gu, 2010). There has been research focusing on
the various motivations that people have in joining online social networks (Ellison et al., 2007), on
sharing information in them (Wasko and Faraj, 2005; Brandtzæg et al., 2010), or on the relationship
between online and offline social networks and how these influence each other (Ellison et al., 2007;
Ledbetter et al., 2011; Oxendine et al., 2003). Much of this research is guided by the inherent realization
that SNS can serve to complement and modify peoples’ social interactions and behavioral patterns, and
can result in beneficial outcomes (Zhao and Rosson, 2009). However, there is little understanding of the
actual process through which the benefits associated with participating in such online social networks can
be realized. Assuming that feeling socially connected, or social connectedness, facilitates the accrual of
social capital related benefits, we would like to better understand the factors that can enhance social
connectedness, and the process through which social connectedness is created in online social networks.
Furthermore, in this study we address the question when and for which groups of SNS users the effects
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
3
are strongest. To address this question, we focus our attention specifically on potential moderating effects
of usage frequency and the size of the online social network.
Using the concept of social capital, which encompasses the norms and networks facilitating
collective actions for mutual benefits (Woolcock, 1998) as the overarching framework, we identify two
key factors that engender feelings of connectedness among the users of a SNS. We then empirically
examine our theoretical model using data collected from Twitter users using a combination of survey data
and usage data collected directly from Twitter. Survey research is a highly accepted research method in
many disciplines such as management, information systems, and various social science disciplines, as
well as HCI (e.g., Ren et al., 2011; .Ji et al., 2010; Morris et al., 2010). Twitter is currently one of three
dominant SNS platforms alongside Facebook and Google+ that allows its users to manage social
relationships among each other and provides functionalities to establish an online social network
(Thompson, 2007; Twitter, 2011). Further, Twitter can be considered as a technology-mediated
communication medium that substitutes face-to-face communication to some extent. There is an ongoing
debate on the effectiveness of such media as communication tools and different contexts in which they
can be used as suitable replacements. Our research contributes to this debate by investigating the extent to
which SNS usage results in feelings of connectedness among its users.
This research focuses on understanding the interplay between users’ perceptions regarding SNS
usage and their actual use of it in enhancing social connectedness (Carlson and Zmud, 1999). We address
this question in the context of Twitter by examining both structural and functional characteristics of an
individual’s online social network and the individual’s communication behavior within the network. In
particular, we investigate the interplay between three fundamental constructs of social interactions: social
presence, social awareness, and social connectedness. Furthermore, we study the concurrent direct and
moderating effect of two structural constructs characterizing peoples’ social network (Scott, 2000):
network size and frequency of usage. While prior research has examined and highlighted the importance
of social presence, social awareness, and social connectedness, there is little empirical evidence on how
they influence each other (Rettie, 2003). This lack of knowledge impedes academic and practical
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
4
understanding on SNS and their implications, and limits the ability to systematically design information
systems to address specific goals. Testing our theoretical model thus allows us to address the research
questions (a) how do social presence, social awareness, and social connectedness influence each other,
and (b) when and for whom the effects of social presence and social awareness are most strongly related
to positive outcomes in social connectedness. We also present additional descriptive analyses of users’
usage behavior. Based on the findings of our analyses, we propose design implications that incorporate
our study findings and analyses, and can be used by human computer interaction (HCI) designers for
informed design of online SNS.
The remaining sections of the paper are organized as follows: Section 2 provides an overview of
previous research on online social networking, focusing on social connectedness and social capital.
Section 3 outlines our research model and depicts the theoretical development of our research hypotheses.
Section 4 presents the data collection. Section 5 presents analyses and results of the theoretical model and
Section 6 provides additional descriptive analyses of Twitter usage behavior. Implications of our research
findings are discussed in Section 7.
Literature Review
Online social networking has gained significant popularity over the last decade. It is a form of
computer-mediated communication, facilitated by advances in information and communication
technologies, and the advent of Internet-based applications that allow its users to individually present
themselves through profiles, display their social networks, and establish or maintain connections with
others (Ellison et al., 2007). This novel form of communication and social interaction is increasingly
consuming a significant and growing portion of an individual’s time and attention (Bapna et al., 2011).
Social networking sites that enable individuals to follow the lives of friends, acquaintances and families,
to find experts, and to engage in commercial transactions have grown exponentially in their user numbers
since the turn of this century (Huberman et al., 2008). Examples of prominent SNS are Twitter, Facebook,
and most recently Google+, the three of them represent the most dominant SNS by 2011 in terms of
number of active users.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
5
Previous Research on Online Social Networks
Early research on computer-mediated communication focused on virtual communities has been
based on the assumption that people participating in them used these platforms to connect with
individuals outside their pre-existing social group or location (Wellman and Gulia, 1999), primarily based
on shared interests or shared life scenarios. For example, the role of virtual communities in providing
informational, emotional, and appraisal support to elderly and patient groups have been investigated
(Leimeister et al., 2008). Current SNS can be distinguished from these virtual communities by the fact
that they are primarily used for the maintenance of existing (offline) social ties, although these platforms
also allow for the formation of purely online ties. For instance, it was found that Facebook was used for
strengthening and intensifying offline relationships that people already have (Ellison et al., 2007; Lampe
et al., 2006).
Research has investigated different aspects of online social networking, such as privacy and
security concerns that users of such networks have and how they affect their usage behavior (Acquisti and
Gross, 2006), different motivations of using SNS, such as for finding and connecting to strangers, how
these sites can be used for keeping up with acquaintances (Köbler et al., 2011; Lampe et al., 2006), how
various functionalities provided by SNS impact social interactions (Boyd and Ellison, 2007), and the use
of social networking sites in a work context (Zhao and Rosson, 2009).
Studies have focused on the formation and dynamics of online social networks and how online
profiles can be created and managed to generate potential offline benefits for individuals, e.g., increasing
the feeling of social connectedness (Köbler et al., 2010), building social capital among its users (Lampe et
al., 2006; Koroleva et al., 2011; Wellman and Gulia, 1999), and creating social support (Leimeister et al.,
2008). Research indicates that social connectedness within SNS is a qualitative source of social capital,
since social connectedness activates and maintains social relationships and in turn results in beneficial
outcomes for individuals, such as a higher social capital (Koroleva et al., 2011). Further, frequent usage
and interaction with other SNS members can result in higher social trust and civic participation among
SNS users (Valenzuela et al., 2009). Accordingly individuals’ offline and online lives overlap and
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
6
influence each other through the use of SNS (Subrahmanyam et al., 2008; Ellison et al., 2007). SNS are
recognized as a complement network of relationships to the offline world by “providing a platform for
active communication between friends and more passive observation through aggregated streams of social
news” (Burke et al., 2010, p. 1909). Participation in SNS is associated with greater feelings of bonding
social capital, emotional support and lower levels of loneliness (Burke et al., 2010; Ellison et al., 2007).
However, none of the studies investigated in detail how this is achieved exactly.
In addition to the several beneficial aspects of SNS use, potential negative consequences from
participation in SNS have also been identified. It has been suggested that information overload caused by
active participation can result in reduced levels of user activity, a negative attitude towards shared
information that potentially lowers the possibility of beneficial aspects such as social capital (Koroleva et
al., 2010). In some cases, information overload might even cause confusions and dysfunctional effects in
the form of stress and anxiety among users of SNS (Eppler and Mengis, 2004). Other studies have
documented that active participation on SNS may result in depression or envy and jealousy of other
individuals (Muise et al., 2009).
The above discussion indicates the possible social benefits of SNS use, however, not every type of
SNS use (Burke et al., 2010) and not every social network (Granovetter, 1973) hold the same potential to
create social value for an individual. Drawing from the concept of social capital, we examine the factors
that enhance and/or explain social capital creation in SNS, which we then empirically investigate using
data collected from users of the Twitter as an example.
Social Capital and Social Connectedness
Social capital is an outcome of network membership and refers to the value of social relations
within and between social networks and the ability to produce collective or individual (economic) benefits
from network members. Networks are the causal agents or moderators of social capital (Williams, 2006).
A social network represents the social ties within which an individual lives and can be characterized by its
structural properties and functional properties (Scott, 2000). The structural properties describe the social
network at large, and are generally measured by the objective characteristics of social networks (Cohen
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
7
and Syme, 1985). For example, the number of ties that a person has in a social network, and frequency of
interactions with other members of the network are examples of structural properties of the social
network. The functional properties are subjective, and can be measured by capturing an individual’s
perceptions regarding the availability and adequacy of resources that can be obtained from the social
network (Cohen and Syme 1985). Therefore, functional properties can influence the general well-being of
network members by providing them with social goods, such as social support and connectedness, and
several forms of social capital (Huysman and Wulf, 2004). Previous research has identified several
beneficial outcomes of social support and social connectedness and has shown that these factors help
people in overcoming social disadvantages like isolation, and strengthen their integration into society
(Cornwell et al., 2008; Pfeil et al., 2011).
