The role of attachment style in building social capital from a social

networking site: The interplay of anxiety and avoidance

Doo Young Lee

Department of Interaction Science, Sungkyunkwan University, 53 Myungryun-dong Jongro-gu, International Hall, Seoul 110-745, South Korea
a r t i c l e i n f o
Article history:
Available online 5 March 2013
Keywords:
Attachment theory
Attachment style
Social capital
Social networking site (SNS)
a b s t r a c t
Drawing on attachment theory, the present study examine s the attachment styles of individuals relative
to two ways of building social capital bonding social capital and bridging social capital. In trying to
relate attachment theory to the use of SNS, the present study argues that bonding social capital is
reected in the use of SNS for forming attachment bonds from trust-based strong ties, while bridging
social capital is reected in the use of SNS for causal afliations among more socially distant people.
The concep tual model was validated through an online survey completed by 368 Facebook users. Two
hierarchical multiple regression analyses indicated several results. First, avoidance attachment was sig-
nicant and negatively predictive of both bonding social and bridging social capital. Second, both bonding
social capital and bridging social capital reported by respo ndents appeared to be greatest under condi-
tions of low anxiety attachment coupled with low avoidance attachment. Third, levels of Facebook usage
were signicant and independently predictive of bridging social capital.
2013 Elsevier Ltd. All rights reserved.
1. Introduction
A very recent study by Egan and Moreno (2011) reported an
interesting observation that people tend to use online social net-
working sites (SNSs) such as Facebook as outlets for relieving their
concerns. The nding is notable because the study demonstrates
that seeking out comfort or emotional support by individuals man-
ifests in the developmen t of interpers onal relationshi ps with oth-
ers through the use of SNS.
Drawing on attachme nt theory (Bowlby, 1969, 1973, 1980 ), the
present study seeks to relate the perspective of attachment styles
to the use of SNS. This work was inspired by the idea that, [in]
adulthood, attachment appears to be driven less by biological
needs. . . and more by interpersonal needs . . . (Welch & Houser,
2010, p. 354 ). In this study, attachment styles are examine d relative
to two ways of building social capital bonding social capital and
bridging social capital in order to address how SNS are used for
two different purposes (Ellison, Steineld, & Lampe, 2007 ). In trying
to relate attachment theory to the use of SNS, the present study ar-
gues that bonding social capital is reected in the use of SNS for
forming attachment bonds from trust-based strong ties, while
bridging social capital is reected in the use of SNS for causal afl-
iations among more socially distant people (Putnam, 1995, 2000;
Williams, 2006 ).
The purpose of this study is twofold: rst, to develop a concep-
tual framework for identifying the role of individual differences in
attachme nt styles in building bonding and bridging forms of social
capital from SNS, and second, to determine the validity of the
framewor k by testing a series of research hypotheses. A survey
study was utilized to collect sample data. The next section reviews
attachme nt theory, attachment styles, and bonding and bridging
forms of social capital. Developmen t of the conceptual framework
and a series of research hypothes es follow, and the design of the
survey study and empirica l ndings are subsequently discussed.
Conclusio n is presented in the nal section.
2. Literature review
2.1. Attachme nt theory
Attachment theory has its origin in understand ing the develop-
ment mechanism of strength of feeling between infants and their
main caregivers (i.e., attachment gures) (Bowlby, 1969, 1973,
1980). Attachment refers to [ . . .] an enduring affectional [or devo-
tional] bond of substanti al intensity (Armsden & Greenberg, 1987,
p. 428 ). A basic tenet of this theory is that infants are motivated to
engage in attachment behavioral systems that promote and main-
tain their proximity to attachme nt gures in order to achieve felt
security (Sroufe & Waters, 1977 ), particularly in times of defense
against separation or loss (Bowlby, 1988; Meredith, Ownsworth, &
Strong, 2008; Wilkinson, 2004 ). The theory postulate s that re-
peated interactions between infants and attachment gures con-
tribute to the developmen t of an internal working model of
attachme nt. It is an internal representat ion, or mental model, of
the understand ing of the infants themselves and their attachment
0747-5632/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chb.2013.01.012

