Download Understanding Members` Attachment to Social Networking Sites: An

2014 47th Hawaii International Conference on System Science
Understanding Members’ Attachment to Social Networking Sites: An
Empirical Investigation of Three Theories
Eric T.K. Lim
Australian School of Business
University of New South Wales
[email protected]
Dianne Cyr
Beedie School of Business
Simon Fraser University
[email protected]
need [33], most, if not all, of the information,
companionship, and entertainment accessible from
these SNSs are dependent on the voluntary
contributions of participants [11, 34]. Unlike
formalized relationships, SNSs exercise little authority
and control over the behavior of participating members
[9]. For SNSs to prosper, members must be proactive
in generating new content, maintaining community
cohesion and providing emotional support to others
[8]. Although the exponential growth of SNSs has been
fueled by individuals’ response to mass invitations
from friends and/or acquaintances, a majority of
members tend to be docile in their involvement. Ren et
al. [34] have thus appealed for better strategies to
foster active participation among members if SNSs
hope to continue to strive. Likewise, Wasko et al. [51]
observed that electronic networks of practice are
sustained through exchanges among a critical mass of
active members. To this end, this study endeavors to
answer the following research question: “How can
social networking sites foster active participation
among members?”
To address the aforementioned research question,
we draw on the theories of Social Identity, Social
Bond and Social Comparison to postulate that
members’ motives for participating within SNSs stem
from three distinct forms of attachment to online
communities, namely identity-based attachment, bondbased attachment and comparison-based attachment.
Specifically, we advance a theoretical model that
positions identity-based, bond-based and comparisonbased attachments as salient drivers affecting
members’ participation within SNSs. The theoretical
model is then empirically validated via an online
survey of 787 active members of SNSs. This study
therefore expands on extant literature of online
communities where communal attachments have been
revealed to be deterministic of members’ participation
within such communities [e.g., 6, 8, 17, 34, 51].
Abstract
Social Networking Sites (SNSs) are pervasive
phenomena in today’s society. With greater
connectivity and interactivity enabled through
emerging technologies, SNSs provide communication
platforms for individuals to bridge spatial and
temporal differences when making friends, sharing
experiences, socializing with others and much more.
This study therefore endeavors to shed light on this
growing trend by decomposing members’ motives for
participating within SNSs into identity-based, bondbased and comparison-based attachments. Each of
these forms of attachment in turn affects members’
cooperative and competitive mentality towards others
within SNSs. We further construct a theoretical model
of members’ communal attachments within SNSs that is
then empirically validated via an online survey of 787
active members of SNSs. Empirical findings suggest
that members’ communal attachments play an
instrumental role in sustaining their continued
participation within SNSs.
1. Introduction
The tendency of people to congregate and form
communities is inherent in the fabric of society; and
the ways in which such communities take shape and
evolve over time is a recurring theme that fascinates
both academics and practitioners alike [9]. The advent
of the Internet has bolstered the interconnectivity
among members of society and in turn, contributes to a
rapid growth of online communities. Online social
interaction flourishes as Social Networking Sites
(SNSs) become increasingly prevalent. Almost
overnight, social interactivity sites such as MySpace
[http://www.myspace.com/],
Facebook
[http://www.facebook.com/]
and
LiveJournal
[http://www.livejournal.com/] as well as media-sharing
sites such as Flickr [http://www.flickr.com/] and
YouTube [http://www.youtube.com/] have sprung up
to cater to the myriad of networking intents and usage
patterns among members.
Because SNSs represent web-enabled communities
whose members share a common purpose, interest, or
978-1-4799-2504-9/14 $31.00 © 2014 IEEE
DOI 10.1109/HICSS.2014.82
Chee-Wee Tan
Department of IT Management
Copenhagen Business School
[email protected]
2. Theory
According to Ren et al. [34], the design of SNSs for
usability and sociability often involves sophisticated
trade-offs that ultimately dictate the manner through
614
which members are motivated to participate. These
complications arise from difficulties in structuring
SNSs to satisfy members with divergent social motives
[34]. In line with the theories of social identity, social
bond and social comparison, we argue that
discrepancies among these social motives are largely
driven by distinct forms of attachment exhibited by
members participating in such online communities,
namely identity-based, bond-based and comparisonbased attachments. Distinctions among identity-based,
bond-based and comparison-based attachments
originate from members’ personal reasons for
participating within a community.
devoted to culinary matters, also concluded that social
ties to a reference group constitutes a core determinant
of interpersonal influence. As alleged by de Valck et
al. [11], the more socially involved an individual is
with an online community, the greater the likelihood
that he/she will be subjected to communal influence.
Tribe.net [http://www.tribe.net/tribes] is an
example of a SNS that promotes community over self
by facilitating the formation of special interest groups
around topical themes (e.g., hobbies, music and travel).
Membership in these special interest groups is
relatively fluid and member departures do not erode
the groups’ communal identity. Identity-based
attachment can hence be construed as one of the
motivational forces behind members’ commitment to
SNSs and it is defined in this study as the “extent to
which a member identifies with certain groups within a
social networking site”.
2.1. Identity-Based Attachment
The Social Identity Theory (SIT) holds that an
individual’s social identity consists of “those aspects of
an individual’s self-image that derive from the social
categories to which he perceives himself as belonging”
[44 p. 16]. Social identity is cultivated from the selfsegmentation of individuals into social categories [46].
Criteria for such self-segmentation can be both
objective (e.g., gender, race and geographical location)
or subjective (e.g., shared hobbies, political views or
religious beliefs) [2]. For an individual to identify with
a particular social category, he/she must be convinced
that his/her self-image aligns with qualities exhibited
by members belonging to the category [7].
