Computers in Human Behavior

Volume 30, January 2014, Pages 69–78

How does online social networking enhance

life satisfaction? The relationships among
online supportive interaction, affect,
perceived social support, sense of community,
and life satisfaction
http://www.sciencedirect.com.ux4ll8xu6v.use
accesscontrol.com/science/article/pii/S074756
3213002926
 Hyun Jung Oha, ,
,
 Elif Ozkayab, ,
 Robert LaRoseb,

 a
Michigan State University, 309 Communication Arts & Sciences, East Lansing, MI
48824, USA
 b
Michigan State University, 409 Communication Arts & Sciences, East Lansing, MI
48824, USA
 Abstract

The purpose of this study is to examine whether supportive interactions on social networking
sites mediate the influence of SNS use and the number of SNS friends on perceived social
support, affect, sense of community, and life satisfaction. Employing momentary sampling, the
current study also looked at the relationship between supportive interaction and immediate affect
after the interaction over a period of 5 days. An analysis of 339 adult participants revealed a
positive relationship between supportive interaction and positive affect after the interaction. A
path model revealed positive associations among the number of SNS friends, supportive
interactions, affect, perceived social support, sense of community, and life satisfaction.
Implications for the research of online social networking and social support are discussed.

Keywords

 Social networking sites;

 Social support;
 Affect;
 Sense of community;
 Life satisfaction;
 Momentary sampling

1. Introduction

Social networking today is one of the most popular activities on the Internet. A recent report
found that visiting social networking sites (SNSs) is the most frequent online activity of Internet
users (comScore., 2011). Over 90% of adult users in the US have visited social media (Simmons,
2011). On average, social networking accounts for one of every 6 minutes that people spend
online (comScore, 2011).

Considerable scholarly attention has focused on the psychological outcomes of online social
networking, but the results have been inconsistent. Although researchers have studied various
psychological outcomes (Burke et al., 2010, Ellison et al., 2007, Kim and Lee, 2011, Steinfield et
al., 2008, Valenzuela et al., 2009 and Valkenburg et al., 2006), major indicators of online social
networking (i.e., number of SNS friend, the amount of social networking) often failed to predict
positive outcomes (Kalpidou et al., 2011, Kim and Lee, 2011, Klingensmith, 2010, Valenzuela et
al., 2009 and Vitak et al., 2011). Inconsistent findings suggest two possibilities: First, prior
online social networking studies have focused on varying predictors of psychological outcomes,
some of which may be more valid than others. Second, there may be a third variable that either
mediates or moderates the relationships between the predictors and the outcomes.

The current study explores one possible mechanism through which peoples’ online social
networking leads to psychological outcomes. Specifically, this study proposes that the amount of
supportive interaction mediates the relationship between general SNS use and users’
psychological states. The traditional social network literature also highlights the importance of
the quality of interaction among members in a network (Pinquart & Sörensen, 2000) and the
perception of available social support from other members (Young, 2006). Pursuing this
argument, the current study proposes several mediating variables that include supportive
interaction, affect, and perceived social support.
The present study focuses on two psychological outcomes, sense of community and life
satisfaction, as they appear frequently in online communication and social network literature
(Kim and Lee, 2011, Klingensmith, 2010, Köbler et al., 2010 and Young, 2006). Using a
momentary sampling technique, which allows an examination of affect immediately after a
behavior, this study first examines the direct relationship between the amount of supportive
interaction and the positive affect following that interaction for five days. Secondly, the study
uses a path model to explore the possible mechanism through which online social networking
can lead to a greater sense of community and life satisfaction. By doing both, we expect to
contribute to a better understanding of the positive outcomes of online social networking and
clarify the issues related to mixed findings in the existing SNS literature.

2. Theoretical background
2.1. Online social networking

Social network sites (SNSs) are defined as web-based applications that allow their users to
construct a profile that other users can see and also list connections with other users (Ellison et
al., 2007). Although the goal of an online social network remains the same as an offline social
network, i.e., connecting people and sharing resources in the network, online networking further
benefits users with its unique features related to computer-mediated communication. With the
growing use of SNSs, certain studies have found that there are positive psychological outcomes
from engaging in social networking on various social networking sites (Burke et al., 2010,
Ellison et al., 2007, Kim and Lee, 2011, Steinfield et al., 2008, Valenzuela et al.,
2009 and Valkenburg et al., 2006). The number of SNS friends and the amount of SNS use are
currently the two most studied indicators that represent the amount of online social networking.