Corresponding author. Tel.: +82 10 3649 5964; fax: +82 2 740 1856.
E-mail address: bugabuga@skku.edu
Computers in Human Behavior 29 (2013) 14991509
Contents lists available at SciVerse ScienceDi rect
Com puters in Human Behavior
j our nal homepage: www. el sevi er . com/ l ocat e/ comphumbeh
gures (Bowlby, 1980; Bretherton, Ridgewa y, & Cassidy, 1990 ). If
secure attachme nt is established due to responsive, sensitive, and
consistent caregiving by attachment gures, for example, then
the infants are likely to form positive internal working models of
themselves as worthy of care and of others as available , trustwor-
thy, and not rejecting of them (Lopez, 2001; Patterson, 1997; Zil-
ber, Goldstein, & Mikulincer, 2007 ). Conversel y, infants develop
negative models of themselves as unlovable as a result of more
unresponsive and unreliable interactio n with their attachment g-
ures. Those with negative internal models tend to develop a fear of
interpersonal rejection, and consequently, the expectati on of rejec-
tion motivates them to become reluctant to close, intimate, or
dependent relationship s with others in order to avoid rejection
(Keefer, Landau, Rothschild, & Sullivan, 2012; Mikail, Henderson,
& Tasca, 1994; Mikulincer & Shaver, 2007 ).
2.2. Attachment style: Anxiety and avoidanc e
Importantly, internal working models developed during infancy
are assumed to inspire coherent and adaptive patterns of attach-
ment behaviors and expectations in other interpers onal relation-
ships througho ut the life span of individua ls (Ainsworth, Blehar,
Waters, & Wall, 1978; Armsden & Greenberg, 1987; Bowlby,
1969; Cassidy, Ziv, Mehta, & Feeney, 2003; Klohnen & John, 1998;
Main, 2000; Main, Kaplan, & Cassidy, 1985; Rholes, Simpson, Tran,
Martin, & Friedman , 2007 ). Moreove r, individuals who demonstrat e
social behaviors in much the same patterns as the attachment
behaviors of their infancy have stability in their attachment styles
(Bartholomew, 1993 ). Attachment style is described as [ . . .] the
propensity to establish affectional bonds with other people [that]
guide behavior and perception in relationship [s] (Bakker, Van
Oudenhove n, &Van Der Zee, 2004, p. 388 ). Anumber of studies have
been framed around attachment styles to understand a wide variety
of social relationship s, including friendshi p (Bippus & Rollin, 2003 ),
romantic relationship s (Feeney & Noller, 1990; Moore & Leung,
2002), marriage (Senchak & Leonard, 1992 ), mentorship (Bernier,
Larose, & Soucy, 2005 ), and online relationships (Ye, 2007 ).
Several attachment styles have been conceptualized as impor-
tant indicators of early childhood attachment patterns and subse-
quent behavior (Ainsworth et al., 1978; Main & Solomon, 1986;
Main et al., 1985 ). In the context of attachment styles in adulthood ,
a four-category model of attachment styles has received consider-
able attention in the eld of social science (Bartholomew, 1993;
Bartholomew & Horowitz, 1991; Meredith et al., 2008 ). The model
is built upon two dimensions of anxiety and avoidance, with par-
ticular reference to the internal working models of self and others.
For example, classications of high anxiety or low anxiety reect
negative or positive views of self, whereas classications of high
avoidance or low avoidance are associated with negative or posi-
tive attitudes toward others, respectively . Thus the model is
crossed with two levels of anxious attachment and two levels of
avoidance attachment, yielding four possible categories of attach-
ment styles. The four possible attachme nt styles are (1) secure
(low anxiety and low avoidance), (2) preoccupied or anxious
ambivalent (high anxiety and low avoidance), (3) dismissing or dis-
missing-avo idant (low anxiety and high avoidance), and (4) fearful
or fearful-a voidant (high anxiety and high avoidance) (Bartholo-
mew & Horowitz, 1991 ).
2.3. Social capital
Social capital broadly, social networks is conceptualized as a
set of actual or potential resources embedded in relationships
among social actors of mutual acquaintance in a group (Bourdieu ,
1986; Wasko & Faraj, 2005 ). The basic idea behind social capital is
that an individual can gain value through accumulating relation-
ships with other people (Coleman, 1988 ). In general, the more re-
sources people access when needed, the greater social capital
individua ls believe they possess, and consequentl y, the greater
chance that they can achieve their goals (Paxton, 1999; Yang &
Farn, 2009 ).
The notion of social capital has been used in a wide range of
social science settings (Adler & Kwon, 2002 ). Due to the nature of
its ambiguity and multi-dimensi onality, a variety of different
operation alizations and variables have contributed to the different
representat ions of social capital (Bourdieu, 1986; Coleman, 1988 ).
Although there has been much debate about the utility of simple
dichotom ized assertion (Lin, 2008 ), a conceptualizat ion of
opposing bridging and bonding forms of social capital has gained
wide acceptab ility in literature (Ellison et al., 2007; Ko & Kuo,
2009; Putnam, 1995, 2000; Steineld, Ellison, & Lampe, 2008 ).
Bonding social capital refers to resources obtained from within-
group ties (Yuan & Gay, 2006 ). Bonding social capital builds strong
links between like-mind ed people, such as groups of close friends
or families (Schuller, Baron, & Field, 2000 ). In contrast, bridging
social capital is a pattern of resources that can be accessed through
external ties with people. Bridging social capital builds weak,
loose, or fragile connections between heteroge neous groups
lacking internally cohesive or emotionally close relationship s
(Granovet ter, 1982 ).
2.4. Bonding and bridging forms of social capital and the use of a social
networkin g site
In the context of SNS, developmen t of bonding social capital can
be found in cases where users create relationship s with others in
order to share values and prospective goals. The benets of bond-
ing social capital include emotional support and ample resource s
whenever needed (Pfeil, Arjan, & Zaphiris, 2009 ). On the other
hand, SNS users tend to expand their social networks by inviting
others to join and/or by gaining access to the contacts of others,
which corresponds to the development of bridging social capital.
A study by Donath and Boyd (2004) indicated that current technol-
ogy affords SNS users a way to increase the size of weak ties over
distance. From a system-oriente d point of view, for example,
advancemen t in Web technolo gies is presumed to easily create
weak and distant ties due to great control over user customiz ation
of interface appearance, navigation, and privacy settings (Yonder &
Stutzman , 2011 ). Several studies have discussed the potential for
large networks of these weak ties. Studies by Ellison et al. (2007)
and Seibert, Kraimer, and Liden (2001) pointed out that bridging
social capital may provide individuals with a source of useful infor-
mation or new perspectives about one another. Florida (2002) and
Ulhi (2005) noted that openness to new ideas relies greatly on
weak social ties. Taking the ideas of existing studies into account,
the present study argues that bonding social capital developed in
the context of SNS is the product of tangible resources based on
interpers onal trust-based strong ties, while bridging social capital
reects causal afliation-based weak ties among SNS users.
3. Developmen t of hypotheses
3.1. The relationsh ip between attachment style and bonding social
capital
There is an important distinction between attachme nt behav-
iors and attachment bonds. The former denotes forms of behavior
that attract, promote, and maintain proximity to the attachment
gures of individuals, while the latter refers to the interpretation
of relationship s with attachment gures by individuals (Flaher-
ty ?tul?> & Sadler, 2011, p. 115 ). The use of SNS by individuals
1500 D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509
for the developmen t of bonding social capital may be due to the as-
pect of attachment behavior that seeks to establish coherent and
consistent relationshi ps. In turn, bonding social capital itself cap-
tures the attachment bonds of users interpretations of their strong
relationship s with others.
According to the four-catego ry model of attachment styles dis-
cussed above (Bartholomew, 1993 ; Bartholomew & Horowitz,
1991), the style of anxious attachme nt is marked by fear of being
rejected or abandoned by attachme nt gures, and consequently,
by judging oneself as unworthy or unlovable. On the other hand,
individuals with high scores for the style of avoidance attachme nt
tend to demonstrat e discomfort in interacting with counterparts or
peers, and to view others as unavailable or untrustworthy . Existing
literature has proven a pattern that anxious and/or avoidance
attachment dimensio ns are negatively related to efforts at support
seeking and belonging to a given domain (Collins & Feeney, 2000;
Mikulincer & Shaver, 2009; Simpson, Rholes, & Nelligan, 1992 ).
Among reghters, for example, a study by Landen and Wang
(2010) found that high scores in both the anxiety and avoidance
dimensions were negatively associated with perceived work cohe-
sion, which is [ . . .] an individuals emotional connectio n with and
perceived support from coworkers [. . .] (p. 146).
In retrospect, attachment theory postulates that people with po-