Through self-segmentation, an individual assumes
the communal identity of a social category and
voluntarily depersonalizes oneself so much so that
most (if not all) behavioral judgments are anchored on
group norms and practices [22]. Due to a strong
emphasis on community over self, social identity
weakens relationships among individual members
belonging to the same social category [46]. In fact,
when members feel attached to a community based on
social identity, they regard others in the group as being
interchangeable or substitutable such that the
communal identity remains intact in the face of
turnover in membership [46]. Within the context of
online communities, Dholakia et al. [13] have supplied
empirical evidence testifying to the role of social
identity in determining members’ participation within
such communities. Likewise, Eckhardt et al. (2009)
noted that individuals’ non-adoption decisions towards
technology can be attributed to pressure from peer
groups of non-adopters. Ren et al. [34] observed that
members of many online health support groups are
attached to the groups because of their shared identity
as sufferers or survivors of a particular illness or
treatment. In such communities, Ren et al. [34]
documented that shared experiences takes precedence
over who the audiences might be. de Valck et al. [11],
in surveying active members of an online community
2.2. Bond-Based Attachment
Participants may become engaged in communities
because they feel connected to other individuals in the
group—what Ren et al. [34] term as bond-based
attachment. According to the Social Bond Theory
(SBT), communities constitute fertile breeding grounds
for the development of relationships among individuals
because through frequent interactions, it becomes easy
for people to get acquainted and become familiar with
one another [34]. The same applies to SNSs. As
documented in Utz’s [47] investigation of online
gaming: the longer the period of participation for an
online gamer, the greater is the likelihood of him/her
developing a bond with other players. Ganley and
Lampe [17], in examining the network structure of
Slashdot [http://slashdot.org], uncovered that members,
who are newer to the site, tend to associate with others
in different circles whereas those who have been
involved with the site for a much longer period are
often affiliated with closely-knitted networks. In other
words, one’s contacts within online communities could
become less diverse and more concentrated over time
[17]. Bond-based attachment is thus amplified
whenever members in online communities experience
social co-presence: a subjective feeling of being with
others in a communal environment [40].
For bond-based attachments, members feel more
connected to one another and much less to the
community as a whole. Consequently, members do not
associate themselves with the collective identity of the
community: should their friends leave the group, they
are likely to drift away as well [34]. Similarly, Hahn et
al. [19] discovered that members from open source
software developer networks display a greater
likelihood of joining a software development project if
615
they have strong collaborative ties with the project
initiator in the past.
Blogster.com [http://www.blogster.com] can be
seen as an example of a SNS that promotes bonding
between bloggers and followers. Because bloggers tend
to build up intimate relationships with followers over
time due to intensive interactions on a one-to-one
basis, their exit may spell an exodus of members from
the site. We hence conceive bond-based attachment as
a second motivational force behind participants’
commitment to SNSs and define it as the “extent to
which a member connects with specific other(s) of a
social networking site”.
that promote comparison and competition among
members through rankings based on peer evaluations.
To become a member of BeautifulPeople, dating
community, applicants are required to be voted in by
existing members of the opposite sex over a 48 hour
period. Only upon securing enough votes from
members who found an applicant to be ‘beautiful’
would he/she be granted membership. Conceivably,
SNSs such as BeautifulPeople draw on comparisonbased attachment as the primary motivational force in
promoting members’ participation and it is defined in
this study as the “extent to which a member is attracted
to his/her standing within the online community of a
social networking site”.
2.3. Comparison-Based Attachment
3. Consequence of Social Attachments
The Social Comparison Theory (SCT) states that
humans possess a drive for self-evaluation and selfimprovement [30]. Festinger [15] maintained that there
is an intrinsic motivation in oneself to know if “one’s
opinions are correct and to know precisely what one is
and is not capable of doing” (p. 217). The same
sentiments were expressed by Suls et al. [42], who
alleged that people desire an accurate assessment of
their opinions and performance so much so that in the
absence of objective standards, people look to others
(preferably those who they deemed to be similar) for
information about their relative standing. Moreover,
because people seek to confirm rather than disconfirm
their opinions of themselves, they are inclined to
perform upward comparison (comparing oneself to
someone who is better off) and crave higher standing
relative to others [4]. This upward comparison process
generally yields motivational incentives for selfimprovement [21].
Blanton et al. [5] discovered that the academic
performance of Dutch school children tend to improve
if they compared their examination grades with high
performing students. Similarly, Vrugt and Koenis [49]
showed that upward comparison is predictive of the
future scientific productivity of faculty members in
academic institutions due to a penchant for setting
higher personal goals. Conceivably, social comparison
acts as a crucial motivational force behind members’
participation in communal settings because it not only
satisfies one’s desire for an accurate assessment of
one’s performance relative to others in the community,
but it also compels one to seek progress through
comparison with better performing members [32].
In the context of online communities, Shepherd et
al. [39] demonstrated that members’ comparison-based
attachment to their standing relative to others within
electronic brainstorming teams acts as an effective
deterrent against social loafing. For instance,
BeautifulPeople [http://www.beautifulpeople.com], a
selective online dating website, is exemplary of SNSs
In this section, we synthesize extant literature in
advancing a theoretical model that delineates between
cooperative and competitive mentality of members as
plausible consequences that could arise from their
communal attachments in SNSs. We further postulate
the impact of these opposite mental states on members’
continual usage intentions towards these SNSs.