The number of friends, the main source of social support (Boyd, 2006), is one of the widely
studied constructs in the online social networking literature (Ellison et al., 2007 and Kim and
Lee, 2011). These studies, however, reported inconsistent results when they treated the number
of SNS friends as a direct predictor of psychological outcomes, thus suggesting the possibility
that either mediating or moderating variables do exist (Kalpidou et al., 2011 and Kim and Lee,
2011). For example, Kim and Lee (2011) found an inverted U relationship between the number
of SNS friends and perceived social support among college students, where the number of total
SNS friends was positively associated with perceived social support until the number of friends
reached a certain point.

Facebook intensity, developed to measure the amount of Facebook use (Ellison et al., 2007), has
also been frequently measured in the SNS literature. It combines the number of Facebook
friends, hours or frequency of using Facebook, and the emotional attachment to Facebook, as
well as indictors of habit strength and compulsive use (LaRose, Wohn, Ellison, & Steinfield,
2011). Studies using this construct often suffered from its weak explanatory power when it was
included in a regression model (Valenzuela et al., 2009). Facebook intensity showed either
negative (Klingensmith, 2010) or non-significant (Vitak et al., 2011) relationships with other
indicators of well-being.

On the other hand, clinical studies have reported negative consequences associated with using
Facebook (e.g., O’Keeffe and Clarke-Pearson, 2011 and Pantic et al., 2012). A terminology
“Facebook Depression” was developed by O’Keeffe and Clarke-Pearson (2011) in their clinical
report to highlight a phenomenon that excessive use of Facebook leads to depression among
certain groups of young people. Although this report has been criticized due to its exaggerated
notion without clear evidence showing a direct causality between Facebook use and depression
(e.g., Magid, 2011), there is still an ongoing debate on whether online social networking leads to
clinical disorders such as depression. For example, Pantic et al. (2012) found a positive
relationship between frequency of Facebook use and indicators of depression among high school
students. Another recent survey study found no evidence suggesting the relationship between
SNS use and clinical depression (Jelenchick, Eickhoff, & Moreno, 2013). These studies,
however, used shallow measures of SNS use such as frequency of SNS use or time spent on SNS
without considering actual activities or communication exchanged on those sites, which might
have caused empty debates based on inaccurate evidence.

To address the need for nuanced measures for SNS use, Burke et al. (2010) distinguished
between person-directed communication and consumption of content on a social networking site.
This study found that person-directed communication lowered loneliness and enhanced bonding
social capital, which is a perception of available emotional and/or tangible aids from one’s social
ties (Ellison et al., 2007). The solitary consumption of content, however, reduced social capital
and enhanced loneliness. This study provided strong evidence that the mere use of social
networking sites might not be a good indicator of online social networking, because it includes
the solitary consumption of content which is not any different from general Internet use.

In sum, the type of online interaction should be examined to better understand online social
networking that produces positive psychological outcomes. To further address this issue, the
current study focuses on the amount of supportive interaction; namely, the amount of directed
supportive communication a user engages in with another user. Admittedly, the amount of
supportive interaction is likely associated with the level of general SNS use and the number of
SNS friends, as people who have many SNS friends and use SNS frequently will have a greater
probability of being involved in supportive interaction with other users. This study, therefore,
begins with proposing that the number of SNS friends and the frequency of online social
networking are positively associated with the amount of supportive interaction on social
networking sites.

H1. The number of SNS friends is positively associated with the amount of supportive
interaction on the social networking site.

H2. The frequency of online social networking is positively associated with the amount of
supportive interaction on the social networking site.

2.2. Supportive interaction and social support on social networking sites

Social support, defined as the resources or aids exchanged between individuals through
interpersonal ties (Cohen & Hoberman, 1983), is one of the key benefits that SNS users perceive
from online social networking (Park, Kee, & Valenzuela, 2009). This concept is considered as
one of the most widely used concepts across various domains in social science, and has
expanded its scope into the online context to explain the positive outcomes of online social
networking.

Studies on online social networking have examined the exchange of social support through
online social networking and its outcomes. Acquiring social support from others in a social
network is found to be one of the most important reasons for online social networking (Park et
al., 2009). In a recent report, SNS users perceived a greater level of emotional support and
companionship than did general Internet users (Hampton, Goulet, Rainie, & Purcell, 2011), at a
level that was almost equivalent to the amount that married or cohabitating Americans normally
perceive from their live-in partners.