sitive internal working models of self and others are more likely to
use their signicant social relationship s as dependable , reliable, and
available resources for gaining emotional bonds in comparison to
people with negative views of self and/or others (Ainsworth,
1989; Bowlby, 1969, 1973, 1980 ). Accordingly, individua ls with
low scores in both the anxiety and avoidance attachment dimen-
sions may develop bonding social capital in the context of SNS with
a sense of condence that they will gain emotional support, such as
comfort or encouragem ent, from specic people. Conversely, fear-
ful-avoidan ts are characterized by having mixed feelings of both
strong belief in social independence and fear of rejection (Bartholo-
mew & Horowitz, 1991; Carvallo & Gabriel, 2006; Simpson, Rholes,
& Phillips, 1996 ). Thus this category of people may be less likely to
develop bonding social capital in the context of SNS with the same
type of condence that is demonst rated by the secure category of
people. Interestin gly, a study by Mikulincer and Selinger (2001)
illustrated that people that are high anxiety seek attachment goals
exclusively. This nding is partly in line with the work of Wang,
Noe, Wang, and Greenberger (2009), who found that the willingness
of protgs to participate in mentoring relationship s was signi-
cantly inuenced by their levels of avoidance, but not by levels of
anxiety. Dewitte and Houwer (2008) noted that [ . . .] anxious indi-
viduals stay committ ed to proximity goals [. . .] (p. 676). Accord-
ingly, it becomes possible to accept that anxiety impacts bonding
social capital at high levels of avoidance attachment more than at
low levels of avoidance attachment. Consideri ng these ndings in
aggregate, it is hypothesize d that avoidance attachment is nega-
tively associated with the extent to which an individua l develops
bonding social capital, and that avoidance interacts with anxiety
to inuence bonding social capital.
Hypothesis 1. Avoidanc e negatively affects bonding social capital.
Hypothesis 2. The interaction between anxiety and avoidance
occurs such that anxiety impacts bonding social capital at high lev-
els of avoidance more than at low levels of avoidance.
3.2. The relationship between attachmen t style and bridging social
capital
Afliation refers to the strength of the desire to be with others
(Mawson, 2007, p. 34 ), even if a person does not feel particularly
close to other people. Activities that tend to afliate participa nts
with others may be analogous to the developmen t of bridging social
capital in that the process is characterized by a lack of depth in trust
(Cheung & Kam, 2010; Williams, 2006 ). A study by Mikulincer and
Selinger (2001) noted that afliation behavior is the behavior that
[aims] at attaining [. . .] stimulation, and knowled ge, [. . .] with
whom one can exchange information and services [. . .] (p. 84).
Several studies have discussed the distinction between attach-
ment behavior and afliation behavior (Cassidy, 1999; Mikulincer
& Selinger, 2001 ). Rosenthal and Kobak (2010) argued that attach-
ment behavior is based on the subjectiv e past experience of an indi-
vidual and is thus highly selective regarding the individuals
perception of attachment gures, whereas afliation behavior is
less selective in the sense that an individual may be open to seeking
a wide range of people with whom to interact. Using the balance
paradigm , studies by Mikulincer and Selinger (2001) and OConnor
and McCartne y (2007) have explained how the attachment system
interacts with the afliation system (i.e., the exploratory system,
which is the latter works term for the afliation system, p. 459).
These studies argue that an individuals perception of attachment
bonds to other(s) is likely to deactivate the attachment system
and activate the afliation system, thereby encourag ing individuals
to interact with their environment. Consequentl y, individual differ-
ences in attachme nt styles may be likely to determine the extent to
which individuals engage in afliation behavior. With regard to the
secure category of people, for example, Mikulincer and Selinger
(2001) noted that secures may be able to switch between the
attachme nt system and afliation system with more exible and
adaptive patterns. The adaptability of secure people relies on their
condence that they will receive support from attachme nt gures
whenever needed. Based on the nding that high anxiety people
exclusivel y do not seek afliation goals (Mikulincer & Selinger,
2001), people who score high on dimensio ns of anxiety and low
on dimensions of avoidance (anxiousambivalent), and people
scoring high in both the anxiety and avoidance dimensions (fear-
ful-avoidan ts) may tend not to develop bridging social capital as
much as do secure people. In line with the claims of Mikulincer
and his colleagues (Mikulincer &Selinger, 2001; Mikulincer, Shaver,
& Horesh, 2006 ), the present study predicts that avoidance attach-
ment is negatively associated with the extent to which an individ-
ual develops bridging social capital, and that anxiety and
avoidance interact to inuence bridging social capital.
Hypothesis 3. Avoidance negatively affects bridging social capital.
Hypothesis 4. The interaction between anxiety and avoidance
occurs such that avoidance impacts bridging social capital at low
levels of anxiety more than at high levels of anxiety.
4. Study methodol ogy
4.1. Study context
The common understa nding is that people do not usually use a
SNS to form connectio ns with new friends or strangers due to con-
cerns about privacy or intrusivenes s (Boyd & Ellison, 2007; Ellison
et al., 2007; Lampe, Ellison, & Steineld, 2007 ). Instead, users seem
to prefer SNS for the maintenance of existing ofine contacts (Din-
dia & Canary, 1993; Subrahman yam, Reich, Waechter , & Espinoza,
2008). However , literature also indicates that new interpers onal
relationship s in the context of SNS are facilitate d by representing
interests or concerns among new contacts with whom individuals
intend to share connections (Peluchet te & Karl, 2008; Utz & Beuke-
boom, 2011 ). In other words, any SNS offers users the potential to
befriend strangers to meet their own interpers onal needs. Thus it
D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509 1501
may be interesting to study the use of SNS in the context of its pur-
pose to seek out new relationships to meet interpersonal needs.
The primary concern of the present work is the creation of new
relationship s in the context of SNS, rather than the maintenanc e
of existing ofine relationship s among users.
4.2. Development of measurement instruments
The measurement instruments used to assess the two construct
measures of bonding and bridging forms of social capital were
developed based on the adaptation and modication of two ve-
item scales pre-validate d in a study by Ellison, Steineld, and
Lampe (2007) (see Table 1). A seven-po int Likert scale anchored
from 1 (=strongly disagree) to 7 (=strongly agree) was used.
The method used for measuring attachment style was directly
adapted from the Adult Attachment Questionnai re (AAQ) used in
a study by Simpson et al. (1996) (see Table 1). The questionnair e
consisted of 17 items, all measured on a seven-point Likert scale
ranging from 1 (=strongly disagree) to 7 (=strongly agree). Nine
items on the AAQ (AAQ4, and AAQ10AAQ17) measure dimensions
of anxiety attachme nt while the remaining eight items assess
dimensions of avoidance attachme nt (Hazan & Shaver, 1987; Simp-
son et al., 1996 ). The measureme nt instrument also contained
nominal item scales to register demogra phic information including
age, gender, grade level, and amount of Facebook use per day.
The English version of the questionnaire was translated to Kor-
ean by the author. A panel of domain experts reviewed the rst
version of the questionnaire for clarity, and several items were re-
vised accordin g to comments from the experts. Subsequentl y, the
revised instrument was pilot-tested on a convenience sample of
30 undergradu ate students (15 females and 15 males) at Korea
University with characteristics similar to the target sample. The
purpose was to ensure internal consistency among each construct.
Cronbachs alpha coefcients were calculated for all four of the
construct measures. All of the subscales had alpha coefcients of
0.7 or more, providing preliminary evidence for high internal con-
sistency (DeVellis, 2003 ).
4.3. Sampling
The present study used the quota and non-random convenience
sampling method. A contracte d survey research company adminis-
tered the collection of data.
1
The target sample included represen ta-
tives of (1) a gender-bala nced group, (2) South Korean
undergra duate students aged 1925 years enrolled in a 4-year uni-
versity, and (3) people having at least 1 year of experience with Face-
book use. Invitations to particip ate in the online survey were sent to
7804 individual s. Out of the total, 2074 qualied individu als ac-
cessed the online survey , and 440 respondent s completed all of the
question s. Usable data was collected from 368 respond ents, yielding
a net response rate of 4.72%.
5. Results
The data was analyzed using SPSS (version 20.0) and SAS (ver-
sion 9.2) for Windows where appropriate . A series of scale rene-
ment processes, including calculations for internal consistency and