Multiple definitions of the concept of social capital can be found in sociology (Woolcock, 1998). It
has been defined as the source of positive externalities for members of a social network that are generated
through shared trust, norms, values, and their consequent effects on expectations and behavior and
broadly refers to the benefits received from social relationships (Adler and Kwon, 2002b). Similar to
physical, financial, and human capital, social capital may facilitate productive activity to individuals or
groups (Coleman, 1988; White, 2002). The beneficial effects of social capital could be of psychological,
emotional, and economical nature (Lin, 1999; 2000).
Social capital that emerges from the use of modern information technology has been referred to as
socio-technical capital (Resnick, 2001), and can be thought of as an artifact of users’ technology usage
behavior. Online social network use can thus enhance users’ social capital (Koroleva et al., 2011).
Therefore, we are interested in understanding the determinants of social capital among the users online
social network sites. In particular, our research identifies the role played by users’ communication
behavior and structural aspects of the network in determining social connectedness.
Social connectedness is a well-established concept in psychology that offers affective benefits like
intimacy, sense of sharing, and stronger group attraction (Ijsselsteijn et al., 2003). A number of studies on
social capital investigated connectedness within communities as an indicator of social capital and
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
8
identified “sense of community” as an indicator and catalyst for behavioral dimensions of social capital
(Stone, 2001; Long and Perkins, 2007). Social connectedness can be viewed as an estimator of the quality
of an individual’s social network and a precursor to social capital (Putnam, 2000), or directly linked to
social capital itself (Bandiera et al., 2008). Therefore, the extent to which network members can
appropriate social capital from the social network will be determined by their connectedness within the
network.
Human beings have the fundamental need to belong and feel connected (Rettie, 2003).
Accordingly, social connectedness is often viewed as a precursor to success in life, and refers to an
individual's attitude and relationship to society. Social connectedness has been identified as one of the
fundamental underlying motivating principles behind social behavior (Smith and Mackie, 2000).
Individuals can assess their social relationships based on the extent to which they feel socially connected.
Connectedness is characterized by a feeling of staying in touch with ongoing social relationships (Romero
et al., 2007), and has been associated with many social and health related benefits and is therefore a
desirable property for individuals to achieve.
SNS as Means of Enhancing Social Connectedness
Social connectedness can be described as the feeling of belonging to a social group and implies the
creation of bonding relationships. Communication is often seen as a means of feeling connected and being
in touch, and computer-mediated communication is increasingly becoming an important way of achieving
social connectedness. Accordingly, social connectedness is considered a potential key concept not only in
the analysis of communication but the development of communication technologies and applications,
such as SNS (Rettie, 2003). Communication can be classified as connectedness-oriented communication
where the focus is mainly on maintaining and enhancing social relationships, and content-oriented
communication, where the goal is exchange of information (Kuwabara et al., 2002). Connectedness-
oriented communication allows people to be aware of each other and contribute to maintain social
relationships (Rettie, 2003).
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
9
It is less clear, however, how social connectedness is achieved exactly. Two candidate constructs
have been suggested in various studies, social presence and social awareness, which we aim to
consolidate in our theoretical model. Social presence is considered an important factor in creating a sense
of community among online communities (Aragon, 2003), while the role of social awareness is
particularly critical in mediated environments given that there is scarcity of real life visual cues regarding
what others are doing in such environments (Bardram and Hansen, 2004). However, there is a lack of
clarity regarding the relationship between these theoretical constructs, the process through which feelings
of social connectedness are created, and further, there is no empirical validation of these relationships in
the context of online SNS. A theoretically grounded understanding of the relationship between these
variables will help in furthering the role of online SNS as significant sources of social capital and in
formulating guidelines for designing SNS and other communication environments to better support the
creation of social connectedness.
Research Model
Our research model outlines the relationship between the structural and functional characteristics of
a SNS using social capital and social connectedness as the theoretical lens. Based on the preceding
review, we propose that social awareness and social presence, as well as the size of an online social
network and usage of SNS influence social connectedness (see Figure 1).
----- Figure 1 -----
Social Awareness and Social Presence
Awareness is defined as an understanding of the activities of others which defines one’s own
activities (Dourish and Bellotti, 1992). Social awareness relies on knowing the (social) context of an
individual; as a result, individual A perceives social awareness of individual B, if A has access to the
(social) context of individual B (Bardram and Hansen, 2004).
Social awareness on a SNS refers to users being conscious about the activities of other users of the
SNS. This consciousness or awareness can be facilitated by certain functionalities provided by the SNS,
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
10
or also by the way people choose to use the platform or communicate over it. When people are not co-
located, they can create a sense of awareness by mediating information regarding their (social) contexts
through the use of technology, such as email, instant messaging, or SNS. It has been shown, for example,
that online and status information in instant messaging applications raised awareness regarding
availability of individuals in the network (Nardi et al., 2000; Bardram and Hansen, 2004).
In contrast to awareness, social presence is defined as “degree of salience of the other person in a
mediated communication and the consequent salience of their interpersonal interactions” (Short et al.,
1976, p. 65). Social presence is usually considered as a prerequisite for effective communication. Social
presence within a computer-mediated environment is defined as user’s subjective sensation of “being
there,” which is also referred to as “telepresence” or “being together” (Biocca and Levy, 1995; Biocca,
1997; de Greef and Ijsselsteijn, 2001). Others have linked the notion of social presence to concepts of
“immediacy” (Weiner and Mehrabian, 1968) and “intimacy” (Argyle and Dean, 1965), where actions
within a medium that facilitate immediacy are used to create and maintain intimacy, and these actions
also enhance the feeling of social presence within a medium (Gunawardena, 1995).
Literature reflects the interconnection between social presence and awareness by introducing
definitions and terms such as “presence awareness” (Tollmar et al., 1996), which can be regarded as the
awareness of the presence of social ties, and “affective awareness” which can be defined as a general
sense of being connected (Liechti and Ichikawa, 1999). While social presence implies awareness of the
other individual, it has been proposed that “awareness can occur without either social presence or
connectedness” (Rettie, 2003, p. 4). Social presence has also been described as the state in which two
individuals are aware of each other in a virtual environment, and consequently, their “mutual awareness is
the essence of social presence” (Biocca et al., 2003, p. 463).
In computer-mediated environments, it is the medium that allows a sense of social presence (Short
at al., 1976). In particular, media might differ in the extent to which they allow users to experience others
as being present (Cyr et al., 2009). This might play a particularly important role in Twitter as message
length is severely restricted (140 characters). Therefore, social presence can be facilitated or limited by
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
11
certain functionalities of Twitter. Further, people’s perception regarding awareness will also increase the
perception of social presence among Twitter users. Therefore, we hypothesize:
H1: Perception of social presence is positively associated with social awareness.
Social Awareness and Social Connectedness
In the context of computer-mediated communication, awareness can be thought of both as a
perception of other users in the system, and as an aspect of the system that facilitates that perception
(Rettie, 2003). Accordingly, social awareness in SNS refers to people being conscious about other users
in the network. It has been proposed that raised awareness regarding availability of other individuals in
the network, and features that help in monitoring this information, creates and maintains a sense of social
connectedness (Nardi et al., 2000). Therefore, a SNS as a medium will facilitate social connectedness by
allowing for enhanced awareness among its users. Accordingly, we hypothesize:
H2: Social awareness will lead to social connectedness.
Social Presence and Social Connectedness
Social presence has been referred to as a psychological connection that users feel among each other
after an interaction, and can be described as a temporary judgment of an interaction as augmented or
limited by the medium (Biocca et al., 2003). Therefore, increased perceptions regarding social presence
among SNS users will enhance the connection that users will feel with each other. This will result in
increased feelings of social connectedness. Moreover, social connectedness is also described as the
emotional experience, which is generated by the perceived social presence of other individuals in a
medium (Rettie, 2003). Therefore, we hypothesize:
H3: Social presence will lead to social connectedness.
Effects of Network Size and Frequency of Use
Social connectedness has been characterized by the size of the social network (2008; Cornwell et
al., 2008; Goswami et al., 2010). Accordingly, network size will play a role in the extent to which users
feel socially connected. A user who has many connections will be more ensconced within the network,
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
12
and therefore, feel more socially connected. Further, having more members in one’s social network means
that at any point in time the user will have a heightened awareness regarding other members of the
network, and this in turn will result in increased feelings of being connected. Therefore, network size will
also moderate the relationship between social awareness and social connectedness. Accordingly, we
hypothesize:
H4a: Network size is positively associated with social connectedness.
H4b: Network size positively moderates the relationship between social awareness and social
connectedness.
Frequency and amount of communication play an important role in maintaining social relationships
(Rettie, 2003). Previous research has suggested that social connectedness is enhanced by active
information sharing among members (Köbler et al., 2010). Therefore, more frequent usage would result in
users experiencing higher levels of social connectedness. Further, the more people interact with each
other, the more it is likely that they will experience psychological and emotional connection with other
members of the network. Since social presence also refers to the psychological connection that users feel
after an interaction, a higher frequency of use will result in a stronger relationship between social
presence and social connectedness. Accordingly we hypothesize:
H5a: Frequency of use is positively associated with social connectedness.
H5b: Frequency of use positively moderates the relationship between social presence and social
connectedness.
Method and Data
The research model was tested using survey data as well as actual usage data collected from
Twitter.
Background on Twitter
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
13
Twitter, founded in 2006, describes itself as a real-time information network (Twitter, 2012).