Although social networking sites may provide greater opportunities for exchanging social
support, not all the elements of these sites are associated with social support. SNS users can use
direct communication services, such as chatting and direct messaging, but they can also engage
in third-party activities, such as updating profiles, viewing photographs of celebrities, searching
for events at commercial companies, and playing games. Measuring general SNS use and linking
the use with perceived social support is likely to produce a weak relationship because of the
redundant indicators of online social networking. A recent study corroborates this argument, as it
failed to reveal any significant relationship between Facebook intensity and social support (Vitak
et al., 2011).

The current study examines the amount of supportive interaction that individuals engage in with
others through direct user-to-user communication. Since the actual amount of received or
exchanged support seldom fully translates into recipients’ perception of social support (Haber,
Cohen, Lucas, & Baltes, 2007), the positive affect generated after the interaction is
conceptualized as a mediator that links the amount of supportive interaction and perceived social
support. A detailed discussion on affect and its relationship to social support appears in the
following section. Hypotheses 3 and 4 address the relationship between the amount of supportive
interaction, positive affect generated after the interaction, and perceived social support.

H3. The amount of supportive interaction on the social networking site is positively associated
with positive affect after the interaction.

H4. The positive affect generated after supportive interaction on the social networking site is
positively associated with perceived social support from SNS friends.

2.3. Psychological outcomes of social support: Life satisfaction, and a sense of community

In the social support literature, positive outcomes of social support are often more psychological
than physical (Schaefer, Coyne, & Lazarus, 1981). Psychological well-being and life satisfaction
are the most studied topics in this domain. Life satisfaction is the overall evaluation of one’s
environment, which can be either positive or negative (Scheufele & Shah, 2000). Traditionally,
this concept is frequently linked with the quality of a social network and the perceived social
support acquired from the network (Pinquart and Sörensen, 2000 and Young, 2006) as a result of
interpersonal communication that produces positive affect (Diener, Sandvik, & Pavot, 1991).
Following prior studies on social support and life satisfaction, the current study tests the
relationship between perceived social support and life satisfaction. Perceived social support
should be positively related to life satisfaction, as the existing studies suggest that perceived
social support, but not necessarily objective amount of supportive interaction, is strongly
associated with life satisfaction (Haber et al., 2007 and Kazarian and McCabe, 1991).

H5. Perceived social support is positively associated with life satisfaction.

Positive affect, as opposed to negative affect, is a more proactive measure that is closely linked
to subjective well-being and life satisfaction. Affect is a combination of moods and emotions and
is seen as “on-line evaluation toward the events that occur in lives” (Diener, Suh, Lucas, &
Smith, 1999, p. 277). For decades, positive emotions and affect have been studied and confirmed
as key predictors of life satisfaction (Fredrickson & Joiner, 2002). Moreover, affect was studied
as being closely related to social support, either as another indicator (Aanes, Mittelmark, &
Hetland, 2010), or as a mediator that transfers the influence of social support to positive health
behavior (Gonzalez et al., 2004) and life satisfaction (Greenglass & Fiksenbaum, 2009).
Therefore, it is proposed that positive affect derived from supportive interaction is also positively
associated with life satisfaction.

H6. Positive affect is positively associated with life satisfaction.

Another important psychological outcome of perceived social support is the sense of community.
It is defined as a feeling of belonging to a group or community within which members perceive
themselves as similar in characteristics and dependent on each other (Sarason, 1974). Studies on
the relationship between SNS use and indicators of sense of community have often produced
mixed findings. On one hand, SNS users who frequently update their statuses were reported to
feel a greater a sense of connectedness with other SNS users than those with less frequent
updates (Köbler et al., 2010). On the other, Facebook intensity was sometimes found to be
negatively associated with sense of belonging (Klingensmith, 2010). As examining mere
frequency of interactions produces weak arguments, Kwak, Shah, and Holbert (2004)
highlighted the need for measuring the type of interaction in order to better predict social
relationships. To further address this argument, it is proposed that perceived social support from
supportive interaction is closely associated with a sense of community.

H7. Perceived social support acquired from online social interaction is positively associated with
sense of community.