explorator y factor analysis (EFA), were performed to eliminate
extraneous item measures. A total of four research hypothes es
were tested through two ordinary hierarchical multiple regression
analyses. P-values less than .05 were considered to be signicant in
all cases.
5.1. Sample demographi cs
The nal sample consisted of 171 males (46.5%) and 197 fe-
males (53.5%). The majority of respondents (99.2%) were 20
Table 1
Measurement instruments: bonding social capital, bridging social capital, and the Adult Attachment Questionnaire (AAQ).
Description of construct/item
Bonding social capital
BND1: I use Facebook because there is someone (whom I do not know in real life) that I can turn to for advice about making very important decisions
BND2: I use Facebook because the people (whom I do not know in real life) that I interact with on Facebook would share their last dollar with me
BND3: I use Facebook because the people (whom I do not know in real life) that I interact with on Facebook would put their reputation on the line for me
BND4: I use Facebook because the people (whom I do not know in real life) that I interact with on Facebook would help me ght an injustice
BND5: I use Facebook because there are several people (whom I do not know in real life) on Facebook that I trust to help solve my problems
Bridging social capital
BRD1: I use Facebook because interacting with people (whom I do not know in real life) makes me feel connected to the bigger picture
BRD2: I use Facebook because interacting with people (whom I do not know in real life) makes me interested in what people unlike me are thinking
BRD3: I am willing to spend time to support Facebook activities with people (whom I do not know in real life)
BRD4: I use Facebook because interacting with people (whom I do not know in real life) makes me want to try new things
BRD5: I use Facebook because interacting with people (whom I do not know in real life) makes me feel like part of a larger community
Adult Attachment Questionnaire (AAQ)
AAQ1. I nd it relatively easy to get close to others
AAQ2. Im not very comfortable having to depend on other people
AAQ3. Im comfortable having others depend on me
AAQ4. I rarely worry about being abandoned by others
AAQ5. I dont like people getting too close to me
AAQ6. Im somewhat uncomfortable being too close to others
AAQ7. I nd it difcult to trust others completely
AAQ8. Im nervous whenever anyone gets too close to me
AAQ9. Others often want me to be more intimate than I feel comfortable being
AAQ10. Others often are reluctant to get as close to me as I would like
AAQ11. I often worry that my partner(s) dont really like me
AAQ12. I rarely worry about my partner(s) leaving me
AAQ13. I often want to merge completely with others, and this desire sometimes scares them away
AAQ14. Im condent others would never hurt me by suddenly ending our relationship
AAQ15. I usually want more closeness and intimacy with others than others want with me
AAQ16. The thought of being left by others rarely enters my mind
AAQ17. Im condent that my partner(s) like me just as much as I like them
1
The survey research company Embrain Co., Ltd. (http://www .embrain.com )
administered the surv ey from 22 to 29 May 2012.
1502 D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509
25 years of age (M= 22.5, SD = 1.46). Most respondents were soph-
omores (n = 50, 13.6%), juniors (n = 139, 37.8%) or seniors (n = 173,
47.0%) in 4-year university programs. Of the respondents, 82.1% of
the respondents reported using Facebook at least 30 min per day.
Another sizable group of respondents (n = 66, 17.9%), however , re-
ported using Facebook less than 30 min per day. The sample demo-
graphics are presente d in Table 2.
5.2. Descriptive statistics of the measurem ent instrument
As outlined in Table 3, the mean of the measure of bridging social
capital (M= 4.84, SD = 1.267) was rated highest among the four con-
struct measures, followed by bonding social capital (M= 4.00,
SD = 1.502). The means of avoidance attachment (M= 3.38,
SD = 1.068) and anxiety attachme nt (M= 3.51, SD = 1.104) were
rated below the scale midpoint, suggesting moderate levels of com-
fort with intimacy or unconcern with respect to rejection by others
among respondents, respectivel y. Cronbachs alpha coefcients cal-
culated for all subscales for each construct ranged from .797 to .924,
which is indicative of acceptable internal consistency (DeVellis,
2003).
5.3. Common method bias
The possibilit y of common method bias was a serious concern in
the sample data because all of the dependent and independen t vari-
ables were collected from the same respondents (Lee, 2009; Podsak-
off, MacKenzie, &Lee, 2003 ). The potential for common method bias
was assessed using Harmans single factor test (Podsakoff & Organ,
1986). All item measures were entered into a principal component
analysis, and the unrotated factor solution was examine d. Because
one single factor accounted for only 18.2% of the total variance and
because one single factor did not emerge, the threat of common
method bias was ruled out in the current study.
5.4. Exploratory factor analysis (EFA)
An EFA was performed to examine the factorial structure of the
measureme nt instrument.
2
The appropriaten ess of conducting fac-
tor analysis was conrmed by both the KaiserMeyerOlkin Mea-
sure of Sampling Adequacy (KMO = .842) and Bartlett s test of
sphericity (p = .000) (Hair, Black, Babin, & Anderson, 2010; Malhotra,
1999). Principal axis factoring with Promax rotation was used as an
extraction method (Fabrigar , Wegener, MacCal lum, & Strahan, 1999 ).
The EFA on measures extracted four componen ts with eigenvalues
exceeding 1.0. The acceptabl e threshold of 0.50 was used as the cri-
terion to retain an item (Hair et al., 2010; Perrot, Deloney, Hastings,
Savell, & Savidge, 2001; Wong & Keung, 2000 ). Four items, AAQ9,
AAQ10, AAQ13, and AAQ14, had factor loadings below the cutoff of
0.50 and were dropped from further considerati on. The four compo-
nents extracted togethe r explained 47.9% of the total varian ce. The
results of the EFA are shown in Table 4.
5.5. Testing assumptions of multiple regression
It is essential to examine if the data meets several underlying
statistical assumptions for multiple regressio n analysis (Ethington ,
Thomas, & Pike, 2002; Pedhazur, 1997 ). Firstly, the assumption of
homosced asticity was checked using the Levenes Test for Equality
of Variances (Snedecor & Cochran, 1989 ). The test ensured no sig-
nicant differences in the variance of the two dependent variables
of bonding- and bridging social capital across groups dened by all
independen t variables (p > 0.05 for all cases). Thus, the assumption
of homoscedas ticity was not violated (Lim & Loh, 1996 ). Secondly,
the assumption of normality was examine d through the calcula-
tion of skewness and kurtosis values of each variable. As seen in
Table 3, the skewness values for all variables ranged from .520
to .301 and the kurtosis values ranged from .634 to .689, which
are within the acceptable range of 1 to +1 for normality (George
& Mallery, 2003; Hair, Anderson, Tatham, & Black, 1998 ). Thus, the
violation of the normality assumption was not present in the sam-
ple data. Thirdly, the assumption of independen ce of residuals was
conrmed by the calculation of the DurbinWatson statistics for
the dependent variables of bonding social capital (=1.582) and
bridging social capital (=1.707), which are within the acceptable
range of 1.52.5 for independen ce (Johnson & Wichern, 2006 ).
Lastly, multi-col linearity was evaluated through the assessment
of zero-order correlations among selected measured constructs,
as calculated in Table 5. Harris and Hagger (2007) noted that mul-
ti-colline arity is not a serious issue if none of the correlation coef-
cients between variables exceeds .70. It is apparent that pair-wise
bivariate associations between the two independen t variables of
anxiety and avoidance were not highly correlated with each other,
nor with the depende nt variables of bonding- and bridging social
capital. Accordingl y, multi-collinear ity was dismissed from being
a major concern in the present study (Saxton & Dollinger, 2004 ).
To conclude, the sample data were judged to meet the criteria
for further analysis.
5.6. Hierarchi cal multiple regression analysis of bonding social capital
A hierarchical multiple regressio n analysis was performed with
bonding social capital as a dependent variable, with three sets of
independen t variables entered in each of three steps. Step 1 con-
trolled for gender, age, and the amount of Facebook use. Step 2
Table 2
Demographic proles from N = 368 respondents.
Demographic Frequency Percentage (%)
Gender
Male 171 46.5
Female 197 53.5
Age
19 3 .8
20 27 7.3
21 69 18.8
22 96 26.1
23 74 20.1
24 61 16.6
25 38 10.3
Grade level
Freshman 6 1.6
Sophomore 50 13.6
Junior 139 37.8
Senior 173 47.0
Facebook use per day
Less than .5 66 17.9
.51 h 140 38.0
12 h 73 19.8
23 h 52 14.1
34 h 13 3.5
45 h 13 3.5
Over 5 h 11 3.0
Total N = 386 100.0
2
The present study was particularly intere sted in identifying two orthogonal
attachment dimensions of anxiety and avoidance that underlie the AAQ (Simpson
et al., 1992 ). Thus two separate EFAs were performed on the measure of social capital
and the AAQ. That is, one EFA was performed on the ten items for social capital, and a