Twitter had around 175 million registered users at the end of 2010 with 500.000 new accounts created
every day in 2011 (Twitter, 2011). Twitter allows its users to manage relationships among each other and
thus establish an online social network (Thompson, 2007; Twitter, 2012).
At the heart of the Twitter network are short messages termed tweets of a maximum of 140
characters in length. In prior research on Twitter, the messages exchanged on the SNS have been
classified as mostly consisting of status updates of what users are doing, where they are, how they are
feeling, or hyperlinks to other sites (Kietzmann et al., 2011). Therefore, Twitter is more suited for
connectedness-oriented communication rather than content-oriented communication. Guided by the
theories on social capital, social presence, and social awareness, we look for evidence of mechanisms
within Twitter that can strengthen social connectedness.
Tweets can contain any type of textual information, including hyperlinks (e.g., Uniform Resource
Locator information) to other resources. By default, all messages posted are publicly available in an
unrestricted fashion in the global stream of Twitter messages. To filter this global stream, users can
choose to follow other Twitter users, i.e., users subscribe to the message stream of other users to get these
messages displayed on their Twitter clients (e.g., browser-based and mobile clients). These users are
termed followers. Followers and following are means of forming a network resulting in a directed graph
of social ties within Twitter. In this social graph, Twitter users represent nodes and following-follower
relationships represent directed edges between the nodes with the follower relationship resulting in an
incoming connection (i.e., in-degree) and the following relationship in an outgoing connection (i.e., out-
degree).
In addition to a standard message type, two special types of messaging are possible: re-tweets and
replies. Re-tweets are citations of existing tweets that are sent without editing to a user’s followers. This
type of message is a means of spreading information from others through a user’s own network of
followers. Replies are messages directed at a specific user while still being publicly available. In addition
to the browser-based client, Twitter also maintains an application programming interface (API) which
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
14
allows third-party developers to offer software clients to access Twitter and to integrate the service into
other application offerings, e.g., mobile clients.
Online social networks, and particularly Twitter, exemplify the mode of communication that
primarily caters to connectedness-oriented communication, since a very small message size typically pre-
empts the conveyance of content. In order to support this connectedness-oriented communication a high
degree of social awareness and social presence is necessary. Consequently, SNS have also been classified
as awareness or presence systems (Rettie, 2003). In this research, we examine the relationship between
social awareness, social presence, and social connectedness in the context of Twitter.
Data Collection
An online survey was developed to measure social awareness, social presence, and social
connectedness. The link to the survey was distributed through student mailing lists of under graduate and
graduate students enrolled in a large German university, and by posting the survey link in various SNS
(i.e., Facebook), and sending out tweets on various lists of Twitter users. Similar sampling approach is
often used for online surveys (e.g., Ledbetter et al., 2011). All respondents to our survey were Twitter
users. The survey was carried out over a two months period in mid 2011. The survey research
methodology is particularly suitable for testing relationships between multiple constructs simultaneously
and enhances the generalizability of results (Dooley, 2001). In the survey, respondents were asked to
provide their Twitter user names. This information was then used to gather network size and usage
information directly from respondents’ Twitter accounts by employing the Twitter API. The survey
consent form clearly explained this data collection procedure to respondents. After combining the survey
data with the information collected directly from respondents’ Twitter profiles, we deleted the Twitter
user names from our records in order to anonymize the dataset. The details of the data collected directly
from Twitter are provided below in the discussion on operationalization of variables.
Operationalization of Variables
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
15
Most of our research variables were operationalized using multi-item survey scales, while others
(such as tenure, usage, and network size) were usage data obtained from Twitter, retrieved by Twitter API
calls. In order to enhance validity of the survey instrument, we used tested and validated scales from
previous studies to measure our research variables (Stone, 1978).
The initial face and content validity of the measurement items was assessed by researchers and
faculty members of the computer science department of a large German university. Their feedback was
used to refine the items. Following this, two rounds of questionnaire sorting exercise (labeled and
unlabeled) were carried out based on Moore and Benbasat (1991). Four graduate students participated in
each sorting exercise. For the unlabeled sorting exercise, the labels or descriptions that the sorters came
up with closely corresponded with the actual construct names and on average more than 70% of the items
were correctly sorted into their intended constructs. The results of the unlabeled sorting exercise were
then used to drop some of the measurement items (which were wrongly sorted or found to be ambiguous)
and to refine the scale. The labeled sorting exercise – in which the sorters were provided with the name
and definition of each construct – resulted in an even higher percentage of the items getting correctly
sorted into their intended constructs, thus indicating a high level of face and content validity. All
constructs were measured on a seven-point Likert scale ranging from “Strongly Disagree” to “Strongly
Agree.” The following paragraphs provide a brief definition of the constructs and Table 1 reports the
items that were used to measure the various constructs, with the various sources from which they were
derived.
Social Connectedness. Connectedness refers to being connected by having social, professional, or
commercial relationships. People having high levels of social connectedness tend to feel close to others,
identify with others and participate in social groups and activities (Lee et al., 2001). Social connectedness
in the current context is defined as the connectedness that is achieved by using Twitter as a
communication medium. We drew from previous research measuring online social connectedness
(Ledbetter, 2009; Ledbetter et al., 2011) in order to measure social connectedness using nine items.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
16
Social Presence. Social presence is usually defined as users’ subjective sensation of “being there”
in a scene, i.e., the illusion of having a non-mediated communication when in reality the communication
is mediated. Based on previous research on social presence (e.g., Biocca et al., 2003; Shen and Khalifa,
2009) we conceptualize and operationalize social presence as a user experience, i.e., a phenomenal nature
of the medium as perceived by the user. The items used to measure social presence were derived from
Shen and Khalifa (2009).
Social Awareness. We define social awareness as the extent to which users are aware of the
presence of others, and react to their presence. Therefore, social awareness can be thought of as
perception of other users and is facilitated by certain aspects of the systems (Rettie, 2003). In computer-
mediated communication such as in online communities, microblogging platforms, and SNS, social
awareness is dependent on factors such as the use of status messages and symbols.
Table 1 shows the items that were used to measure social connectedness, social presence, and
social awareness.
----- Table 1 -----
Network Size (measured directly from Twitter). Following established practices in graph theory,
we measured network size through a node’s in-degree and out-degree. In the context of Twitter, our
measure was operationalized as a sum of the number of followers and following relationships. The
number of followers and the number of following was obtained directly from Twitter through calls to the
Twitter API. Since this variable ranged from 0 to 152,813, and significantly deviated from a normal
distribution, we performed a natural log transformation to obtain a relatively normal distribution of
network size.
• Following: The number of other Twitter users that the focal user is following (i.e., has subscribed to
their tweets). The variable Following represents the number of other Twitter users in whom the focal
user is interested, and represents the node’s out-degree. The variable was measured as a count
obtained directly from respondents’ Twitter profiles and ranged from 0 to 73,959 in our sample.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
17
• Followers: The number of users who follow a focal user. This is a direct measure of the focal user’s
online social network size regarding incoming connections and represents the node’s in-degree. The
number of followers cannot be controlled by the focal user but are unsolicited expressions of interest
of those who follow. In our sample, the variable followers ranged from 0 to 78,854.
Frequency of Use (measured directly from Twitter). Different measures have been employed in
the field of IS research to measure the usage of an IS application. In the context of the SNS Twitter, we
operationalize our usage measure as the frequency with which the application Twitter has been used.
Frequency of usage is calculated by dividing the total number of tweets a user has posted by the user’s
tenure with Twitter. Both measures have been obtained directly from Twitter through calls to the Twitter
API. Like network size, frequency is transformed using a natural log transformation.
• Tenure (Twitter API): The Twitter API allows reading the sign-up date when a Twitter user account
has been created. We have operationalized this as “(June 26, 2011 – sign-up Date) in days.” In our
sample, the earliest sign-up date was Nov 03 15:01:14 EST 2006; and the most recent sign-up date
was Apr 12 01:23:13 EDT 2011.
• Tweets (Twitter API): The number of all tweets ever posted by the respondent as a natural number
(min 0; max 62,173).
As the number of consecutive calls to the Twitter API is limited to 500 per hour, we had to perform the
data extraction over multiple days.
Data Analysis and Results
We received 141 completed survey responses out of which 121 provided a valid Twitter name
linking to a public profile. These 121 were included in the final analysis. Table 2 presents the
demographic variables of our sample. As mentioned previously, all our respondents were Twitter users,
and the sample was equally distributed between male and females Twitter users. The demographic
distribution of our sample corresponds closely to the overall demographic distribution found on Twitter
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
18
(Twitter Demographics, 2012). While we cannot rule out potential selection bias, the demographic
characteristics of our sample with regards to gender, age, and education are representative of the overall
population of Twitter users.
----- Table 2 -----
Partial least squares (PLS) analysis was used to perform the data analysis using the software
SmartPLS Version 2.0. (Ringle et al., 2005). PLS has enjoyed increasing popularity in recent years
because its ability to model latent constructs under conditions of non-normality and in small to medium-
sized samples (Chin, 1998). It allows researchers to specify the relationships among the conceptual
factors of interest and the measures underlying each construct, resulting in a simultaneous analysis of the
measurement model (i.e., how well the measures relate to each construct) and the structural model (i.e.,
whether the hypothesized relationships at the theoretical level are empirically true). Using the two-stage
approach, we first examined the measurement model, and then the structural model.