Cooperative Mentality: Both identity-based and
bond-based attachments appear to have similar positive
effects on members’ evaluation of and commitment to
communities as a whole. Studies have shown that
members’ exhibiting identity-based and/or bond-based
attachment share a tendency to view their communities
as being cohesive, thereby leading to a more favorable
evaluation of these communities as compared to other
communities [31]. Scholars found that both identitybased and bond-based attachments bolster members’
positive feelings towards their communities, which in
turn led to a boost in their level of participation in the
community, and an increased likelihood of remaining
in the group [27]. The more members who are in
regular contact with one another within a community,
the more likely they are to develop a ‘habit of
cooperation’ and act collectively [35]. We therefore
define cooperative mentality as the “extent to which a
member shares a tendency to cooperate with others
within a social networking site” and hypothesize that:
Hypothesis 1: Members’ identity-based attachment
to communal groups within social networking sites will
positively influence their cooperative mentality
towards others on the site.
Hypothesis 2: Members’ bond-based attachment to
other individual members participating within a social
networking site will positively influence their
cooperative mentality towards others on the site.
The same cannot be said for comparison-based
attachment. For comparison-based attachments,
individuals’ sense of self-worth is derived from a
craving of higher social standing relative to others in a
616
community [4]. According to the SCT, members in
comparison-based communities share a tendency to
emphasize distributive justice—amount of output
relative to the level of input invested by an individual
[41]. Yet, an over-emphasis on distributive justice will
fixate members on self-achievements and lessen the
likelihood of cooperative mentality manifesting. We
therefore hypothesize that:
Hypothesis 3: Members’ comparison-based
attachment to their relative communal standing on a
social networking site will negatively influence their
cooperative mentality towards others on the site.
Competitive Mentality: Identity-based attachments
tend to be associated with greater collaboration and
less competition among members in communities.
Sassenberg [37] revealed that members who exhibit
identity-based attachment in communities display
greater congeniality and higher behavioural
compliance to group norms as opposed to those
exhibiting bond-based attachment. For instance, Ren et
al. [34] alleged that the cooperative nature of identitybased communities is exemplified through their
welcoming stance towards newcomers. In contrast, a
greater number of membership obstacles may be
enacted for bond-based communities due to their
closely-knitted nature [34]. The same deduction was
made by Lakhani and Hippel [26], whose work on
open source development communities uncovered that
old-timers are often willing to offer help to newcomers,
even though these newcomers have not yet contributed
to the community.
Evidently, empirical studies have demonstrated that
members contribute more resources to the achievement
of public good, work harder to attain mutual goals and
slack off less when they feel committed to their
communities [24]. They also prefer equal rewards for
their contribution, a clear indicator of the absence of
competition among members [24]. We therefore define
competitive mentality as the “extent to which a
member shares a tendency to compete with others
within a social networking site” and hypothesize that:
Hypothesis 4: Members’ identity-based attachment
to communal groups within a social networking site
will negatively influence their competitive mentality
towards others on the site.
Although bond-based attachment also decreases
competition among members within SNSs, it is for a
different reason. Studies have proved that members
belonging to bond-based communities are more likely
to render assistance to specific others [26] and feel less
obligated to compensate for others’ lack of effort [48].
Members exhibiting bond-based attachment within
SNSs share greater affinity with one another and much
less to the online community as a whole [34]. As a byproduct of their apathy to the collective identity and
general functioning of the online community [25],
members exhibiting bond-based attachment may not
hold much interest in competing with uninterested
others. We therefore hypothesize that:
Hypothesis 5: Members’ bond-based attachment to
other individual members participating within a social
networking site will negatively influence their
competitive mentality towards others on the site.
Conversely, anecdotal evidence from previous
studies on comparison-based attachment in both offline
[e.g., 5, 49,] and online [e.g., 36] communities have
suggested that mindfulness of one’s standing relative to
others in these communities fosters competition among
members in a bid to outperform one another. Shepherd
et al. [39] established a causal relationship between the
quantity of ideas generated in an electronic
brainstorming session and feedback about participants’
performance. They found that participants, whose
performance is compared to a ‘dummy’ average group,
are able to generate 63% more ideas during the session
because nobody wanted to be seen as being below
‘average’ [39]. Further, participants, who engage in
comparison, continue to exhibit above-average levels
of productivity throughout the session [39]. We
therefore hypothesize that:
Hypothesis 6: Members’ comparison-based
attachment to their relative communal standing within
a social networking site will positively influence their
competitive mentality towards others on the site.
Continual Usage Intentions: Continual usage
intentions—the “extent to which a member intends to
continue using a social networking site in the future”—
governs whether members are likely to persist with a
SNS [12]. Bandura [3] attested that making regular and
high quality contributions to a community assures
members that they have a positive impact on others
within the community and reinforces members’ own
self-image as efficacious individuals. This image of
self-efficacy in turn prompts members to contribute
further on a constant basis [3]. The same observation
was recorded by Wang and Fesenmaier [50], who not
only affirmed self-efficacy as an antecedent to
members’ willingness to contribute within an online
travel community, but also uncovered reciprocity as
another critical determinant behind members’ active
participation. Similarly, the work of Skyes et al. [43]
alluded to cooperation as a core determinant of system
utilization. They discovered network density
(reflecting ‘get-help’ ties for an employee) and
network centrality (reflecting ‘give-help’ ties for an
employee) to be predictive of system usage within the
organization [43]. Consistent with the aforementioned
empirical findings, we define continual usage
intentions as the “extent to which a member intends to
617
will negatively influence their continual usage
intentions towards the site.
continue using a social networking site in the future”
and hypothesize that:
Hypothesis 7: Members’ cooperative mentality
towards participation within a social networking site
will positively influence their continual usage
intentions towards the site.