discrete EFA was performed on the seventeen item s for attachment style. The EFA
pertaining to social capital extracted two components with eigenvalues exceeding 1.0
(i.e., one set of bonding social capital and one set of bridging social capital). In the case
of the EFA regarding the AAQ, a two-component solution with eigenvalues exceeding
1.0 was extracted from the data set. The present study uses the result from a single
EFA conducted on the entire survey response set.
D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509 1503
controlled for anxiety and avoidance. Step 3 controlled for the
interaction of anxiety and avoidance. Gender was dummy-cod ed
as 1 for male and 2 for female. The two attachment dimensions
of anxiety and avoidance were centered by subtracting the sample
mean from each AAQ score prior to creating the interaction term
(Aiken & West, 1991; Cohen, Cohen, West, & Aiken, 2003 ). Table 6
presents the results of hierarchical multiple regressio n analysis on
bonding social capital. Hypotheses 1 and 2 were tested.
At step 1, demogra phic variables explained .3% of the variance
in bonding social capital. Inspectio n of the beta weights indicated
that age, gender, and the amount of Facebook use were not signif-
icant in any of the regression analyses. At step 2, inclusion of anx-
iety and avoidance attachment dimensions explained an additional
5.0% of the variance. Avoidance was the only signicant predictor
of bonding social capital. The negative beta coefcient indicates
that the lower the levels of avoidance attachment among respon-
dents, the more likely they will develop bonding social capital.
Avoidanc e remained a signicant independen t predictor of bond-
ing social capital in the nal model. Thus the results support
hypothes is 1. However , anxiety attachment did not emerge as a
signicant predictor of bonding social capital in the second and -
nal regressio n models. At step 3, addition of the interaction term to
the regressio n did not signicantly improve the model t, explain-
ing an additional .4% of variance. As a post hoc analysis, the pattern
of interaction between avoidance and anxiety was plotted by
means of simple-slope (Aiken & West, 1991 ), in which the effects
of anxiety on bonding social capital were assessed across low
(1 SD) and high (+1 SD) levels of avoidance in the total sample.
As shown in Fig. 1, the observed pattern appears to be consistent
with the prediction that anxiety impacts bonding social capital at
high levels of avoidance more than at lower levels of avoidance.
Table 3
Descriptive statist ics of the measurement instrument: mean, standard deviation, skewness, kurtosis, and Cronbachs alpha coefcient.
Construct Item Mean SD Skewness Kurtosis Cronbachs a
Bonding social capital (Mean = 4.00, SD = 1.502) BND1 3.74 1.490 .033 .634 .924
BND2 4.12 1.467 .208 .474
BND3 4.30 1.532 .302 .332
BND4 4.01 1.456 .056 .421
BND5 3.81 1.504 .045 .503
Bridging social capital (Mean = 4.84, SD = 1.267) BRD1 4.48 1.270 .256 .129 .879
BRD2 4.98 1.223 .520 .480
BRD3 4.81 1.257 .477 .255
BRD4 5.01 1.203 .414 .031
BRD5 4.90 1.314 .396 .050
Avoidance attachment (Mean = 3.38, SD = 1.068) AAQ1 3.34 1.123 .131 .415 .797
AAQ2 3.45 1.155 .076 .425
AAQ3 3.41 1.091 .022 .532
AAQ5 3.18 1.014 .301 .260
AAQ6 3.38 1.091 .112 .353
AAQ7 3.45 1.073 .168 .326
AAQ8 3.47 1.049 .208 .178
AAQ9 3.50 .907 .011 .689
Anxiety attachment (Mean = 3.51, SD = 1.104) AAQ4 3.53 1.075 .031 .156 .833
AAQ10 3.43 1.062 .081 .428
AAQ11 3.66 1.231 .164 .488
AAQ12 3.58 1.167 .058 .140
AAQ13 3.50 1.041 .116 .322
AAQ14 3.19 .918 .288 .316
AAQ15 3.37 1.193 .111 .338
AAQ16 3.46 1.102 .062 .297
AAQ17 3.47 1.064 .031 .398
Table 4
An Exploratory Factor Analysis (EFA): structure matrix.
Item Component
1 2 3 4
BND1 .753 .109 .055 .178
BND2 .873 .137 .112 .188
BND3 .868 .122 .095 .247
BND4 .916 .142 .067 .218
BND5 .806 .070 .071 .178
BRD1 .078 .768 .109 .167
BRD2 .100 .704 .059 .049
BRD3 .135 .782 .092 .116
BRD4 .126 .864 .066 .072
BRD5 .090 .739 .041 .086
AAQ1 .069 .089 .019 .632
AAQ2 .077 .055 .014 .653
AAQ3 .079 .019 .075 .588
AAQ5 .019 .009 .026 .627
AAQ6 .126 .074 .119 .571
AAQ7 .056 .016 .036 .574
AAQ8 .048 .148 .045 .600
AAQ9
a
.041 .079 .043 .332
AAQ4 .079 .057 .673 .011
AAQ10
a
.182 .082 .440 .075
AAQ11 .173 .089 .733 .019
AAQ12 .164 .038 .643 .084
AAQ13
a
.158 .119 .427 .205
AAQ14
a
.136 .154 .198 .274
AAQ15 .078 .054 .739 .039
AAQ16 .130 .049 .774 .007
AAQ17 .081 .034 .695 .032
Note. The exploratory factor analysis technique was principal axis factoring with
Promax rotation and Kaiser normalization. Rotation converged in ve iterations.
Bold values presents factor loadings above 0.5.
a
AAQ9, AAQ10, AAQ13, and AAQ14 were eliminated from further analyses.
Table 5
Zero-order correlations among measured constructs.
Constructs 1 2 3 4
1. Bonding social capital 1.000
2. Bridging social capital .123
*
1.000
3. Anxiety .074 .083 1.000
4. Avoidance .207
**
.114
*
.043 1.000
*
p < 0.05.
**
p < 0.01.
1504 D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509
However, the effect for anxiety was not signicant, neither at high
levels of avoidance (t
+1SD
(361) = 1.704, p = 0.089 > 0.05) nor at
the low levels of avoidance (t
1SD
(361) = 0.059, p = 0.953 > 0.05).
To gain a better understa nding of the interaction between anx-
iety and avoidance, the signicance regions and 95% condence
intervals for simple slope of anxiety was plotted across the stan-
dardized values of avoidance (Bauer & Curran, 2005; Gnambs &
Batinic, 2012; Preacher, Curran, & Bauer, 2006 ). In Fig. 2, regions
of signicance are dened by a lower bound of .3, indicating that
simple slope of anxiety on bonding social capital is signicantly
different from 0 for the standardized values of the avoidance
attachment greater than .3. In turn, the signicant differenc es were
not evident across the entire range of the moderating variable of
avoidance attachment. Overall, the results fail to support hypothe-
sis 2. The nal regression model explained 5.7% of the total vari-
ance in bonding social capital.
5.7. Hierarchi cal multiple regression on bridging social capital
Bridging social capital was regressed rst on demographic vari-
ables, second on anxiety and avoidance, and third on the interac-
tion of anxiety and avoidance. Again, the two attachment
dimensio ns of anxiety and avoidance were mean-cente red.
Hypothes es 3 and 4 were tested. Table 7 presents the results from
the hierarchical multiple regression analyses of bridging social
capital (see Fig. 3).
Age, gender, and the amount of Facebook use were entered at
step 1. These variables accounted for 2.7% of the total variance.
Interestin gly, the amount of Facebook use was found to be signi-
cantly predictive of bridging social capital in the model. Inspection
of the beta weights indicated that Facebook use remained a signif-
icant and independen t predictor of bridging social capital in the
subsequent two regression models. The addition of anxiety and
avoidance in step 2 signicantly increased the explained amount
of variance by 4.6%. Avoidance emerged as a signicant and nega-
tive predictor of bridging social capital, while the effect of anxiety
was not signicant for bridging social capital. Avoidance remained
a signicant predictor in the nal model. Thus the results support
hypothes is 3. The subsequent addition of the interaction term at
step 3 did not signicantly improve the model t. The interaction
of anxiety and avoidance with regard to bridging social capital is
plotted in Fig. 3. As predicted in hypothesis 4, the overall pattern
is consistent with the observation that avoidance seems to inu-
ence bridging social capital at low levels of anxiety more than at
high levels of anxiety. However, the effect for avoidance was not
signicant, neither at low levels of avoidance (t
1SD
(361) = 1.775,
p = 0.077 > 0.05) nor at the high levels of avoidance (t
+1SD
(361) = 0.103, p = 0.918 > 0.05).
To probe the interaction effect, the signicance regions and 95%
condence intervals for simple slope of avoidance was plotted
across the standardized values of anxiety. Fig. 4 demonstrates that
signicance regions are dened by an upper bound of .1. According
to the plot, simple slope of avoidance on bridging social capital is
signicantly different from 0 for the standardi zed values of the
anxiety attachment less than .1, which in turn, the signicant dif-
ferences were not observed over the full range of the moderator of
anxiety attachment. The results do not support hypothesis 4. The
explained total variance in the nal model was increased to 5.6%.
6. Discussion
Attachment theory has been instrumental in guiding research
about social developmen t througho ut the lives of individuals. The
Table 6
Hierarchical multiple regression of bonding social capital on age, gender, Fa cebook use, anxiety, avoidance, and interaction of anxiety and avoidanc e.
Predictor B Std. error b b
nal
Adj. R
2
DAdj. R
2
F
change
Step 1 .003 .003 F
(3, 364)
= .348
Age .027 .053 .030 .009
Gender .154 .153 .059 .055
Facebook use .006 .046 .006 .006
Step 2 .053 .050 F
(2, 362)
= 9.486
***
Anxiety .124 .077 .083 .083
Avoidance .375 .091 .211
***
.212
***
Step 3 .057 .004 F
(1, 361)
= 1.680
Anxiety Avoidance .130 .100 .066 .066
B: unstandardized coefcient; b: standardized coefcient; b
nal
: standardized coefcient in the nal model.
***
p < 0.001.
Low High
1.00
2.00
3.00
4.00
5.00
6.00
7.00
Anxiety
Low Avoidance
High Avoidance
B
o
n
d
i
n
g