Measurement Model
Internal consistency, convergent validity, and discriminant validity were assessed to validate the
instruments (Gefen and Straub, 2005). Internal consistency was examined using composite reliability,
which in PLS relies on the actual loading to calculate the factor scores and is a better indicator of internal
consistency than Cronbach’s alpha (Ranganathan et al., 2004). As shown in Table 3, the composite
reliability for the constructs in the model were all above the suggested threshold of 0.7 (Chin, 1998;
Straub, 1989), thus supporting the reliability of the measures.
Convergent validity indicates the extent to which the items of a scale that are theoretically related
are also related in reality. Convergent validity measures the correlation among item measures of a given
construct using different methods of measurement. Table 3 presents information about the factor loadings
of the measures of our research model. All items have significant path loadings at the 0.001 level. The
average variance extracted (AVE) values are all above the recommended value of 0.5 (Fornell and
Larcker, 1981). Therefore, convergent validity of the scales is acceptable.
----- Table 3 -----
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
19
Table 4 shows the correlation between constructs, with the diagonal elements being the square
roots of the average variance extracted (AVEs). This indicates that the variance shared by the items of a
particular construct is higher than the variance that is shared by the construct with another construct,
which is used as an indicator of discriminant validity in PLS.
----- Table 4 -----
Another method of assessing discriminant validity is by checking the factor loadings and cross
loadings of the measurement items (see Table 5). Scanning down the columns indicates that the item
loadings in their corresponding columns are all higher than the loadings of items used to measure the
other constructs. Scanning across rows indicate that item loadings are higher for their corresponding
constructs than for other constructs. Thus, the measurement items of this model satisfy the two criteria for
discriminant validity suggested by Chin (1998). Overall, the data provides empirical support for
reliability, as well as convergent and discriminant validity of the scales of our measurement model.
----- Table 5 -----
Structural Model
Table 6 presents the result of our model testing. We first tested a model incorporating only the
main effects and then introduced the moderating effects in a second model. Overall, our models explain
59% and 63% of the variance in social connectedness in Twitter. In addition to testing the main effects of
social awareness and social presence on social connectedness, we also test for a mediating effect of social
presence on the relationship between social awareness and social connectedness. Our results indicate that
both social awareness and social presence have significant direct effects on social connectedness and that
the effect of social awareness is partially mediated by social presence.
----- Table 6 -----
While network size does not have a significant direct effect on social connectedness, it positively
moderates the relationship between social awareness and social connectedness. Frequency of use has a
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
20
significant direct effect. However, the moderating effect on the relationship between social presence and
social connectedness was not significant.
Understanding Usage Behavior
The above analysis relies on partial least squares analysis to test a theoretic model linking Twitter
usage to perceived benefits such as social connectedness. To better understand our sample population’s
usage behavior of Twitter, we performed several additional analyses both on actual usage data collected
from Twitter as well as responses elicited through the online survey. First, we investigate what other
communication media are used and in which frequency. Figure 2 shows communication media usage.
Face-to-face, social media, and cell phone text messages are the dominant communication media with
around 50% of the respondent reporting “very frequently.” Given the sample population of Twitter users,
the usage frequency of other electronic communication media such as email, cellular phone text messages
is, not surprisingly, very high (around 45% responding “very frequently”).
----- Figure 2 -----
To better understand whether the respondents to our survey were exclusive users of Twitter, we
collected data on their usage of other social networking sites (Figure 3). Almost all respondents were also
users of Facebook (93%). The only other widely-used SNS was LinkedIn (54%). Other SNS had only
limited reported usage of below 25%.
----- Figure 3 -----
In order to get a better understanding of users’ feeling of connectedness, we further analyzed self-
reported details regarding the nature of connections users have on Twitter. We investigate different
connection types in the panels of Figure 4. Panel 1 shows the composition of connections. Most
connections are with celebrities and prolific people (22%), friends (21%), and acquaintances and casual
contacts (21%). Very few connections are with family members or romantic relationships (2% and 0.3%
respectively). Work-based relationships account for 13%. Panel 2 shows the classification of connections
regarding their physical distance. Twitter connections are mostly between local friends (34% reported
“almost all” or “all”). The classification of long-distance friends is largely consistent (42% reporting
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
21
“none” or “very few” long-distance connections) which increases the reliability of this finding. This
indicates that most Twitter users rely on Twitter as a medium for remaining connected to friends, casual
acquaintances, and other prominent Twitter users rather than using it for keeping in touch with family.
Therefore, Twitter can be considered as a tool for establishing and maintaining weak network ties, rather
than for strong network ties.
----- Figure 4 -----
Finally, we investigate individuals’ usage behavior by analyzing each respondent’s last 200 tweets
(Figure 5 shows a random sample of 50 respondents from our overall sample of 121). To illustrate
frequency of use, we plot the tweets as a time series: the shorter the time series, the higher an individual’s
usage frequency (the elapsed time between the last 200 tweets is shorter). For a high-frequency user such
as “User 1,” the last 200 tweets span only a period of 3.6 days resulting in an average of 55 tweets per
day. For a low-frequency user, such as “User 50,” the last 200 tweets span almost one and a half years
resulting in an average of 0.3 tweets per day. The majority of tweets are connectedness-oriented mentions
of other users accounting for 45% of all tweets. Another category of connectedness-oriented tweets is re-
tweets with additional mention of another Twitter user (mostly the source of the re-tweet), which account
for another 20% of tweets. Re-tweets without mentioning of another user are very rare accounting for
below 1% of tweets. Finally, plain tweets account for 35% of the tweets. This supports our argument that
Twitter is a connectedness-oriented social network and provides additional credibility to our findings of
the ability to build social relationships through the network.
----- Figure 5 -----
Discussion and Implications
Discussion
In this paper, we developed a model theorizing relationships between social presence, social
awareness, and social connectedness. Additionally, we investigated direct and moderating influences of
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
22
users’ network size and frequency of use. We tested the model using survey data collected from users of
the online social network Twitter. The effects of network size and frequency of use were measured
directly from Twitter. While prior research has examined and highlighted the importance of social
presence, social awareness, and social connectedness, there is little empirical evidence on how they
influence each other (Rettie, 2003). Testing our theoretical model, we address the research questions (a)
how do social presence, social awareness, and social connectedness influence each other, and (b) when
and for whom the effects of social presence and social awareness are most strongly related to positive
outcomes in social connectedness.
Our data show a positive and significant effect of social awareness on social presence, thus
supporting H1. Furthermore, our empirical results find both social awareness and social presence have a
significant positive effect on social connectedness, thus supporting H2 and H3. Furthermore, we find that
social presence partially mediates the relationship between social awareness and social connectedness.
Thus, social awareness functions as a fundamental building block for the higher-level constructs of social
presence and social connectedness, while social presence functions as a mechanism through which the
positive effect of social awareness on social connectedness is realized. This finding thus explains the
process through which social connectedness is created and has important implications on the designs of
SNS. We argue that social connectedness is the higher-level construct, which is influenced by both the
functional and structural characteristics of the social network, and the communication among individuals
within the network.
In the context of computer-mediated communication, social presence theory advocates that the
medium should be as proximate to face-to-face communication as possible. However, previous research
in social presence in computer-mediated communication has often been criticized for taking a too media-
centric view on the occurrence of social presence (Shen and Khalifa, 2009), with social presence often
being considered an inherent property of the medium itself (the richer the medium, the more social
presence it is believed to have). Our research overcomes this criticism by considering the mechanism
through which social presence can be enhanced in a not so rich medium (Twitter). By facilitating the
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
23
exchange of very short connectedness-oriented messages and status updates, Twitter enhances feelings of
awareness regarding other users in the medium, and this in turn enhances social presence. Further, both
social presence and social awareness significantly influence social connectedness, an excellent indicator
of the potential to appropriate social capital from the network.
To provide a more complete picture addressing the question when and for which user groups
feelings of social connectedness are most strongly affected by social presence and social awareness, we
also study direct and moderating effects of network size and usage frequency through data collected
directly from user profiles. Regarding network size we find no significant direct effect but a significant
moderating effect on the relationship between social awareness and social connectedness. Therefore, H4a
is not supported, while H4b is supported. This indicates, that having a large network is not necessarily a
pre-condition to feeling socially connected, however, the larger the network size, the stronger the
relationship between social awareness and social connectedness. The relationship between social
awareness and social connectedness can be summarized as a feeling of being connected in the network
because of a better understanding regarding the activities of others. A larger network results in increased
understanding regarding the activities of others and therefore increased feeling of connectedness.
Frequency of use has a significant direct effect but the moderating effect on the relationship
between social presence and social connectedness is not significant, thus hypothesis H5a is supported
while H5b is not supported. We speculate that this could be due to the fact that not every type of usage
and not every type of social network results in having higher levels of psychological connection as a
result of interactions within the network (Granovetter, 1973; Burke et al., 2010). In Twitter, for instance,
not just the amount or frequency of usage, but rather the fundamental motivation behind usage of the
medium, may have a better explanatory power of the relationship between social presence and social
connectedness.