Competitive mentality, on the other hand, tends to
promote aggression and hostility among community
members [29]. Due to heightened arousal caused by
competitive instincts, Zillmann et al. [53] admitted that
individuals may hold little regard for consequences and
undertake aggressive and retaliatory measures in
response to competitors. The significance of group
cohesion in affecting one’s behavioral actions is
corroborated in Yoo and Alavi’s [52] work in which
they demonstrated that the absence of group cohesion
decreases members’ willingness to engage in task
participation. Conceivably, SNSs with competitive
members would create an antagonistic and intimidating
atmosphere that hastens the departure of others. We
therefore hypothesize that:
Hypothesis 8: Members’ competitive mentality
towards participation within a social networking site
4. Methodology
To validate our theoretical model, data was
gathered via an online survey questionnaire on a panel
of active SNS members. Respondents were instructed
to recall an SNS for which they have actively
participated during the past six months and to evaluate
their experience with this site.
Constructs from our theoretical model have been
extensively investigated in past studies and measures
can be readily obtained from extant literature with
minor modifications whenever necessary. Measures for
bond-based attachment were adapted from Jenkins
[23]. Measures for identity-based attachment were
adapted from Triandis and Gelfand [45]. Measures for
comparison-based attachment were adapted from Allan
and Gilbert [1]. Measures for cooperative mentality
were adapted from Triandis and Gelfand [45].
Measures for competitive mentality were adapted from
Lim [28]. Finally, measures for continual usage
intention were adapted from Deng et al. [12]. The list
of measurement items is summarized in Table 1.
Table1:ListofAdaptedMeasurementItems
Construct
Definition(Extentto
whichamember…)
Mean
[S.D.]
MeasurementItems
Factor
Loading
SocialAttachments
BondBased
Attachment(as
adaptedfrom
Jenkins[23])
Connectswithspecific Ifeelclosetocertainindividualswithinthesocialnetworksite.
5.03[1.40] 0.873
other(s)ofasocial
Icarealotaboutwhatcertainindividualswithinthesocialnetworksitethink
4.52[1.68] Dropped
networkingsite
aboutme.
Iwanttobethekindofpersonthatcertainindividualswithinthesocial
networksiteare.
Ifindcertainindividualswithinthesocialnetworksitetobeimportanttome. 5.08[1.44]
0.907
Iliketospendtimewithcertainindividualswithinthesocialnetworksite.
5.07[1.46]
0.902
Ifavorcertainindividualswithinthesocialnetworksite.
5.17[1.43]
0.905
Iprefercertainindividualswithinthesocialnetworksite.
5.20[1.40]
0.918
4.61[1.55]
0.895
4.69[1.46]
0.893
4.49[1.56]
0.924
4.32[1.66]
0.933
4.08[1.74]
0.879
3.63[1.83]
0.930
IdentityBased
Identifieswithcertain Iidentifyverymuchwithcertaingroupswithinthesocialnetworksite.
Attachment(as
groupswithinasocial Ifitwellintocertaingroupswithinthesocialnetworksite.
adaptedfrom
networkingsite
ThegroupsIbelongtowithinthesocialnetworksiteareanimportant
TriandisandGelfand
reflectionofwhoIam.
[45])
ThegroupsIbelongtowithinthesocialnetworksiteareimportanttomy
senseofwhatkindofpersonIam.
ThegroupsIbelongtowithinthesocialnetworksitehavealottodowith
howIfeelaboutmyself.
ComparisonBased
Attachment(as
adaptedfromAllan
andGilbert[1])
4.28[1.64] Dropped
Isattractedtohis/her Ioftencomparemyselfwithotherswithinthesocialnetworksite.
standingwithinthe
Ilikecomparingmyselfwithotherswithinthesocialnetworksite.
onlinecommunityofa
Iliketoknowmystandingamongotherswithinthesocialnetworksite.
socialnetworkingsite
IliketoknowhowIrankrelativetootherswithinthesocialnetworksite.
3.51[1.84]
0.949
3.73[1.81]
0.944
3.58[1.86]
0.955
4.79[1.40]
0.935
MentalStates
Cooperative
Mentality(as
adaptedfrom
Sharesatendencyto Itendtocooperatewithotherswithinthesocialnetworksite.
cooperatewithothers Itendtoengageincooperationwithotherswithinthesocialnetworksite.
withinasocial
Itendtoworkwithotherswithinthesocialnetworksite.
618
4.73[1.41]
0.948
4.71[1.43]
0.940
TriandisandGelfand networkingsite
[45])
Itendtohelpotherswithinthesocialnetworksiteevenifthere’snothingin
itforme.
4.81[1.46]
0.906
Competitive
Mentality(as
adaptedfromLim
[28])
Itendtowanttoperformbetterthanotherswithinthesocialnetworksite.
3.79[1.72]
0.892
Itendtobeannoyedwhenotherswithinthesocialnetworksiteperform
betterthanIdo.
3.32[1.84]
0.947
ContinualUsage
Intention(as
adaptedfromDeng
etal.[12])
Sharesatendencyto
competewithothers
withinasocial
networkingsite
Itendtopitmyselfagainstotherswithinthesocialnetworksite.
3.24[1.85]
0.957
Itendtostrivetooutdootherswithinthesocialnetworksite.
3.23[1.91]
0.956
5.61[1.31]
0.913
Intendstocontinue
Iintendtocontinueusingthesocialnetworksiteinthefuture.
usingasocial
Iwillalwaystrytousethesocialnetworksiteaspartofmyroutine.
networkingsiteinthe
IwillkeepusingthesocialnetworksiteasregularlyasIdonow.
future
Survey respondents were recruited via a
commercial marketing research firm with a track
record in online surveys. Incentives for participation in
the survey were arranged through the marketing
research firm and they take the form of a point-based
system. Through taking part in such surveys,
respondents accumulate points that are redeemable for
prizes from the marketing research firm. The computer
logs of the web server on which the electronic survey
5.19[1.47]
0.923
5.40[1.33]
0.914
was hosted recorded a total of 1,183 visits to the
questionnaire, some of which may not be unique. Of
the 1,183 visitors, 818 completed the entire
questionnaire. Therefore, a conservative estimate of
the response rate is 69.15% of invited respondents.