S
o
c
i
a
l

C
a
p
i
t
a
l
Fig. 1. Simple slopes for the interaction between anxiety and avoidance in bonding
social capital.
-1.0 -0.5 0.0 0.5 1.0
-0.4
-0.2
0.0
0.2
Avoidance
S
i
m
p
l
e

S
l
o
p
e

o
f

A
n
x
i
e
t
y
o
n

B
o
n
d
i
n
g

S
o
c
i
a
l

C
a
p
i
t
a
l
Region of
significance
Fig. 2. Regions of signicance and condence intervals for simple slope of anxiety
on bonding social capital as a function of Z-standardized avoidance. Note: Gray lines
reect the lower and upper bounds of the 95% condence interval; dashed vertical
line represents regions of signicance.
D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509 1505
primary contribution of the present study is to adapt and apply
attachment theory to understa nding the developmen t of interper-
sonal relationships in the context of SNS. Given the popularity of
SNS, a body of research has been devoted to identifying the moti-
vational factors that lead individuals to engage in SNS (Kalpidou,
Costin, & Morris, 2011 ; Kim, 2011; Lin & Lu, 2011; Raacke &
Bonds-Raac ke, 2008; Wilson, Fornasier, & White, 2010 ). The ap-
proach of much of the existing work is characteri zed by a stream
of research that attempts to determine and explain the system
features and options most valued by users. By contrast, there is a
limited set of empirical studies that account for the role of individ-
ual differences in our natural propensity for intimacy in the use
of SNS (Amichai-Ham burger & Vinitzky, 2010; Correa, Hinsley,
& de Ziga, 2010 ). The primary focus of this small set of studies
is the role of individual differences associated with attachment
styles (Buote, Wood, & Pratt, 2009; Ye, 2007 ), which is a particular
concern given that individual differences in attachment styles
may lead users to form a range of different patterns of interper-
sonal relationship s. Conseque ntly, these different patterns may
different ially impact the ways in which users approach SNS.
The motive behind the utilization of the term social capital in
the present study is that social capital may be conceptu alized as
an internalizat ion of the set of outcome variables that individuals
seek to create via interpersonal relationships. The term social cap-
ital is further classied into bonding social capital and bridging so-
cial capital to reect the different patterns of affectional ties that
develop among individua ls through participatio n in a SNS. It was
hypothes ized that the attachment styles of individuals would be
evident in the developmen t of these two types of social capital in
a highly predictable manner.
Consisten t with the prediction, individua l differences in avoid-
ance attachment were found to play a predictive role in the devel-
opment of social capital. Specically, higher or lower levels of
avoidance attachment are signicantly related to lesser or greater
developmen t of bonding and bridging forms of social capital,
respectivel y. The ndings suggest that the notion of avoidance is
consisten t with a determinist ic view of the negative internal work-
ing models of individuals about others. That is, those individuals
with high levels of avoidance attachme nt are likely to be uncom-
fortable with intimacy. This study nds the effects of anxiety
attachme nt to be insignicant for both the developmen t of bond-
ing social capital and bridging social capital. Taken together, the
ndings suggest that avoidance of intimacy may be the most inu-
ential motivating factor in the pursuit of the fulllment of interper-
sonal goals, regardless of any relationship expectati ons in the
context of SNS. However, it can be argued that a computer -medi-
ated communication channel may appeal to individuals high in
avoidance attachment as they seek to meet their interpersonal
needs. This is reasonabl e because individuals high in avoidance
attachme nt may enjoy more open communi cation over the Inter-
net or within SNS environments due to the lack of non-verbal or
supercial physical cues (Siegel, 2007; Ye, 2007 ). The present
study, however, was not designed to examine whether the use of
the Internet or SNS compensates for the shortcomin gs of the ten-
dencies of avoidants to avoid intimacy. Interestin gly, an investiga -
tion of beta coefcients in the two hierarchical multiple
regressio ns revealed that the effect of avoidance attachment was
greater for bonding social capital than for bridging social capital.
It seems likely that the extent to which developmen t of bonding
social capital is subject to barriers to closeness is greater than
the extent of this phenomeno n on the developmen t of bridging so-
cial capital. This result implies that the development of bonding so-
Table 7
Hierarchical multiple regression of bridging social capital on age, gender, Facebook use, anxiety, avoidanc e, and interact ion of anxiety and avoidance.
Predictor B Std. error b b
nal
Adj. R
2
DAdj. R
2
F
change
Step 1 .027 .027 F
(3, 364)
= 3.358
*
Age .041 .041 .058 .052
Gender .161 .119 .078 .078
Facebook use .099 .036 .144
**
.137
**
Step 2 .046 .019 F
(2, 362)
= 3.664
*
Anxiety .098 .061 .083 .083
Avoidance .162 .072 .116
*
.115
*
Step 3 .056 .010 F
(1, 361)
= 3.613
Anxiety Avoidance .150 .079 .097 .097
B: unstandardized coefcient; b: standardized coefcient; b
nal
: standardized coefcient in the nal model.
*
p < 0.05.
**
p < 0.01.
Low High
1.00
2.00
3.00
4.00
5.00
6.00
7.00
Avoidance
Low Anxiety
High Anxiety
B
r
i
d
g
i
n
g