Since social support and sense of community are often referred to as byproducts of social
networks, both have been investigated as determinants of psychological well-being of people
within real-world social networks (Vieno, Santinello, Pastore, & Perkins, 2007). Many studies
have confirmed the role of social support and sense of community as predictors of life
satisfaction (Kutek et al., 2011 and Manago et al., 2012), and prior research has also proposed a
mechanism whereby the effect of social support on psychological outcomes is mediated by that
sense of community (Vieno et al., 2007). Therefore, the current study tested the direct
relationship between sense of community and life satisfaction.

H8. Sense of community is positively associated with life satisfaction.

Online social networking platforms are designed to engage in different supportive interactions
(i.e., share information, provide encouragement, and show liking), which may produce
differential outcomes and be linked to differential social support dimensions. For example,
positive feedback received from SNS friends was found to be the source of enhanced self-esteem
(Valkenburg et al., 2006). Different tools and applications on social networking sites may
accommodate supportive interactions in different forms. For example, esteem support can be
expressed by “liking” a friend’s new profile picture, whereas companionship can be conveyed by
inviting a friend to a social networking group. Prior studies have reported distinctive roles of
different social support dimensions in offline environments (Richmond et al., 2007 and Young,
2006), but little attention has been paid to the nuanced effects of social support exchanges online.
This study sorts out the unique proportion of such outcomes explained by individual support
dimension (i.e., esteem, appraisal, and companionship). To further explore the differential
relationships of social support types, following research questions are proposed.
RQ1. How does positive affect after the interaction is related to different dimensions of social
support?

RQ2. How does each dimension of social support, namely companionship, appraisal, and esteem
support, is associated with sense of community?

RQ3. How does each dimension of social support, namely companionship, appraisal, and esteem
support, is associated with life satisfaction?

Fig. 1 depicts a path model to examine the eight hypotheses and three research questions
proposed above.

Fig. 1. 

Conceptual model with proposed hypotheses.

Figure options

2.4. Test of daily supportive interaction and immediate affect after the interaction

For a more robust test of the relationship between supportive interaction and affect, this study
examines the relationship between the amount of supportive interaction and the affect
immediately after the interaction on a daily basis. Specifically, the study examines whether
general SNS use variables including (1) the frequency of SNS use, (2) the number of SNS
friends, (3) the number of close SNS friends, and (4) the history of SNS use have direct impacts
on the daily affect of participants immediately following the interaction. Also, this study
explores whether four general SNS use variables can moderate the relationship between the
amount of supportive interaction on SNS and user affect after that interaction.

One additional hypothesis and two research questions are, therefore, examined. Hypothesis 9
addresses the relationship between supportive interaction and positive affect. As discussed, the
amount of supportive interaction should be associated with positive affect, but not with negative
affect. Research questions 4 and 5 examine whether general SNS usage variables have direct
impacts on users’ affective states or moderate the relationship between their supportive
interaction and affect. Here, general SNS use variables represent the variables that are measured
only once, and assumed to be stable over a course of 5 days, whereas momentary variables (i.e.,
supportive interaction and affect following the interaction) are those that are measured each day
on a daily basis.
H9. The amount of daily supportive interactions on social networking sites is positively
associated with positive affect, but not with negative affect following the interaction on each
day.

RQ4. Are indicators of general SNS use (i.e., total number of SNS friends, number of close SNS
friends, frequency of SNS use, and history of SNS use) directly associated with positive affect
after the interaction on each day?

RQ5. Do indicators of general SNS use (i.e., total number of SNS friends, number of close SNS
friends, frequency of SNS use, and history of SNS use) moderate the relationship between the
amount of daily supportive interaction and positive affect after the interaction on each day?

3. Method
3.1. Participants

Participants were recruited using a snowball sampling technique. Undergraduate students in a

media research course at a major Midwestern university were asked to recruit their friends and
family members. The recruitment was a part of a class project: Students had to recruit five
people from their peers and five people from their non-peer networks (i.e., family). This
recruitment yielded 339 participants in total. The age of the participants ranged from 18 to 81
with a mean of 28 years old (SD = 11.77, Median = 23 years old). More than 50% were women
(51.2%). Also, most of the participants had a personal computer (92.6%), whereas 72.1% had a
smart phone and 15.6% owned a tablet PC.