After deleting another 31 responses due to data runs,
we arrive at an eventual sample of 787 (66.53%) data
points for analysis. Table 2 gives a detailed breakdown
of descriptive statistics for the data sample.
Table2:DescriptiveStatisticsforDataSampleofOnlineSurvey[SampleN=787]
DemographicCharacteristic
No.ofRespondents[%]
ExperiencewithSNSs
FrequencyofVisitstoRecalledSNS
Male
336(42.69%)
3years<t<4years
>Onceperweek
Female
451(57.31%)
3years<t<4years
>Onceperday
0(0.00%)
Gender
Unwillingtodisclose
Age
Age1929
59(7.50%)
4years<t<5years
>Onceperday
Age3049
377(47.90%)
4years<t<5years
>Onceperday
Age5064
274(34.82%)
3years<t<4years
>Onceperday
Age65+
75(9.53%)
3years<t<4years
>Onceperday
Unwillingtodisclose
2(0.25%)
4years<t<5years
>Onceperweek
Lessthancollegeeducation
220(27.95%)
3years<t<4years
>Onceperday
Collegeeducationorhigher
557(70.78%)
3years<t<4years
>Onceperweek
10(1.27%)
3years<t<4years
>Onceperweek
$0$30,000
175(22.24%)
3years<t<4years
>Onceperday
$30,000$50,000
181(23.00%)
3years<t<4years
>Onceperday
$50,000$75,000
187(23.76%)
3years<t<4years
>Onceperweek
$75,000+
210(26.68%)
3years<t<4years
>Onceperweek
34(4.32%)
3years<t<4years
>Onceperday
EducationalLevel
Unwillingtodisclose
Income
Unwillingtodisclose
Table 3 depicts the spread of social networking
sites upon which respondents’ answers were based. As
can be seen from Table 3, respondents targeted a wide
variety of social networking sites in replying to the
online questionnaire, thereby assuring full variance on
the constructs of interest.
Table3:DescriptiveStatisticsforListofSocialNetworkingSites(SNSs)RecalledinSurvey[SampleN=787]
SocialNetworkingSite(SNS)
No.ofRespondents[%]
619
ExperiencewithSNSs
FrequencyofVisitingRecalledSNS
Facebook[https://www.facebook.com]
LinkedIn[http://www.linkedin.com]
655(83.23%)
3years<t<4years
>Onceperday
50(6.35%)
3years<t<4years
>Onceperweek
Twitter[https://twitter.com]
28(3.56%)
2years<t<3years
>Onceperweek
Others
24(3.05%)
3years<t<4years
>Onceperweek
GooglePlus+[https://plus.google.com]
15(1.91%)
1years<t<2years
>Onceperweek
MySpace[http://www.myspace.com]
8(1.02%)
3years<t<4years
>Onceperweek
Flickr[http://www.flickr.com]
5(0.64%)
4years<t<5years
>Onceperweek
LiveJournal[http://www.livejournal.com]
2(0.25%)
t>5years
>Onceperday
the 0.70 mark for the dataset, and even then, its value
is still much lower than the square root of intraconstruct AVE for each (see Table 5). This indicates
that respondents are able to distinguish among the
various constructs when answering the questionnaire.
5. Data Analysis
Partial Least Squares (PLS) analysis was employed
to analyze data gathered through the online survey
[18]. Because survey methodologies may be plagued
by common method bias, we applied Harman’s [20]
one-factor extraction test to the data sample. No single
factor accounted for more than 50% of total variance
explained, implying that common method bias is not a
threat in this empirical study.
The verification of the measurement model
involves estimation of internal consistency as well as
the convergent and discriminant validity of the
measurement items included in the survey instrument.
Because reflective items capture the effects of the
construct under scrutiny, internal consistency can be
assessed through standard estimates of Cronbach’s
alpha, composite reliability and the Average Variance
Extracted (AVE) [16]. After dropping 2 measurement
items due to low factor loadings (i.e., < .80) (see Table
1), the latent constructs exceed prescribed thresholds
(see Table 4), thus supporting convergent validity.
Table5:InterConstructCorrelationMatrix
AVE
[>0.50]
SocialAttachments
0.72
0.95
0.93
IdentityBasedAttachment
0.82
0.96
0.94
ComparisonBasedAttachment
0.89
0.97
0.96
MentalStates
CooperativeMentality
0.87
0.96
0.95
CompetitiveMentality
0.88
0.97
0.95
ContinualUsageIntention
0.84
0.94
0.91
0.85
For sufficient discriminant validity, the AVE from
each construct should be greater than the variance
shared between the construct and other constructs in
the model [18]. Based on the inter-construct
correlation matrix generated from SmartPLS, all
constructs display sufficient discriminant validity (see
Table 5). Only the correlation between comparisonbased attachment and competitive mentality surpass
620
CompetitiveMentality(CMM)
0.41 0.79 0.94
CooperativeMentality(COM)
0.62 0.27 0.16 0.92
ContinualUsageIntentions(CUI)
0.66 0.49 0.41 0.53 0.93
IdentityBasedAttachment(IBA)
0.70 0.70 0.58 0.50 0.67 0.91
Discriminant and convergent validity are further
confirmed when individual items load above 0.5 on
their associated factors and there is a minimum
difference of 0.10 between loadings within constructs
and cross-loadings among constructs [18]. Based on
the factor loading matrix accessible through SmartPLS,
we observe that all items load above 0.70 on their
targeted constructs (see Table 1), and that these
loadings are much higher than any cross-loadings on
any other untargeted constructs, thus supporting
convergent and discriminant validity [18].