Further analysis of Twitter usage behavior provides a better indication of the structural
characteristics of the Twitter network, and shows that users rely on connectedness-enhancing Tweets in
order to establish and maintain weak ties with other Twitter users. Previous research has suggested that
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
24
social capital is associated with both strong and weak ties that an individual has within a system
(Granovetter, 1973; Nahapiet and Ghoshal, 1998). This indicates that Twitter can indeed serve as means
of enhancing social capital for its users.
Overall, our use of two distinct but complementary data sources provides a richer picture of the
studied phenomenon and follows suggested practice in IS research (Sharma et al., 2009). This study is
one of the few examples of research that combines measures of psychological perceptions with actual
measures of usage and network size from the SNS to investigate the link between subjective and objective
variables. Our results show that usage of an online social network, such as Twitter, can indeed result in
tangible outcomes by increasing users’ social capital. Hence, we provide empirical evidence to the
discussion on whether the use of online social networks translates into real-world benefits.
Limitations
In this study of social capital formation in SNS, we relied on two very basic measures of usage
behavior and network size. We did not, in particular, investigate additional structural properties of the
online social network such as the networks density or the centrality of the focal user within the
relationship graph. As relationships on Twitter are directed rather than undirected (as it would be the case,
for example, in Facebook) this might provide additional insights which could be explored in future
research. Furthermore, our usage measure did not take the nature or the content of the communicated
information into consideration. Consequently, as the motivation to join Twitter and the information
shared on Twitter is likely not identical to that of other SNS such as Facebook or Google+, future
research could be extended in this direction to investigate more closely in how far our results can are
generalizable to other SNS. Also, we acknowledge that users may be simultaneously active on multiple
SNS, and their overall social connectedness is a cumulative effect. However, in this research, we were
primarily interested in focusing on social connectedness and its determinants that can be attributed to one
particular SNS, and therefore we only asked respondents about their Twitter usage behavior. We believe
that this parallels real life social network membership and its consequences as people simultaneously
belong to multiple networks (e.g., school friends, work colleagues, local community) and they are likely
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
25
to feel different levels of connectedness in each of these networks. Researchers have already started
analyzing the content of communication in online social networks, and these could provide a richer
understanding of different types of usage behaviors. However, only few of them have started to
investigate the link between content and perceptual responses.
Most studies of Twitter are population-level studies (e.g., Golder and Macy, 2011; Perreault and
Ruths 2011). Our paper provides a valuable contribution as it attempts to establish user-level inferences
between functional and structural features of the social networking site on users’ feeling of
connectedness. This is a difficult task due to sample selection and can potentially introduce biases, which
can limit the degree to which inferences can be drawn. However, our analysis of demographic
information from Twitter shows that the distribution of the various demographic characteristics in Twitter
(Twitter Demographics 2012) is very similar to their distribution in our research sample. We therefore
believe that this user-level study can lead to more meaningful insights and consequently design
implications.
We study an actual, global online social network with several million users and their established
social networks, rather than an experimental system in a laboratory setting. While we acknowledge that
this type of research using observational data, and self-reported responses can have certain limitations, the
fact that we study an actual system, with actual users, and their natural usage behavior should reflect
positively on the concerns of external validity.
Given the observational nature of our study, we acknowledge potential limitations to causal
inference. However, we objectively measure actual Twitter usage and social network size, and are thus
improving on earlier research on user behavior in online social networks that often relies only on survey-
based measures. For example, in their study of communication behavior Ledbetter et al. (2011) rely on
users’ self-reported usage metrics instead of measurements taken directly from the system. Consequently,
we expect very little measurement error in determining users’ network size and usage frequency.
Furthermore, using multiple data sources and measurement methods, we address potential concerns of
common method variance.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
26
Research Implications
Our work contributes to the stream of literature on the formation of social capital in online social
networks (Ledbetter et al., 2011; Koroleva et al., 2011) by specifically investigating moderating effects of
usage frequency and online social network size on the relationship between social awareness, social
presence, and social connectedness.
Extending early research on social capital formation process, we go beyond the study of direct
effects but specifically study both mediating and moderating effects in order to deepen our understanding
how key constructs social presence, social awareness, and social connectedness influence each other and
the conditions that affect the direction and/or strength of the underlying relationship. As current research
on the formation of social capital via SNS usage lacks validated measurement instruments that are
specifically developed to capture social capital outcomes (Koroleva et al., 2011), we built our research
model on existing and established psychological concepts such as social presence, social awareness and
social connectedness as potential predictors of social capital among SNS users. Our literature review
indicated that prior empirical research missed to generate a clear understanding of potential relationships
between these constructs. Therefore, our research contributes to a better understanding of the
relationships between the three constructs by disentangling the empirical effect of social presence and
social awareness on social connectedness. We thus contribute to and extend social psychology literature,
which has investigated these constructs (Rettie, 2003; Biocca et al., 2003; Shen and Khalifa, 2009;
Koroleva et al., 2011).
Design Implications
From a practical perspective, our finding could have major implications in the field of HCI where
researchers consistently attempt to develop systems to improve various aspects of social benefits of
applications. Building on the concept of evidence-driven social design, we aim at translating the findings
resulting from the test of our theoretical model into concrete design implications (Kraut and Resnick,
2011a; Riedl, 2011). The goal of this being to help designers incorporate these social science research
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
27
findings into informed design choices for online communities and online social networks. Table 7
provides a summary of design implications that can be drawn from this research.
----- Table 7 -----
Conclusion
In this study we presented results that extend our understanding of the theoretical and empirical
relationship between social presence, social awareness, and social connectedness in the use of online
social network sites. Drawing from literature on social psychology, we developed a model that relates the
key constructs influencing social connectedness in the offline world to the realm of computer-mediated
communication in online social networks. Furthermore, we contribute to our understanding of which user
groups are most strongly affected by the interaction of social presence, social awareness, and social
connectedness, and the underlying process through which social connectedness gets created. The
empirical results using data collected from Twitter users largely support the proposed research model. We
also investigated the direct and moderating effects of network size and usage frequency, finding a direct
effect of usage frequency on social connectedness and a moderating effect of network size on the
relationship between social awareness and social connectedness. We argued that by increasing users’
feeling of social connectedness online social networks such as Twitter can make valuable contribution to
users’ well being. Thus, we provide insights about how specific aspects of computer-mediated
communication in the form of online social networks affect and moderate users’ perception of social
connectedness and provide design implications that can inform the design of online social networking
sites.
References
Acquisti, A. & R. Gross (2006). Imagined communities: Awareness, information sharing and privacy on
the Facebook. Proceedings of 6th Workshop on Privacy Enhancing Technologies, Cambridge, UK.
Adler, P. S. & S. W. Kwon (2002a). Social Capital: Prospects for a New Concept. Academy of
Management Review (27) 1, pp. 17-40.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
28
Adler, P. S. & S. W. Kwon (2002b). Social Capital: The Good, the Bad, and the Ugly. In E. L. Lesser
(Ed.) Knowledge and Social Capital. Woburn, MA: Butterworth-Heinemann, pp. 89-115.
Aragon, S. R. (2003). Creating social presence in online environments. New Directions for Adult and
Continuing Education (100), pp. 57-68.
Aral, S. & D. Walker (2011). Creating social contagion through viral product design: A randomized trial
of peer influence in networks. Management Science (57) 9, pp. 1623-1639.
Argyle, M. & J. Dean (1965). Eye-contact, distance and affiliation. Sociometry (28) 3, pp. 289-304.
Bandiera, O., I. Barankay, & I. Rasul (2008). Social capital in the workplace: Evidence on its formation
and consequences. Labour Economics (15) 4, pp. 725-749.
Bapna, R., A. Gupta, S. Rice, & A. Sundararajan (2011). Trust, Reciprocity and the Strength of Social
Ties: An Online Social Network based Field Experiment. Proceedings of Conference on Information
Systems and Technology, Charlotte, NC.
Bardram, J. E. & T. R. Hansen (2004). The AWARE Architecture: Supporting Context-Mediated Social
Awareness in Mobile Cooperation. Proceedings of the 2004 ACM Conference on Computer
Supported Cooperative Work, pp. 192-201.
Biocca, F., C. Harms, & J. Burgoon (2001). Criteria and Scope Conditions for a Theory and Measure of
Social Presence. Proceedings of 4th Annual International Workshop on Presence, Philadelphia, PA.
Biocca, F., C. Harms, & J. K. Burgoon (2003). Toward a more robust theory and measure of social
presence: Review and suggested criteria. Presence: Teleoperators & Virtual Environments (12) 5, pp.
456-480.
Biocca, F. & M. R. Levy (1995). Communication in the Age of Virtual Reality. Hillsdale, NJ: Erlbaum
Associates.
Bourdieu, P. (1986). The forms of capital. In J. Richardson (Ed.) Handbook of Theory and Research for
the Sociology of Education. New York, NY: Greenwood, pp. 241-258.