3.2. Procedure

The current study employs a momentary sampling technique. Momentary sampling allows
participants to document feelings and behaviors as they go about their daily activities (Carpentier
et al., 2008). This technique is beneficial when recording affect that results from certain
behaviors because the time lag that occurs between the experience and reporting is very brief.
Studies have shown various benefits of the momentary sampling technique over other self-report
methods in terms of capturing the affect people feel as a result of their media use, as it reduces
errors inherent in retrospective reports (Carpentier et al., 2008 and Wood et al., 2002).

3.2.1. Diary

Respondents were given a diary survey booklet and asked to record their first session of SNS use
each evening over a 5-day period (Sunday–Thursday). The title page of the booklet included a
brief introduction about the study including the purpose of the survey and general instructions
for keeping the diary. The second page then contained questions about participants’ general SNS
use and detailed instructions on how to respond to the questions about their SNS use and feelings
after their SNS session on each day. Starting from the third page, participants were asked to
record their first social networking session for each evening (after 6:00 pm) including (1) the title
of SNS, (2) the time the session started and ended, (3) the features they used (i.e., wall post,
message, etc.), (4) the type of supportive interaction and support exchanged during the
interaction, and (5) the affect they experienced right after that session. Social networking sites
were defined here as web-based applications such as Facebook, MySpace, and Cyworld that
enable people to connect with others by creating a profile, inviting friends to access the profile,
and exchanging messages. In the last page, participants were asked to provide their gender and
age. They also reported their general perception about social support, sense of community, and
life satisfaction.

3.3. Measures

3.3.1. Person-level variables

First, respondents were asked to indicate the total number of their SNS friends (M = 574.26,
SD = 377.61) and how many of them were their close friends (M = 46.78, SD = 58.76). The total
number of close friends was later log-transformed as the data were significantly right-skewed
(4.57). The frequency of using SNSs was measured with a 6-point ordinal scale ranging from (1)
once a week or less to (6) more than 5 times a day. Responses were well-distributed: 25%
reported that they use SNS more than 5 times a day and other options fell between 15% and 20%
except for “once a week or less” (5.3%). The history of SNS use was measured by asking
respondents how long they have been using SNSs. A majority of respondents indicated that they
have been using SNSs for more than 2 years (74.3%).

Perceived social support was measured with 9 items reflecting three dimensions of perceived
social support: appraisal support, companionship, and esteem support. Each item was measured
with a 7 point Likert-type scale ranging from (1) definitely false to (7) definitely true. The items
were adopted from Eastin and LaRose (2005), which modified Cohen’s original Interpersonal
Support Evaluation List (ISEL) (Cohen, Mermelstein, Kamarck, & Hoberman, 1985). The
tangible support dimension was excluded as it was found to be irrelevant in prior online social
support studies (Braithwaite, Waldron, & Finn, 1999). Therefore, we measured three
dimensions: appraisal support (M = 6.18, SD = .94, α = .87), companionship support (M = 5.57,
SD = 1.28, α = .75), and esteem support (M = 5.52, SD = 1.19, α = .72) in this study and treated
each dimension as a separate construct. A confirmatory factor analysis with the three factors
explained by the nine items produced an acceptable fit: χ2 (24) = 40.80, p = .02, CFI = .99,
RMSEA = .05.

Sense of community was measured with three items (Sheldon & Gunz, 2009) using a 7 point
Likert-type scale ranging from (1) definitely false to (7) definitely true (M = 5.69, SD = 1.04,
α = .77). Originally, the sense of community scale had 6 items, but the pretest results showed
that the three reversed items (i.e., I am lonely, I feel unappreciated by one or more important
people, and I have disagreements with people I get along with) significantly decreased the
reliability of the scale, so they were excluded from the scale for further analysis.

Life satisfaction was measured with four items (e.g., I am satisfied with my life) using a 7 point
Likert-type scale ranging from (1) definitely false to (7) definitely true (M = 5.22, SD = 1.02,
α = .86). The items were adopted from the Satisfaction with Life Scale (SWLS) (Diener,
Emmons, Larsen, & Griffin, 1985). Table 1 shows the actual items for all the variables with their
descriptive information and reliability scores.

Table 1.

Means, standard deviations, factor loadings, and alpha coefficients of person-level and
momentary variables.