The test of the structural model include estimates of
the path coefficients that indicate the strengths of the
relationships between independent and dependent
variables as well as R2 values that capture the amount
of variance explained by the independent variables on
its dependent counterpart. The bootstrap re-sampling
technique was employed to generate 500 random
samples from the original data set to compute for
standard errors. Figure 1 depicts statistical results from
analyzing my proposed theoretical model of members’
attachments within SNSs.
From the data analysis, a majority of hypothesized
relationships are substantiated by the empirical
evidence (see Figure 1). Identity-based attachment ( =
0.42, p < 0.001) and bond-based attachment ( = 0.37,
p < 0.001) exert positive and significant effects on
members’ cooperative mentality within SNSs whereas
comparison-based attachment ( = -0.02, p > 0.05) has
Composite Cronbach’s
Reliability Alpha()
[>0.70]
[>0.70]
BondBasedAttachment
BBA CBA CMM COM CUI IBA
BondBasedAttachment(BBA)
ComparisonBasedAttachment(CBA) 0.56 0.94
Table4:InternalConsistencyofLatentConstructs
Construct
no impact on the latter. Together, identity-based
attachment, bond-based attachment and comparisonbased attachment account for 52% of variance in
members’ cooperative mentality within SNSs.
Hypotheses 1 and 2 are hence supported whereas
hypotheses 3 is not supported. Conversely, identitybased attachment ( = 0.12, p < 0.01) and comparisonbased attachment ( = 0.77, p < 0.001) exert positive
and significant effects on members’ competitive
mentality within SNSs whereas bond-based attachment
( = -0.10, p < 0.05) have a significantly negative
impact on the latter. Together, identity-based
attachment, bond-based attachment and comparisonbased attachment account for 64% of variance in
members’ competitive mentality within SNSs.
Hypothesis 4 is not supported whereas hypotheses 5
and 6 are corroborated. Lastly, cooperative mentality
( = 0.56, p < 0.001) exert positive and significant
effects on members’ continual usage intentions
towards SNSs whereas competitive mentality ( = 0.07, p < 0.10) have a significantly negative impact on
the latter. Together, cooperative mentality and
competitive mentality account for 28% of variance in
members’ continual usage intentions towards SNSs.
Hypotheses 7 and 8 are thus substantiated.
SocialAttachments
IdentityBased
Attachment
0.423***
driven by: (1) their identity-based attachment to the
communal purpose of the online community; (2) their
bond-based attachment to specific member(s) of the
online community; (3) their comparison-based
attachment to their relative standing in the online
community, or; (4) any combination of the three. It is
also posited in the theoretical model that depending on
the form(s) of attachment being cultivated within a
SNS, it may culminate in cooperative and/or
competitive behaviors among members, leading to
differences in continual usage intentions towards the
site. Findings from our empirical investigation raise
several points of interest.
First, contrary to our expectations, identity-based
attachment was found to exert a positive and
significant effect on members’ competitive mentality
towards participation within SNSs (see Figure 1). One
plausible explanation for this observation could be due
to members’ tendency in contrasting themselves with
an “in-group target who supports their personal
identity (by serving as a downward social
comparison)” [38 p. 1604]. As noted by Schmitt et al.
[38], members, who identify strongly with a group,
could develop concerns over the erosion of their
personal identity and as a consequence, rely on ingroup comparisons as a means of maintaining their
individuality within the group. In other words,
members’ conflict between communal and personal
identities may drive them to view in-group
comparisons as a form of self-differentiation [38].
Second, empirical findings indicate that members’
comparison-based attachment does not affect their
cooperative mentality (see Figure 1). While surprising,
an explanation for this contradictory observation could
be that comparison-based attachment triggers
physiological arousal on the part of SNS members and
renders them oblivious to cooperative possibilities. As
observed by Malhotra [29], competitive environmental
cues induce physiological arousal among individuals:
narrowing their attention to concentrate on an
immediate ‘desire to win’ while abandoning cognitive
rationality. Therefore, it could be the case that
comparison-based attachment compels members of
SNSs to disregard prospects for cooperation in
exchange for self-gratification.
From a theoretical standpoint, this study answers
Ren et al.’s [34] call for a “social engineering
theoretical approach to community design and treat
online communities as social-technical systems” (p.
400). By decomposing members’ motives for
participation into identity-based attachment, bondbased attachment and comparison-based attachment,
we proffer theoretical explanations and empirical
evidence for why members would be inclined to
participate (in the event of bond-based and identity-
MentalStates
CooperativeMentality
[R2=0.518]
0.116**
0.557***
0.369***
ContinualUsage
Intentions
[R2=0.284]
BondBased
Attachment
0.103*
0.067†
0.016
ComparisonBased
Attachment
0.770***
CompetitiveMentality
[R2=0.636]
Figure1:StructuralModelAnalyticalResults
*** Correlation is significant at the 0.001 level (twotailed); **
Correlationissignificantatthe0.01level(twotailed);*Correlation
issignificantatthe0.05level(twotailed);†Correlationissignificant
atthe0.10level(twotailed).
6. Discussion
Espousing theories of social identity, social bond
and social comparison, we advance a theoretical model
claiming that members’ commitment to a SNS is
621
“Identifying the In group: A closer look at the influence of
demographic dissimilarity on employee social identity”,
Academy of Management Review 29(2), 2004, pp. 180-202.
based attachment) or not to participate (in the event of
comparison-based attachment) within SNSs. In doing
so, this study also hints at a probable reason for the coexistence of SNSs with distinct communal missions as
well as individuals’ choice to participate in more than
one online community simultaneously.