Boyd, D. & N. Ellison (2007). Social network sites: Definition, history, and scholarship. Journal of
Computer Mediated Communication (13) 1, article 11.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
29
Brandtzæg, P. B., M. Luders, & J. H. Skjetne (2010). Too many facebook 'Friends'? Content sharing and
sociability versus the need for privacy in social network sites. International Journal of Human-
Computer Interaction, (26) 11/12, pp. 1006-1030.
Burke, M., C. Marlow, & T. Lento (2010). Social Network Activity and Social Well-being. Proceedings of
the 28th International Conference on Human Factors in Computing Systems, Atlanta, GA, pp. 1909-
1912.
Carlson, J. R. & R. W. Zmud (1999). Channel expansion theory and the experiential nature of media
richness perceptions. Academy of Management Journal (42) 2, pp. 153-170.
Cohen, S. & S. L. Syme (1985). Issues in the Study and Application of Social Support. In S. Cohen and S.
L. Syme (Eds.) Social Support and Health. San Francisco: Academic Press.
Coleman, J. (1988). Social capital in the creation of human capital. American Journal of Sociology (94)
Supplement, pp. S95-S120.
Cornwell, B., E. O. Laumann, & L. P. Schumm (2008). The social connectedness of older adults: A
national profile. American Sociological Review (73) 2, pp. 185-203.
Chin, W. W. (1998). The partial least squares approach for structural equation modeling. In G. A.
Marcoulides (Ed.) Modern Methods for Business Research. Hillsdale, NJ: Lawrence Erlbaum, pp.
295-336.
Cyr, D., M. Head, H. Larios, & B. Pan (2009). Exploring human images in website designs: A multi-
method approach. MIS Quarterly (33) 3, pp. 530-566.
Dabbish, L., & R. E. Kraut (2008). Awareness displays and social motivation for coordinating
communication. Information Systems Research 19(2), pp. 221-238.
Dooley, D. (2001). Social Research Methods. Upper Saddle River, NJ: Prentice-Hall.
Dourish, P. & V. Bellotti (1992). Awareness and Coordination in Shared Workspaces. Proceeding of
ACM Conference on Computer Supported Cooperative Work, Toronto, Canada, pp. 107-114.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
30
Ellison, N., C. Steinfield, & C. Lampe (2007). The benefits of Facebook ‘Friends:’ Social capital and
college students’ use of online social network sites. Journal of Computer-Mediated Communication
(12) 4, article 1.
Ellison, N., R. Heino, & J. Gibbs (2006). Managing impressions online: Self-presentation processes in the
online dating environment. Journal of Computer-Mediated Communication (11) 2, article 2.
Eppler, M. J. & J. Mengis (2004). The concept of information overload: A review of literature from
organization science, marketing, accounting, MIS and related disciplines. The Information Society:
An International Journal (20) 5, 1-20.
Fornell, C. & D. Larcker (1981). Evaluating structural equation models with unobservable variables and
measurement error. Journal of Marketing Research (18) 1, pp. 39-50.
Gefen, D. & D. W. Straub (2005). A practical guide to factorial validity using PLS-Graph: Tutorial and
annotated example. Communications of the AIS (16) 1, pp. 91-109.
Golder, S. A. & M. W. Macy (2011). Diurnal and seasonal mood vary with work, sleep, and daylength
across diverse cultures. Science (333) 6051, pp. 1878-1881.
Goswami, S., F. Köbler, J. M. Leimeister, & H. Krcmar (2010). Using Online Social Networking to
Enhance Social Support and Social Connectedness for the Elderly. Proceedings of International
Conference on Information Systems, St. Louis, USA.
Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology (78) 6, pp.1360-
1380.
de Greef, P. & W. A. Ijsselsteijn (2001). Social presence in a home tele-application. CyberPsychology &
Behavior (4) 2, pp. 307–315.
Gu, Y. J. (2010). HCI and social computing. International Journal of Human-Computer Interaction (26)
11/12, pp. 1003-1005.
Gunawardena, C. (1995). Social presence theory and implications of interaction and collaborative
learning in computer conferencing. International Journal of Educational Telecommunications (1) 2-3,
pp. 147-166.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
31
Huberman, B. A., D. M. Romero, & F. Wu (2008). Social networks that matter: twitter under the
microscope. Working paper, arXiv:0812.1045, last accessed 2012-06-20.
Humphreys, L., 2(007). Mobile social networks and social practice: a case study of Dodgeball. Journal of
Computer-Mediated Communication (13) 1, article 17.
Huysman, M. & V. Wulf (2004). Social Capital and Information Technology. Cambridge, MA: MIT
Press.
IJsselsteijn, W. A., J. van Baren, & F. van Lanen (2003). Staying in Touch: Social Presence and
Connectedness Through Synchronous and Asynchronous Communication Media. Proceedings of the
10th International Conference on Human-Computer Interaction, Crete, Greece, pp. 924-928.
Ji, Y. G., H. Hwangbo, J. S. Yi, P. L. Rau, X. Fang, & C. Ling (2010). The influence of cultural
differences on the use of social network services and the formation of social capital. International
Journal of Human-Computer Interaction (26) 11/12, pp. 1100-1121.
Kietzmann, J. H., K. Hermkens, I. P. McCarthy, & B. S. Silvestre (2011). Social media? Get serious!
Understanding the functional building blocks of social media. Business Horizons (54) 3, pp. 241-251.
Köbler, F., S. Goswami, P. Koene, J. M. Leimeister, & H. Krcmar (2011). "NFriendConnector: Design
and evaluation of an application for integrating offline and online social networking." AIS
Transactions on Human-Computer Interaction 3, (4), pp. 214-235.
Köbler, F., C. Riedl, C. Vetter, J. M. Leimeister, & H. Krcmar (2010). Social Connectedness on
Facebook - An Explorative Study on Status Message Usage. Proceedings of 16th Americas
Conference on Information Systems, Lima, Peru.
Koroleva, K., H. Krasnova, N. Veltri, & O. Günther (2011). It’s All About Networking! Empirical
Investigation of Social Capital Formation on Social Network Sites. Proceedings of International
Conference on Information Systems, Shanghai, China, Paper 24.
Koroleva, K., H. Krasnova, & O. Günther (2010). ‘STOP SPAMMING ME!’ – Exploring Information
Overload on Facebook. Proceedings of the 16th Americas Conference on Information Systems, Lima,
Peru.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
32
Kraut, R. E. & P. Resnick (2011a). Building Successful Online Communities: Evidence-Based Social
Design. Cambridge, MA: MIT Press.
Kraut, R. E. & P. Resnick (2011b). Encouraging contribution to online communities. In R. E. Kraut and
P. Resnick (Eds) Building Successful Online Communities: Evidence-Based Social Design.
Cambridge, MA: MIT Press, pp. 21-76.
Kuwabara, K., T. Watanabe, T. Ohguro, Y. Itoh, & Y. Maeda (2002). Connectedness Oriented
Communication: Fostering a Sense of Connectedness to Augment Social Relationships. Proceedings
of Symposium on Applications and the Internet, pp. 186-193.
Lampe, C., N. Ellison, & C. Steinfield (2006). A Face(book) in the Crowd: Social Searching vs. Social
Browsing. Proceedings of the 20th Conference on Computer Supported Cooperative Work, Banff,
Canada, pp. 167-170.
Ledbetter, A. M. (2009). Measuring online communication attitude: instrument development and
validation. Communication Monographs (76) 4, pp. 463-486.
Ledbetter, A. M., J. P. Mazer, J. M. DeGroot, K. R. Meyer, Y. Mao, & B. Swafford (2011). Attitudes
toward online social connection and self-disclosure as predictors of facebook communication and
relational closeness. Communication Research (38) 1, pp. 27-53.
Lee, R. M., M. Draper, & S. Lee (2001). Social connectedness, dysfunctional interpersonal behaviors, and
psychological distress: testing a mediator model. Journal of Counseling Psychology (48) 3, pp. 310-
318.
Leimeister, J. M., K. Schweizer, S. Leimeister, & H. Krcmar (2008). Do virtual communities matter for
the social support of patients? Information Technology & People (21) 4, pp. 350-374.
Liechti, O. & T. Ichikawa (1999). A Digital Photography Framework Supporting Social
Interaction and Affective Awareness in Home Communication. Proceedings of the International
Workshop on Handheld and Ubiquitous Computing, Karlsruhe, Germany, pp. 186-192.
Lin, N. (2000). Inequality in social capital. Contemporary Sociology (29) 6, pp. 785-795.
Lin, N. (1999). Social networks and status attainment. Annual Review of Sociology (25), pp. 467-487.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
33
Long, D. A. & D. D. Perkins (2007). Community social and place predictors of sense of community: a
multilevel and longitudinal analysis. Journal of Community Psychology (35) 5, pp. 563–581.
Moore, G. C. & I. Benbasat (1991). Development of an instrument to measure the perceptions of adopting
an information technology innovation. Information Systems Research (2) 3, pp. 192-222.
Morris, M. R., J. Teevan, & K. Panovich (2010). What do People Ask their Social Networks, and Why?: A
Survey Study of Status Message Q&A Behavior. Proceedings of the 28th International Conference on
Human Factors in Computing Systems, Atlanta, GA, pp. 1739-1748.