S
o
c
i
a
l

C
a
p
i
t
a
l
Fig. 3. Simple slopes for the interaction between anxiety and avoidance in bridging
social capital.
-1.0 -0.5 0.0 0.5 1.0
-0.4
-0.2
0.0
0.2
Anxiety
S
i
m
p
l
e

S
l
o
p
e

o
f

A
v
o
i
d
a
n
c
e
o
n

B
r
i
d
g
i
n
g

S
o
c
i
a
l

C
a
p
i
t
a
l
Region of significance
Fig. 4. Regions of signicance and condence intervals for simple slope of
avoidance on bridging social capital as a function of Z-standardized anxiety. Note:
Gray lines reect the lower and upper bounds of the 95% condence interval;
dashed vertical line represents regions of signicance.
1506 D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509
cial capital differs in part from the developmen t of bridging social
capital in that the former relies more on trust-based relationship s.
Importantly, both bonding social capital and bridging social
capital among respondents appear to be greatest under conditions
of low anxiety attachment coupled with low avoidance attach-
ment. Even though differences in the means of measures among
the four groups of attachment styles did not reach statistical signif-
icance, the observed pattern is consistent with the notion that
[. . .] secure people would show a rich interpers onal life, which in-
cludes both attachme nt and afliation interactions (Mikulincer &
Selinger, 2001, p. 85 ). Secure individuals may be more likely than
insecure individua ls to develop bonding social capital from social
ties, which is a phenomeno n possibly motivated by the condence
of secure people in the availability of emotional support. In addi-
tion, secure people may be able to exibly move to seek out the
types of shallow relationships (Amichai- Hamburger, 2002, p. 7)
that commonly characterize SNS. This type of social behavior cor-
responds to bridging social capital, primarily because of the ten-
dency to maintain distance from others (Dewitte & Houwer,
2008). The ndings are partly in line with work by Florian, Mikul-
incer, and Bucholtz (1995), in which the authors found that secure
individuals report actively reaching out for emotional and instru-
mental support more than insecure individua ls do.
The present study found no support for the predicted interac-
tion effects between anxiety attachment and avoidance attach-
ment in the measures of bonding and bridging forms of social
capital. Regardless, it is interesting to note that the two interactio n
patterns observed were consisten t with the predictio ns. The
attachment styles of respondents were measured on the basis of
continuously distributed dimensions of attachment, but not on dis-
crete typological or categorical measureme nts (i.e., Relationship
Questionnai re, Bartholomew & Horowitz, 1991 ). With reference
to the work of Fraley and Waller (1998), a study by Meredith
et al. (2008) noted that [ . . .] dimensional assessment methods
might afford a more accurate representation of the adult attach-
ment construct than that provided by the categorical [measures]
(p. 411). However, Fraley and Aron (2004) indicated that [ . . .]
interaction effects involving a continuous variable have very low
power (p. 65). According to studies by Cohen (1992) and Frazier,
Tix, and Barron (2004), R
2
changes less than .02 reect a small ef-
fect of interaction in regression analysis. Based on the results of the
hierarchical regressio n analyses performed in this study, failures to
reach statistical signicance may be due to the possibility that sta-
bility of the attachment classication was not obtained in the pres-
ent study.
Lastly, the mean of the measure of bridging social capital was
found to be greater than the mean of the measure of bonding social
capital. This result is contradictor y to the argument that social net-
working sites are predominantly used for purposes of satisfying
individuals social-em otional needs (Rau, Gao, & Ding, 2008 ). Re-
call, however , that the emphasis of the present work is the devel-
opment of new interpersonal relationships within the context of
SNS, as opposed to the maintenanc e of established ofine relation-
ships. Additionally, there appears to be a signicant difference in
the amount of Facebook use between users that develop bonding
social capital versus bridging social capital in the context of SNS.
Overall, this study implies that higher levels of Facebook use
should be the cost paid for contributi ng to the developmen t of a
wide range of access to broad social ties, only not so in the case
of the highly selective nature of bonding social capital.
7. Conclusion
One of the main ndings of the present study is that attachme nt
theory is a useful theoretical framework for predicting the devel-
opment of different interpers onal relationships among individuals
in the ever-expanding realm of SNS. There were three notable nd-
ings in this study. First, avoidance attachment is the most salient
factor leading to the developmen t of trust-based strong ties and
causal afliation from more socially distant people. Second, secure
individua ls seem to display greater relationa l competence than
anxiousambivalents, dismissing-avo idants or fearful-avoidan ts
(Bartholomew, 1993 ). Third and nally, the developmen t of bridg-
ing social capital increases as the level of SNS use increases.
Acknowled gements
This work was supported by the National Research Foundation
of Korea Grant funded by the Korean Governmen t (NRF-2010-
B00171). This work was also supported by the Social Science Korea
(NRF-2011-330-B00225 ) project by the National Research Founda-
tion of Korean government .
References
Adler, P. S., & Kwon, S. W. (2002). Social capital: Prospects for a new concept.
Academy of Management Review, 27 (1), 1740.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.
Ainsworth, M. D. S. (1989). Attachment beyond infancy. American Psychologist,
44(4), 709716.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of
attachment: A psychological study of the strange situation . Hillsdale, NJ: Erlbaum.
Amichai-Hamburger, Y. (2002). Internet and personality. Computers in Human
Behavior, 18 (1), 110.
Amichai-Hamburger, Y., & Vinitzky, G. (2010). Social network use and personality.
Computers in Human Behavior, 26 (6), 12891295.
Armsden, G. C., & Greenberg, M. T. (1987). The inventory of parent and peer
attachment: Individual differences and their relationship to psychological well-
being in adolescence. Journal of Youth and Adolescence, 16 (5), 427454.
Bakker, W., Van Oudenhoven, J. P., & Van Der Zee, K. I. (2004). Attachment styles,
personality, and Dutch emigrants intercultural adjustment. European Journal of
Personality, 18 (5), 387404.
Bartholomew, K. (1993). From childhood to adult relationships: Attachment theory
and research. In S. Duck (Ed.), Understanding relationship processes . Learning
about relationships (Vol. 2, pp. 3062). Newbury Park, CA: Sage.
Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults:
A test of a four-category model. Journal of Personality and Social Psychology,
61(2), 226244.
Bauer, D. J., & Curran, P. J. (2005). Probing interactions in xed and multilevel
regression: Inferential and graphical techniques. Multivariate Behavioral
Research, 40 (3), 373400.
Bernier, A., Larose, S., & Soucy, N. (2005). Academic mentoring in college: The
interactive role of students and mentors interpersonal dispositions. Research in
Higher Education, 46 (1), 2951.
Bippus, A. M., & Rollin, E. (2003). Attachment style difference in relational
maintenance and conict behaviors: Friends perceptions. Communication
Reports, 16 (2), 113123.
Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of
theory of research for the sociology of education (pp. 241258). New York:
Greenwood Press.
Bowlby, J. (1969). Attachment and loss (Vol. 1). Attachment. New York, NY: Basic
Books.
Bowlby, J. (1973). Attachment and loss (Vol. 2). Separation: Anxiety and anger . New
York, NY: Basic Books.
Bowlby, J. (1980). Attachment and loss (Vol. 3). Loss: Sadness and depression . New
York, NY: Basic Books.
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory . London:
Routledge.
Boyd, D. M., & Ellison, N. B. (2007). Social network sites: Denition, history, and
scholarship. Journal of Computer-Mediated Communication, 13 (1). <http://
jcmc.indiana.edu/vol13/issue1/boyd.ellison.html> (Retrieved 08.05.12).
Bretherton, I., Ridgeway, D., & Cassidy, J. (1990). Assessing internal working models
of attachment relationships: An attachment story completion task for 3-year-
olds. In M. T. Greeberg, K. Cicchetti, & E. M. Cummings (Eds.), Attachment in the
preschool years: Theory, research, and intervention (pp. 273308). Chicago:
University of Chicago Press.
Buote, V. M., Wood, E., & Pratt, M. (2009). Exploring similarities and differences
between online and ofine friendships: The role of attachment style. Computers
in Human Behavior, 25 (2), 560567.
Carvallo, M., & Gabriel, S. (2006). No man is an island: The need to belong and
dismissing avoidant attachment style. Personality and Social Psychology Bulletin,
32(5), 697709.
D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509 1507
Cassidy, J. (1999). The nature of the childs ties. In J. Cassidy & P. R. Shaver (Eds.),
Handbook of attachment: Theory, research, and clinical applications (pp. 320).
New York: Guilford Press.
Cassidy, J., Ziv, Y., Mehta, T. G., & Feeney, B. C. (2003). Feedback seeking in children
and adolescents: Associations with self-perceptions, attachment
representations, and depression. Child Development, 74 (2), 612628.
Cheung, C.-K., & Kam, P. K. (2010). Bonding and bridging social capital development
by social workers. Journal of Social Service Research, 36 (5), 402413.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112 (1), 155159.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/
correlation analysis for the behavioral sciences . Mahwah, NJ: Lawrence Erlbaum
Association.
Coleman, J. S. (1988). Social capital in the creation of human capital. American
Journal of Sociology, 94 , 95120.
Collins, N. L., & Feeney, B. C. (2000). A safe haven: An attachment theory perspective
on support seeking and caregiving in intimate relationships. Journal of
Personality and Social Psychology, 78 (6), 10531073.
Correa, T., Hinsley, A., & de Ziga, H. (2010). Who interacts on the Web? The
intersection of users personality and social media use. Computers in Human
Behavior, 26 (2), 247253.
DeVellis, R. F. (2003). Scale development: Theory and applications . Newbury Park, CA:
Sage Publications.
Dewitte, M., & Houwer, J. D. (2008). Proximity and distance goals in adult
attachment. European Journal of Personality, 22(8), 675694.
Dindia, K., & Canary, D. J. (1993). Denitions and theoretical perspectives on
relational maintenance. Journal of Social and Personal Relationships, 10 , 163173.
Donath, J., & Boyd, D. (2004). Public displays of connection. BT Technology Journal,
22(4), 7182.
Egan, K. G., & Moreno, M. A. (2011). Prevalence of stress references on college
freshmen Facebook proles. Computers, Informatics, Nursing, 29 (10), 586592.
Ellison, N. B., Steineld, C., & Lampe, C. (2007). The benets of Facebook friends:
Social capital and college students use of online social network sites. Journal of
Computer-Mediated Communication, 12 (4). <http://jcmc.indiana.edu/vol12/
issue4/ellison.html> (Retrieved 02.02.12).
Ethington, C. A., Thomas, S. L., & Pike, G. R. (2002). Back to the basics: Regression as
it should be. In J. C. Smart (Ed.), Higher education: Handbook of theory and
research (pp. 263294). New York: Agathon Press.
Fabrigar, L. R., Wegener, D. T., MacCallum, R. C., & Strahan, E. J. (1999). Evaluating
the use of exploratory factor analysis in psychological research. Psychological
Methods, 4(3), 272299.
Feeney, J. A., & Noller, P. (1990). Attachment style as a predictor of adult romantic
relationships. Journal of Personality and Social Psychology, 58 (2), 281291.
Flaherty, S. C., & Sadler, L. S. (2011). A review of attachment theory in the context of
adolescent parenting. Journal of Pediatric Health Care, 25 (2), 114121.
Florian, V., Mikulincer, M., & Bucholtz, I. (1995). Effects of adult attachment style on
the perception and search for social support. The Journal of Psychology, 129 (6),
665676.
Florida, R. (2002). The rise of the creative class, and how its transforming work,
community, and everyday life . New York: Basic Books.
Fraley, B., & Aron, A. (2004). The effect of a shared humorous experience on
closeness in initial encounters. Personal Relationships, 11(1), 6178.
Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns: A test of the
typological model. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory and
close relationships (pp. 77114). New York: The Guildford Press.
Frazier, P. A., Tix, A. P., & Barron, K. E. (2004). Testing moderator and mediator
effects in counseling psychology research. Journal of Counseling Psychology,