Pragmatically, this study could be of interest to
practitioners.
The
delineation
of
members’
participation motives into the three forms of
attachment (i.e., identity-based attachment, bond-based
attachment and comparison-based attachment)
represents a formalized method for practitioners to
strategize about the evolution of SNSs. Through an indepth appreciation of the drivers of members’
participation on SNSs, practitioners can be better
informed in deciding the direction with which to grow
a specific online community depending on its
communal vision. For instance, while open source
software development teams (e.g., Sourceforge
[http://sourceforge.net/]) could benefit from the
inducement of cooperative mentality to secure
members’ sustained participation, gaming communities
(e.g., msn games [http://zone.msn.com/en-us/home])
and
online
dating
sites
(e.g.,
Badoo
[http://badoo.com/en-ca/]) may be better suited for
competitive members where short bouts of active
participation in a tournament setting within a limited
time frame is preferred over protracted involvement.
[8] Cheung, C. M. K., and Lee, M. K. O. “A theoretical
model of intentional social action in online social networks”,
Decision Support Systems 49(1), 2010, pp. 24-30.
[9] Clemons, E. K. “The complex problem of monetizing
virtual electronic social networks”, Decision Support
Systems (48)1, 2009, pp. 46-56.
[10] Coleman, J. Foundations of social theory. Boston,
MA: Harvard University Press, 1990.
[11] de Valck, K., van Bruggen, G. H., and Wierenga, B.
“Virtual communities: A marketing perspective”, Decision
Support Systems 47(3), 2009, pp. 185-203.
[12] Deng, L., Turner, D. E., Gehling, R., and Prince, B.
“User experience, satisfaction, and continual usage intention
of IT”, European Journal of Information Systems 19(1),
2010, pp, 60-75.
[13] Dholakia, U. M., Bagozzia, R. P., and Pearob, L. K.
“A social influence model of consumer participation in
network- and small-group-based virtual communities”,
International Journal of Research in Marketing 21(3), 2004,
pp. 241-263.
[14] Eckhardt, A., Laumer, S., and Weitzel, T. “Who
influences whom? Analyzing workplace referents’ social
influence on IT adoption and non-adoption”, Journal of
Information Technology 24(1), 2009, pp. 11-24.
[15] Festinger, L. “A theory of social comparison
processes”, Human Relations 7(2), 1954, pp. 117–140.
This research was completed with funding support from the Social
Sciences and Humanities Research Council of Canada (SSHRC).
[16] Fornell, C., and Larcker, V. F. “Evaluating structural
equation models with unobservable variables and
measurement error”, Journal of Marketing Research, 18(1),
1981, pp. 39-50.
7. References
[1] Allan, S., and Gilbert, P. “A social comparison scale:
Psychometric
Properties
and
Relationship
to
Psychopathology”, Personality and Individual Differences
19(3), 1995, pp. 293-299.
[17] Ganley, D., and Lampe, C. “The ties that bind: Social
network principles in online communities”, Decision Support
Systems 47(3), 2009, pp. 266-274.
[2] Amichai-Hamburger, Y. “Internet minimal group
paradigm”, Cyber Psychology and Behavior 8(2), 2005, pp.
140–142
[18] Gefen, D., and Straub, D. “A practical guide to
factorial validity using PLS-graph: Tutorial and annotated
example”, Communications of the Association for
Information Systems, 16(5), 2005, pp. 91-109.
[3] Bandura, A. Self-efficacy in changing societies.
Cambridge: Cambridge University Press, 1995.
[19] Hahn, J., Moon, J. Y., and Zhang, C. “Emergence of
new project teams from open source software developer
networks: Impact of prior collaboration ties”, Information
Systems Research 19(3), 2008, pp. 369-391.
[4] Baumeister, R. F. and Bushman, B. J. Social
Psychology and Human Nature, Belmont, CA: Wadsworth,
2008.
[20] Harman, H. H. Modern factor analysis. Chicago, IL:
University of Chicago Press, 1967.
[5] Blanton, H., Buunk, B. P., Gibbons, F. X., and Kuyper,
H. “When better-than-others compare upward: Choice of
comparison and comparative evaluation as independent
predictors of academic performance”, Journal of Personality
and Social Psychology 76(3), 1999, pp. 420–430.
[21] Helgeson, V. S., and Mickelson, K. D. “Motives for
social comparison,” Personality and Social Psychology
Bulletin, 21(11), 1995, pp. 1200–1209.
[22] Hogg, M. A., and Terry, D. J. “Social identity and selfcategorization processes in organizational contexts”,
Academy of Management Review 25(1), 2000, pp. 121-140.
[6] Bruque, S., Moyano, J., & Eisenberg, J. “Individual
adaptation to IT-induced change: The role of social
networks”, Journal of Management Information Systems
25(3), 2008, pp. 177-206.
[7]
[23] Jenkins, P. H. “School delinquency and the school
social bond”, Journal of Research in Crime and Delinquency
34(3), 1997, pp. 337-367.
Chattopadhyay, P., Tluchowska, M., and George, E.
622
and Nunamaker, J. F., Jr. “Invoking Social Comparison to
Improve Electronic Brainstorming: Beyond Anonymity”,
Journal of Management Information Systems 12(3), 1995,
pp. 155–170.
[24] Karau, S. J., and Williams, K. “Social loafing: A metaanalytic review and theoretical integration”, Journal of
Personality and Social Psychology 65(4), 1993, pp. 681–706.
[25] Krackhardt, D., and Porter, L. W. “The snowball
effect: Turnover embedded in communication networks”,
Journal of Applied Psychology 71(1), 1986, pp. 50–55.