Muise, A., E. Christofides, & S. Desmarais (2009). More information than you ever wanted: does
facebook bring out the green-eyed monster of jealousy? CyberPsychology & Behavior (12) 4, pp.
441-444.
Nahapiet, J., & S. Ghoshal (1998). Social capital, intellectual capital and the organizational advantage.
Academy of Management Review 23 (2), pp. 242-266.
Nardi, B., S. Whittaker, & E. Bradner (2000). Interaction and Outeraction: Instant Messaging in Action.
Proceedings of the ACM Conference on Computer Supported Cooperative Work, Philadelphia, PA,
pp. 79-88.
Oxendine, A., E. Borgida, J. L. Sullivan, & M. S. Jackson (2003). The importance of trust and community
in developing and maintaining a community electronic network. International Journal of Human-
Computer Studies (58) 6, pp. 671-696.
Perreault, M. and D. Ruths (2011). The Effect of Mobile Platforms on Twitter Content Generation.
Proceedings of Fifth International AAAI Conference on Weblogs and Social Media, Barcelona,
Spain.
Pfeil, U., K. Svangstu, C. Ang, & P. Zaphiris (2011). Social roles in an online support community for
older people. International Journal of Human-Computer Interaction (27) 4, pp. 323-347.
Preece, J. & B. Shneiderman (2009). The reader-to-leader framework: motivating technology-mediated
social participation. AIS Transactions on Human-Computer Interaction 1(1), pp. 13-32.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
34
Ranganathan, C., J. S. Dhaliwal, & T. S. H. Teo (2004). Assimilation and diffusion of web technologies
in supply-chain management: an examination of key drivers and performance impacts. International
Journal of Electronic Commerce (9) 1, pp. 127-161.
Ren, Y., R. Kraut, S. Kiesler, & P. Resnick (2011). Encouraging commitment in online communities. In
R. E. Kraut & P. Resnick (Eds) Building Successful Online Communities: Evidence-Based Social
Design. Cambridge, MA: MIT Press, pp. 77-124.
Resnick, P. (2001). Beyond bowling together: SocioTechnical capital. In J. Carroll (Ed.) HCI in the New
Millenium. Boston, MA: Addison-Wesley, pp. 247-272.
Resnick, P., J. Konstan, Y. Chen, & R. Kraut, 2011. “Starting New Online Communities,” in R. E. Kraut
and P. Resnick (Eds), Building Successful Online Communities: Evidence-Based Social Design,
Cambridge, MA: MIT Press, pp. 231-281.
Rettie, R., 2003. “Connectedness, Awareness and Social Presence,” Proceedings of 6th Annual
International Workshop on Presence,” Aalborg , Denmark.
Riedl, C. 2011. Tool-Supported Innovation Management in Service Ecosystems. Wiesbaden: Gabler.
Ringle, C. M., S. Wende, & A. Will (2005). SmartPLS 2.0. http://www.smartpls.de, last accessed 2012-
12-08.
Romero, N., P. Markopoulos, J. van Baren, B. de Ruyter, W. Ijsselsteijn, & B. Farshchian (2007).
Connecting the family with awareness systems. Personal and Ubiquitous Computing (11) 4, pp. 299-
312.
Scott, J. (2000). Social Network Analysis. 2nd edition, London, UK: Sage.
Sharma, R., P. Yetton, & J. Crawford (2009). Estimating the effect of common method variance: The
method-method pair technique with an illustration from tam research. MIS Quarterly (33) 3, pp. 473-
490.
Shen, K. N. & M. Khalifa (2009). Design for social presence in online communities: a multidimensional
approach. AIS Transactions on Human-Computer Interaction (1) 2, pp. 33-54.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
35
Short, J., E. Williams, & B. Christie (1976). The Social Psychology of Telecommunications. London, UK:
Wiley.
Smith, E. & D. Mackie (2000). Social Psychology. 2nd Edition, New York, NY: Psychology Press.
Stone, E. F. (1978). Research Methods in Organizational Behavior. Santa Monica, CA: Goodyear.
Stone, W. (2001). Measuring Social Capital: Towards a Theoretically Informed Measurement
Framework for Researching Social Capital in Family and Community Life. Melbourne, Australia:
Australian Institute of Family Studies.
Straub, D. W. (1989). Validating instruments in MIS research. MIS Quarterly (13) 2, pp. 147-169.
Subrahmanyam K., S. M. Reich, N. Wächter, & G. Espinoza (2008). Online and offline social networks:
use of social networking sites by emerging adults. Journal of Applied Developmental Psychology (29)
6, pp. 420-433.
Thompson, C. (2007). Clive thompson on how Twitter creates a social sixth sense. Wired Magazine, July
15th.
Tollmar, K., O. Sandor, & A. Shomer (1996). Supporting Social Awareness @Work - Design and
Experience. Proceedings of Computer Supported Cooperative Work, Boston, MA, pp. 298-307.
Twitter, (2012). About Twitter. http://twitter.com/about, last accessed 2012-06-20.
Twitter, (2011). Happy Birthday Twitter! http://blog.twitter.com/2011/03/happy-birthday-twitter.html
posted 2011-03-21, last accessed 2012-12-08.
Twitter Demographics, (2012). Double Click Ad Planner by Google. https://www.google.com/adplanner/
#siteSearch?uid=domain%253A%2520twitter.com&geo=001&lp=false, last accessed 2012-06-13.
Valenzuela, S., N. Park, & K. Kee (2009). Is there social capital in a social network site? Facebook use
and college students' life satisfaction, trust, and participation. Journal of Computer-Mediated
Communication (14) 4, pp. 875-901.
Vasalou, A., A. N. Joinsona, & D. Courvoisier (2010). Cultural differences, experience with social
networks and the nature of ‘‘True Commitment’’ in Facebook. International Journal of Human-
Computer Studies (68) 10, pp. 719–728.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
36
Wasko, M. M., & S. Faraj (2005). Why should I share? Examining social capital and knowledge
contribution in electronic networks of practice. MIS Quarterly (29) 1, pp. 35 – 57.
Weiner, M. & A. Mehrabian (1968). Language within Language: Immediacy, a Channel in Verbal
Communication. New York, NY: Appleton-Century-Crofts.
Wellman, B. & M. Gulia (1999). Net-surfers don’t ride alone: Virtual communities as communities. In B.
Wellman (Ed.) Networks in the Global Village: Life in Contemporary Communities. Boulder, CO:
Westview Press, pp. 331-366.
White, A. (2002). Social Focus in Brief: Ethnicity. London, UK: Office for National Statistics.
Williams, D. (2006). On and off the 'Net: scales for social capital in an online era. Journal of Computer-
Mediated Communication (11) 2, pp. 593-628.
Woolcock, M. (1998). Social capital and economic development: Toward a theoretical synthesis and
policy framework. Theory and Society (27) 2, pp. 151-208.
Zhao, D., & M. B. Rosson (2009). How and Why People Twitter: The Role that Micro-Blogging Plays in
Informal Communication at Work. Proceedings of GROUP, Sanibel Island, FL, pp. 243-252.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
37
Tables
Table 1. Operationalization of Constructs. Variable Items Source Social Connectedness
I feel distant from people in my Twitter network I do not feel related to most people in my Twitter network I feel like an outsider in my Twitter network I feel disconnected from the Twitter world around me I do not feel I participate with anyone or any group in Twitter I feel close to people on Twitter I am able to relate to my peers on Twitter I am able to connect with other people on Twitter I find myself actively involved in my Twitter friends’ lives
Adapted from Lee et al., 2001.
Social Presence
People on Twitter understand each other I understand the other Twitter users’ opinions The other Twitter users understand what I mean My thoughts are clear to the others on Twitter The other individuals’ thoughts are clear to me
Adapted from Shen and Khalifa, 2009.
Social Awareness
I hardly notice other users on Twitter I feel that other Twitter users are aware of my presence The other individuals do not notice me on Twitter I am often aware of other users in Twitter Others of often aware of me on Twitter
Shen and Khalifa, 2009; Biocca et al., 2001.
Note. Items in italics are reverse coded.
Table 2. Demographic Details (N = 121). Variable Categories Frequencies Gender Male
Female Not specified
60 58
3 Age Under 17
18 - 24 25 - 34 35 - 44 45 - 54 55 or older Not specified
2 46 41 21
8 1 2
Education High School Bachelor Master / Diploma PhD Not specified
20 44 42
7 8
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
38
Table 3. Psychometric Properties of Measurement Model. Construct Item Factor
Loadings Composite Reliability
AVE
Social Connectedness
SC1 SC2 SC3 SC4 SC5 SC6 SC7 SC8 SC9
0.730 0.796 0.713 0.861 0.867 0.871 0.755 0.801 0.715
0.938 0.628
Social Presence SP1 SP2 SP3 SP4 SP5
0.804 0.811 0.828 0.780 0.764
0.897 0.637
Social Awareness SA1 SA2 SA3 SA4 SA5
0.725 0.845 0.770 0.629 0.867
0.879 0.596
Table 4. Correlation Between Constructs. Frequen
cy of Use
Network Size
Social Awareness
Social Presence
Social Connectedness
Frequency of Use Network Size 0.679 Social Awareness 0.488 0.582 0.772 Social Presence 0.323 0.301 0.483 0.798 Social Connectedness
0.539 0.460 0.686 0.555 0.792
Note. Diagonal elements are the square roots of AVE (for latent constructs).