51(1), 115134.
George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and
reference, 11.0 update (4th ed). Boston: Allyn & Bacon.
Gnambs, T., & Batinic, B. (2012). The roots of interpersonal inuence: A mediated
moderation model for knowledge and traits as predictors of opinion leadership.
Applied Psychology . http://dx.doi.org/10.1111/j.1464-0597.2012.00497.x.
Granovetter, M. S. (1982). The strength of weak ties: A network theory revisited. In
P. V. Mardsen & N. Lin (Eds.), Social structure and network analysis (pp. 105130).
Thousand Oaks: Sage Publications.
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data
analysis (5th ed.). Englewood Cliffs, NJ: Prentice Hall.
Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis.
A global perspective (7th ed.). Upper Saddle River, NJ: Pearson Prentice Hall.
Harris, J., & Hagger, M. S. (2007). Do basic psychological needs moderate
relationships within the theory of planned behavior? Journal of Applied
Biobehavioral Research, 12 (1), 4364.
Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as an attachment
process. Journal of Personality and Social Psychology, 52 (3), 511524.
Johnson, R. A., & Wichern, D. W. (2006). Applied multivariate statistical analysis (5th
ed.). Englewood Cliffs, NJ: Prentice Hall.
Kalpidou, M., Costin, D., & Morris, J. (2011). The relationship between Facebook and
the well-being of undergraduate college students. Cyberpsychology, Behavior,
and Social Networking, 14 (4), 183189.
Keefer, L. A., Landau, M. J., Rothschild, Z. K., & Sullivan, D. (2012). Attachment to
objects as compensation for close others perceived unreliability. Journal of
Experimental Social Psychology . http://dx.doi.org/10.1016/j.jesp. 2012.02.007.
Kim, B. (2011). Understanding antecedents of continuance intention in social-
networking services. Cyberpsychology, Behavior, and Social Networking, 14 (4),
199205.
Klohnen, E. C., & John, O. P. (1998). Working models of attachment: A theory based
prototype approach. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory
and close relationships (pp. 115140). New York: Guilford.
Ko, H.-C., & Kuo, F.-Y. (2009). Can blogging enhance subjective well-being through
self-disclosure? CyberPsychology & Behavior, 12 , 7579.
Lampe, C., Ellison, N., & Steineld, C. (2007). A Face(book) in the crowd: Social
searching vs. social browsing. In Proceedings of the SIGCHI conference on human
factors in computing systems (pp. 434444). New York: ACM Press.
Landen, S. M., & Wang, C.-C. (2010). Adult attachment, work cohesion, coping, and
psychological well-being of reghters. Counseling Psychology Quarterly, 23 (2),
143162.
Lee, M.-C. (2009). Predicting and explaining the adoption of online trading: An
empirical study in Taiwan. Decision Support Systems, 47 (2), 133142.
Lim, T. S., & Loh, W. Y. (1996). A comparison of tests of equality of variances.
Computational Statistics & Data Analysis, 22(3), 287301.
Lin, N. (2008). A network theory of social capital. In D. Castiglione, J. W. Van Deth, &
G. Wolleb (Eds.), The handbook of social capital (pp. 5069). New York: Oxford
University Press.
Lin, K.-Y., & Lu, H.-P. (2011). Intention to continue using Facebook fan pages from
the perspective of social capital theory. Cyberpsychology, Behavior, and Social
Networking, 14 (10), 565570.
Lopez, F. G. (2001). Adult attachment orientations, self-other boundary regulation,
and splitting tendencies in a college sample. Journal of Counseling Psychology,
48(4), 440446.
Main, M. (2000). The organized categories of infant, child, and adult attachment:
Flexible vs. inexible attention under attachment-related stress. Journal of the
American Psychoanalytic Association, 48 (4), 10551096.
Main, M., Kaplan, N., & Cassidy, J. (1985). Security in infancy, childhood, and
adulthood: A move to the level of representation. Monographs of the Society for
Research in Child Development, 50 (12), 66104.
Main, M., & Solomon, J. (1986). Discovery of an insecure-disorganized/disoriented
attachment pattern. In T. B. Brazelton & M. W. Yogman (Eds.), Affective
development in infancy (pp. 95124). Westport, CT: Ablex Publishing.
Malhotra, N. K. (1999). Marketing research: An applied orientation . Upper Saddle
River, NJ: Prentice Hall.
Mawson, A. R. (2007). Mass panic and social attachment: The dynamics of human
behavior. Burlington, VT: Ashgate Publishing Co..
Meredith, P., Ownsworth, T., & Strong, J. (2008). A review of the evidence linking
adult attachment theory and chronic pain: Presenting a conceptual model.
Clinical Psychology Review, 28 (3), 407429.
Mikail, S. F., Henderson, P. R., & Tasca, G. A. (1994). An interpersonally based model
of chronic pain: An application of attachment theory. Clinical Psychology Review,
14(1), 116.
Mikulincer, M., & Selinger, M. (2001). The interplay between attachment and
afliation systems in adolescents same-sex friendships: The role of attachment
style. Journal of Social and Personal Relationships, 18 (1), 81106.
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics,
and change . New York: Guilford Press.
Mikulincer, M., & Shaver, P. R. (2009). An attachment and behavioral systems
perspective on social support. Journal of Social and Personal Relationships, 26 (1),
719.
Mikulincer, M., Shaver, P. R., & Horesh, N. (2006). Attachment bases of emotion
regulation and posttraumatic adjustment. In D. K. Snyder, J. A. Simpson, & J. N.
Hughes (Eds.), Emotion regulation in families: Pathways to dysfunction and health
(pp. 7799). Washington, DC: American Psychological Association.
Moore, S., & Leung, C. (2002). Young peoples romantic attachment styles and their
associations with well-being. Journal of Adolescence, 25 (2), 243255.
OConnor, E., & McCartney, K. (2007). Attachment and cognitive skills: An
investigation of mediating mechanisms. Journal of Applied Developmental
Psychology, 28 (56), 458476.
Patterson, D. L. (1997). Adolescent mothering: Childgrandmother attachment.
Journal of Pediatric Nursing, 12 (4), 228237.
Paxton, P. (1999). Is social capital declining in the United States? A multiple
indicator assessment. American Journal of Sociology, 105 (1), 88127.
Pedhazur, E. J. (1997). Multiple regression in behavioral research: Explanation and
prediction. New York: Holt, Rinehart, & Winston.
Peluchette, J., & Karl, K. (2008). Social networking proles: An examination of
student attitudes regarding use and appropriateness of content.
CyberPsychology & Behavior, 11(1), 9597.
Perrot, L. J., Deloney, L. A., Hastings, J. K., Savell, S., & Savidge, M. (2001). Measuring
student motivation in health professions colleges. Advances in Health Sciences
Education, 6(3), 193203.
Pfeil, U., Arjan, R., & Zaphiris, P. (2009). Age differences in online social networking
A study of user proles and the social capital divide among teenagers and older
users in MySpace. Computers in Human Behavior, 25 (3), 643654.
Podsakoff, P. M., MacKenzie, S. B., & Lee, J.-Y. (2003). Common method biases in
behavioral research: A critical review of the literature and recommended
remedies. Journal of Applied Psychology, 88(5), 879903.
Podsakoff, P., & Organ, D. (1986). Self-reports in organizational research: Problems
and prospects. Journal of Management, 11(2), 129139.
Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing
interactions in multiple linear regression, multilevel modeling, and latent curve
analysis. Journal of Educational and Behavioral Statistics, 31 (4), 437448.
Putnam, R. D. (1995). Bowling alone: Americas declining social capital. Journal of
Democracy, 6(1), 6478.
1508 D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community .
New York: London, Simon & Schuster.
Raacke, J., & Bonds-Raacke, J. (2008). MySpace and Facebook: Applying the uses and
gratications theory to exploring friend-networking sites. CyberPsychology &
Behavior, 11(2), 169174.
Rau, P.-L., Gao, Q., & Ding, Y. (2008). Relationship between the level of intimacy and
lurking in online social network services. Computers in Human Behavior, 24 (6),
27572770.
Rholes, W. S., Simpson, J. A., Tran, S., Martin, M., III, & Friedman, M. (2007).
Attachment and information seeking in romantic relationships. Personality and
Social Psychology Bulletin, 33(3), 422438.
Rosenthal, N. L., & Kobak, R. (2010). Assessing adolescents attachment hierarchies:
Differences across developmental periods. Journal of Research on Adolescence,
20(3), 678706.
Saxton, T., & Dollinger, M. (2004). Target reputation and appropriability: Picking
and deploying resources in acquisitions. Journal of Management, 30 (1),
123147.
Schuller, T., Baron, S., & Field, J. (2000). Social capital: A review and critique. In S.
Baron, J. Field, & T. Schuller (Eds.), Social capital: Critical perspectives (pp. 138).
New York, Oxford: Oxford University Press.
Seibert, S. E., Kraimer, M. L., & Liden, R. C. (2001). A social capital theory of career
success. Academy of Management Journal, 44(2), 219237.
Senchak, M., & Leonard, K. E. (1992). Attachment styles and marital adjustment
among newlywed couples. Journal of Social and Personal Relationships, 9(1),
5164.
Siegel, D. (2007). The mindful brain . New York: Mind Your Brain, Inc..
Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support seeking and support
giving within couples in anxiety-provoking situations: The role of attachment
styles. Journal of Personality and Social Psychology, 62 (3), 434446.
Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conict in close relationships: An
attachment perspective. Journal of Personality & Social Psychology, 71 (5),
899914.
Snedecor, G. W., & Cochran, W. G. (1989). Statistical methods (8th ed.). Ames, IA:
Iowa State University Press.
Sroufe, L. A., & Waters, W. (1977). Attachment as an organizational construct. Child
Development, 48 (5), 11841199.
Steineld, C., Ellison, N. B., & Lampe, C. (2008). Social capital, self-esteem, and use of
online social network sites: A longitudinal analysis. Journal of Applied
Developmental Psychology, 29 (6), 434445.
Subrahmanyam, K., Reich, S. M., Waechter, N., & Espinoza, G. (2008). Online and
ofine social networks: Use of social networking sites by emerging adults.
Journal of Applied Developmental Psychology, 29 (6), 420433.
Ulhi, J. P. (2005). The social dimensions of entrepreneurship. Technovation, 25 (8),
939946.
Utz, S., & Beukeboom, C. J. (2011). The role of social network sites in romantic
relationships: Effects on jealousy and relationship happiness. Journal of
Computer-Mediated Communication, 16 (4), 511527.
Wang, S., Noe, R. A., Wang, Z.-M., & Greenberger, D. B. (2009). What affects
willingness to mentor in the future? An investigation of attachment
styles and mentoring experiences. Journal of Vocational Behavior, 74(3),
245256.
Wasko, M. M., & Faraj, S. (2005). Why should I share? Examining social capital and
knowledge contribution in electronic networks of practice. MIS Quarterly, 29 (1),
3557.
Welch, R. D., & Houser, M. E. (2010). Extending the four-category model of adult
attachment: An interpersonal model of friendship attachment. Journal of Social
and Personal Relationships, 27 (3), 351366.
Wilkinson, R. B. (2004). The role of parental and peer attachment in the
psychological health and self-esteem of adolescents. Journal of Youth and
Adolescence, 33(6), 479493.
Williams, D. (2006). On and off the Net: Scales for social capital in an online era.
Journal of Computer-Mediated Communication, 11(2), 593628.
Wilson, K., Fornasier, S., & White, K. M. (2010). Psychological predictors of young
adults use of social networking sites. Cyberpsychology, Behavior, and Social
Networking, 13 (2), 173177.
Wong, S., & Keung, C. (2000). Tourists perceptions of hotel frontline employees
questionable job-related behaviour. Tourism Management, 21 (2), 121134.
Yang, S.-C., & Farn, C.-K. (2009). Social capital, behavioural control, and tacit
knowledge sharing A multi-informant design. International Journal of
Information Management, 29 (3), 210218.
Ye, J. (2007). Attachment style differences in online relationship involvement: An
examination of interaction characteristics and relationship satisfaction.
CyberPsychology & Behavior, 10 (4), 605607.
Yonder, C., & Stutzman, F. (2011). Identifying social capital in the Facebook
interface. In Proceedings of the ACM conference on human factors in computing
systems, Vancouver, BC, Canada.
Yuan, Y. C., & Gay, G. (2006). Homophily of network ties and bonding and bridging
social capital in computer-mediated distributed teams. Journal of Computer-
Mediated Communication, 11(4). <http://jcmc.indiana.edu/vol11/issue4/
yuan.html> (Retrieved 03.02.12).
Zilber, A., Goldstein, A., & Mikulincer, M. (2007). Adult attachment orientations and
the processing of emotional pictures ERP correlates. Personality and Individual
Differences, 43 (7), 18981907.
D.Y. Lee / Computers in Human Behavior 29 (2013) 14991509 1509