[40] Slater, M., Sadagic, A. and Schroeder, R. “Smallgroup behavior in a virtual and real environment: A
comparative study”, Presence, Teleoperators and Virtual
Environments 9(1), 2000, pp. 37–51.
[26] Lakhani, K. R., and Hippel, E. V. “How open source
software works: ‘Free’ user to user assistance”, Research
Policy 32(6), 2003, pp. 923–943.
[41] Song, J., and Kim, Y. J. “Social influence process in
the acceptance of a virtual community,” Information systems
frontier 8(3), 2006, pp. 241-252.
[27] Levine, J. M., and Moreland, R. L. Small groups. In D.
T. Gilbert, S. T. Fiske & G. Lindzey, (Eds.), The handbook
of social psychology (pp. 415–469). Boston, MA: McGrawHill, 1998.
[42] Suls, T., Martin, R., and Wheeler, L. “Social
comparison: Why, with whom, and with what Effect?”
Current Directions in Psychological Science 11(5), 2002, pp.
159-163.
[28] Lim. L. “A two-factor model of defensive pessimism
and its relations with achievement motives”, The Journal of
Psychology 143(3), 2009, pp. 318-336.
[43] Sykes, T. A., Venkatesh, V., and Gosain, S. “Model of
acceptance with peer support: A social network perspective
to understand employees’ system use”, MIS Quarterly 33(2),
2009, pp. 371-393.
[29] Malhotra, D. “The desire to win: The effects of
competitive arousal on motivation and behavior”,
Organizational Behavior and Human Decision Processes
111(2), 2010, pp. 139-146.
[44] Tajfel, H., and Turner, J. C. The social identity theory
of intergroup behavior. In S. Worchel and W. G. Austin,
(Eds.), Psychology of intergroup relations (pp. 7-24).
Chicago, IL: Nelson-Hall, 1986.
[30] Michinov, N., and Primois, C. “Improving productivity
and creativity in online groups through social comparison
process: New evidence for asynchronous electronic
brainstorming”, Computers in Human Behavior 21(1), 2005,
11-28.
[45] Triandis, H. C., and Gelfand, M. J. “Converging
measurement of horizontal and vertical individualism and
collectivism”, Journal of Personality and Social Psychology
74(1), 1998, pp. 118-128.
[31] Michinov, N., Michinov, E., and Toczek-Capelle, M.
C. “Social identity, group processes, and performance in
synchronous computer-mediated communication”, Group
Dynamics – Theory Research and Practice 8(1), 2004, pp.
27–39.
[46] Turner, J. C. Social categorization and the selfconcept: A social cognitive theory of group behavior. In E. J.
Lawler (Ed.), Advances in group processes (pp. 77– 122).
Greenwich, CT7: JAI Press, 1985.
[32] Monteil, J. M., and Huguet, P. Social context and
cognitive performance: Towards a social psychology of
cognition. Hove, East Sussex: Psychology Press, 1999.
[47] Utz, S. “Social identification and interpersonal
attraction in MUDs”, Swiss Journal of Psychology 62(2),
2003, pp. 91–101.
[33] Preece, J. Online communities: Designing usability,
supporting sociability. Chichester: Wiley, 2000.
[48] Utz, S., and Sassenberg, K. “Distributive justice in
common-bond and common-identity groups”, Group
Processes and Intergroup Relations 5(2), 2002, pp. 151–162.
[34] Ren, Y., Kraut, R., and Kiesler, S. “Applying common
identity and bond theory to design of online communities”.
Organization Studies, 28(3), 2007, pp. 377-408.
[49] Vrugt, A., and Koenis, S. “Perceived self-eƥcacy,
personal goals, social comparison, and scientific
productivity. Applied Psychology”, An International review
51(4), 2002, pp. 593–607.
[35] Robert Jr., L. P., Dennis, A. R., and Ahuja, M. K.
“Social capital and knowledge integration in digitally
enabled teams”, Information Systems Research 19(3), 2008,
pp. 314-334.
[36] Roy, M. C., Gauvin, S., and Limayen, M. “Electronic
group brainstorming: The role of feedback on productivity”,
Small Group Research 27(2), 1996, pp. 215–247.
[50] Wang, Y., and Fesenmaier, D. R. “Assessing
motivation of contribution in online communities: An
empirical investigation of an online travel community”,
Electronic Markets, 13(1), 2003, pp. 33-45.
[37] Sassenberg, K. “Common bond and common identity
groups on the Internet: Attachment and normative behavior
in on-topic and off-topic chats”, Group Dynamics 6(1), 2002,
pp. 27–37.
[51] Wasko, M. M., Teigland, R., and Faraj, S. “The
provision of online public goods: Examining social structure
in an electronic network of practice”. Decision Support
Systems 47(3), 2009, pp. 254-265.
[38] Schmitt, M. T., Silvia, P. J., and Branscombe, N. R.
“The intersection of self-evaluation maintenance and social
identity theories: Intragroup judgment in interpersonal and
intergroup contexts”, Personality and Social Psychology
Bulletin, 26(12), 2000, pp. 1598-1606.
[52] Yoo, Y., and Alavi, M. “Media and group cohesion:
Relative influences on social presence, task participation, and
group consensus,” MIS Quarterly 25(3), 2001, pp. 371-390.
[53] Zillmann, D., Bryant, J., Cantor, J. R., and Day, K. D.
“Irrelevance of mitigating circumstances in retaliatory
behavior at high levels of excitation”, Journal of Research in
Personality, 9(4), 1975, pp. 282–293.
[39] Shepherd, M. M., Briggs, R. O., Reinig, B. A., Yen, J.,
623