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
39
Table 5. Factor Loadings and Cross Loadings. Social
Presence Social Awareness
Social Connectedness
SP1 0.804 0.367 0.493 SP2 0.811 0.409 0.451 SP3 0.828 0.432 0.456 SP4 0.780 0.353 0.352 SP5 0.764 0.359 0.456 SA1 0.404 0.725 0.546 SA2 0.441 0.845 0.601 SA3 0.262 0.770 0.483 SA4 0.336 0.629 0.401 SA5 0.394 0.867 0.582 SC1 0.275 0.501 0.730 SC2 0.396 0.593 0.796 SC3 0.375 0.514 0.713 SC4 0.494 0.599 0.861 SC5 0.427 0.600 0.867 SC6 0.544 0.548 0.871 SC7 0.490 0.507 0.755 SC8 0.572 0.537 0.801 SC9 0.347 0.480 0.715
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
40
Table 6. Hypotheses Testing (Dependent Variable = Social Connectedness). Variable Path Coefficients Model 1 Model 2 Implications Independent Variables Social Awareness Social Presence 0.483*** 0.483*** H1 Supported Social Awareness 0.462*** 0.476*** H2 Supported Social Presence 0.267*** 0.335*** H3 Supported Network Size -0.100 -0.115 H4a Not Supported Frequency of Use 0.284*** 0.250** H5a Supported Moderating effect of Network Size Network Size * Social Awareness 0.199** H4b Supported Moderating effect of Frequency of Use Frequency of Use * Social Presence 0.075 H5b Not Supported Control Variables Age 0.037 0.059 Gender -0.049 -0.029 R2 0.585 0.630 Note. Significance levels * p < 0.05; ** p < 0.01; *** p < 0.001.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
41
Table 7. Summary of design implications that can be drawn from this study. Implication Description Social benefits of SNS
Our results indicate that online social networking sites have indeed social benefits by enabling users to build social capital. Users can successfully build social capital and feelings of social connectedness through online systems thus indicating that socially beneficial effects of personal interaction can be transferred from an offline to an online context. Designers of HCI should therefore consider adding social networking features to their website to support the creation of social capital. This design implication is particularly important for designers of online systems that are targeted at users’ social benefits such as those providing companionship and social support.
Social benefits despite limited media richness
Our results indicate that the social benefits of SNS can be achieved even with very limited media such as Twitter, which allows only the exchange of very short broadcast messages. Therefore, designers should exploit even very simple means of communication, given that even tweets are able to successfully convey social awareness, social presence, and social connectedness. System features should especially focus on the exchange of connectedness-oriented messages and status updates. Systems should specifically allow to direct messages at others and provide personal and location information to put messages into a social and geographic context.
Social awareness as fundamental system feature
Social awareness functions as a fundamental building block for higher-level constructs of social presence and social connectedness. Designs should be directed towards incorporating and enhancing awareness creating features within the SNS, as awareness not only has a direct effect on social connectedness but also enhances social connectedness indirectly through increased feelings of social presence. Possible features to consider are status indicators, status messages, and other forms of awareness displays to provide contextual information about the activities of group members (Dabbish and Kraut, 2008).
Social presence as building block for social connectedness
Evidence from our study suggests that social presence functions as an important building block for social connectedness. Consequently, designers should aim at integrating features to enhance social presence in order to enhance users’ feeling of social connectedness. Possible system features to convey the presence of other people are images of members (Resnick et al., 2011). To support social presence, Twitter includes profile pictures next to each tweet and a display of a relative time indicator such as “posted 8 minutes ago” to evoke immediacy.
Direct effect of usage frequency
Our results show that frequency of use has a strong direct effect on users’ feeling of social connectedness. To enhance social connectedness, designers of HCI should integrate system features that motivate users to use the system more. A wide variety of system features can be employed such as low threshold interfaces for easily making small contributions, identifying user contributions by name, leader boards, and conveying to users that they are unique and others in the group cannot make the contributions they are making (Preece and Shneiderman, 2009; Resnick and Kraut, 2011b). Twitter encourages usage frequency for example, through their “Trends” feature that
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
42
shows major trends in current Twitter content and thus invites the user to explore contributions made by other Twitter users who are not in the ego’s immediate network.
Moderating effect of network size
While network size has no direct effect on social connectedness, it is an important moderating factor for the effect of social awareness on social connectedness. To increase the positive effect of users’ network size, designers should implement system features that help users to grow their online social network. Possible system features are to automatically suggest other members as possible friends or import functions that allow importing an address book to grow the online social network (Preece and Schneiderman, 2009). Twitter supports users in building a larger network through the “Who to follow” panel that suggests other Twitter users. In addition to the immediate suggested list of users, the panel also offers a “refresh” button, which will generate a new set of suggested users.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
43
Figures
Social''Presence'
Social'Connectedness'
Social'Awareness'
Usage'Frequency'
Online'Social'Network'Size'
H1+
H3+
H2+
H5b
H4b
H5a+
H4a+
Figure 1. Research Model.
Other online
Cellular phone text messages
Social networking websites
Instant messaging
E−mail
Mobile / cellular phone
Telephone (land line)
Face−to−face 4.42 (0.83)2.49 (1.28)4.03 (1.03)3.98 (1.06)3.58 (1.24)4.24 (0.91)4.09 (1.12)2.44 (1.24)
4.42 (0.83)2.49 (1.28)4.03 (1.03)3.98 (1.06)3.58 (1.24)4.24 (0.91)4.09 (1.12)2.44 (1.24)
4.42 (0.83)2.49 (1.28)4.03 (1.03)3.98 (1.06)3.58 (1.24)4.24 (0.91)4.09 (1.12)2.44 (1.24)
4.42 (0.83)2.49 (1.28)4.03 (1.03)3.98 (1.06)3.58 (1.24)4.24 (0.91)4.09 (1.12)2.44 (1.24)
4.42 (0.83)2.49 (1.28)4.03 (1.03)3.98 (1.06)3.58 (1.24)4.24 (0.91)4.09 (1.12)2.44 (1.24)
2%30%4%2%9%3%5%
31%
2%24%3%9%10%2%6%22%
4%18%16%14%21%7%11%26%
36%22%38%37%33%44%32%16%
56%6%38%38%27%44%47%6%
Mean (SD) Never Very rarely Rarely Frequently Very frequently
Percent020406080100
How often do you use the following media to communicate with friends?
Figure 2. Communication media usage.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
44
BadooHabbo
StudiVZOrkut
QzoneFriendster
XingMyspace
FoursquareLinkedIn
FacebookTwitter
0 20 40 60 80 100Percentage
Netw
ork
What other social networking sites do you use?
Figure 3. Reported usage of other social networking sites.
Romantic relationshipsFamily
Work Accaintances & casula contacts
FriendsOthers
Celebrities & prolific people
0 5 10 15 20Percentage
Type
Mean composition of Twitter connection by type
Mostly long−distance friendsMostly local friends
42%
29%
19%
34%
0 20 40 60 80 100Percentage
ResponseNoneVery fewSomeAlmost allAll
What is the distance of your connections on Twitter
Figure 4. Composition of Twitter connections by type and physical distance between ego and Twitter connections. Percentages denote distribution excluding neutral rating.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
45
Figure 5. Usage frequency of random subset of 50 Twitter users of our survey sample of 121. Each users’ last 200 tweets are shown, classified by tweet type (plain tweet, mention, re-tweet, or mention within a re-tweet). The distribution of tweet types is shown in the bar plot next to it. Numbers to the far right show tweets per day (T/D). For “User 1,” a high-frequency user, the last 200 tweets span only a period of 3.6 days while for “User 50,” a low-frequency user, the last 200 tweets span almost one and a half years.
TWEETING TO FEEL CONNECTED: A MODEL FOR SOCIAL CONNECTEDNESS IN ONLINE SOCIAL NETWORKS
46
Biographical Notes
Christoph Riedl is a post-doctoral fellow at the Institute for Quantitative Social Science and
Harvard Business School at Harvard University. He has been awarded a post-doctoral fellowship of the
German Research Foundation. He holds a PhD in Information Systems from Technische Universität
München, Germany. His research interests include crowd sourcing and collective intelligence.
Felix Köbler is a PhD candidate and research assistant at the Chair for Information Systems of
Technische Universität München. His teaching and research areas include Online Social Networking,
Human-Computer Interaction as well as Ubiquitous and Mobile Computing. He received a B.Sc. and
M.Sc. in Information Systems from Technische Universität München.
Suparna Goswami is a Research Fellow at the Technische Universität München, Germany. She
holds a PhD in Information Systems from the National University of Singapore. Her research interests lie
in the areas of Web 2.0 technologies and their implications, topics in human-computer interaction and
digitally enabled inter-organizational networks.
Helmut Krcmar is a full professor of Information Systems and holds the Chair for Information
Systems at the Department of Informatics, Technische Universität München, Germany. His research
interests include computer supported cooperative work, and information systems in healthcare and e-
government.
top related