Items N M SD Loadingb α
Person-level variables
 Items N M SD Loadingb α
Total SNS friend (network size) 320 574.26 377.61
Close SNS friend 318 46.78 58.76
Frequency of SNS use 295 2.94 1.56
SNS history 288 2.98 .98
Appraisal support 330 6.18 0.94 .87
 There is at least one person I know whose advice I really trust 330 6.32 1.01 .84
 There is really someone who can give me objective feedback… 330 6.16 1.02 .84
 I feel that there is someone with whom I can share my most
330 6.05 1.14 .81
private
Companionship 331 5.57 1.28 .75
 No one I know would throw a birthday party for me a 330 5.91 1.56 .62
 Most people I know don’t enjoy the same things that I do a 330 5.35 1.67 .73
 I feel that I’m on the fringe in my circle of friends a 329 5.44 1.49 .77
Esteem support 332 5.52 1.19 .72
 People don’t have much confidence in me a 330 5.80 1.42 .79
 My friends feel that I’m not very good at helping them solve
331 5.73 1.47 .68
problemsa
 Most of my friends are more successful at making changes in
331 5.05 1.57 .56
their lives than I ama
Sense of community 332 5.69 1.04 .77
 I feel sense of contact with people who care for me 330 5.55 1.31
 I feel connected with others who are important to me 332 5.92 1.13
 I feel intimate with people I spent time with 330 5.60 1.30
Life satisfaction 332 5.22 1.02 .86
 In most ways my life is close to my ideal 332 5.03 1.17
 The conditions of my life are excellent 332 5.16 1.23
 I am satisfied with my life 331 5.50 1.12
 So far I’ve gotten the important things I want in life 332 5.20 1.31
Momentary variables
Number of supportive interaction (weekly average) 1314 2.24 3.31
 Sunday 298 2.77 3.38
 Monday 262 2.15 2.30
 Tuesday 249 1.88 2.91
 Wednesday 256 2.22 3.59
 Thursday 249 2.08 3.28
Positive affect (weekly average) 1285 4.08 .04 .86
 Sunday 294 4.24 .07
 Monday 254 4.00 .08
 Tuesday 241 4.07 .09
 Wednesday 249 4.04 .08
 Thursday 247 4.02 .09
Negative affect (weekly average) 1284 2.47 1.01 .81
 Items N M SD Loadingb α
 Sunday 293 2.42 1.10
 Monday 254 2.45 1.18
 Tuesday 241 2.55 1.14
 Wednesday 249 2.49 1.23
 Thursday 247 2.46 1.11
a

Reversed item.

Standardized coefficient.

Table options

3.3.2. Momentary variables

The study includes two momentary measures that were reported five times. The momentary
measures were the amount of supportive interaction and the affect generated immediately after
the interaction. Supportive interaction was measured by asking participants to indicate whether
they had interacted with other people during their first SNS session in each evening, and whether
they had provided or received each of the following types of support during the interaction: (1)
giving advice, (2) showing empathy, (3) validating thoughts, (4) complimenting, (5) teaching
something new, (6) inviting to a new group, (7) sharing information, (8) giving encouragement,
and (9) inviting to a group plan. Respondents were asked to check all types of support they had
provided to others or received from other users. Then, the total number of supportive interaction
was reconstructed as a single item-variable by counting the number of positive responses
(“yes”).

Affect was measured with the Positive and Negative Affect Schedule (PANAS) scale (Watson,
Clark, & Tellegen, 1988), supplemented by items that came out from a pretest through
undergraduate students in a media research class. Students were asked to indicate other feelings
they had after using social networking sites that were not listed in PANAS, which yielded
thirteen items in total. Each item was measured with a 7-point scale ranging from (1) definitely
false to (7) definitely true. The exploratory factor analysis with the Varimax rotation was
performed to validate our measure. Thirteen items yielded a two-factor solution, namely positive
affect (weekly average = 4.08, SD = .04, α = .86) and negative affect (weekly average = 2.47,
SD = 1.01, α = 81). One item (i.e. Alert) was excluded from the further analysis, as it was cross-
loaded to both positive and negative affect dimensions. Twelve items explained 62.57% of the
total affect variance. Table 1 shows daily average of both positive affect and negative affect.

3.4. Analytic strategies

3.4.1. Path model

A path model was utilized to test the proposed relationships among the number of SNS friends,
the frequency of SNS use, the amount of supportive interactions, affect, perceived social support,
sense of community, and life satisfaction. Because the model included two momentary variables
(i.e., the amount of supportive interaction and affect), they were aggregated across each
participant to create two person-level variables by averaging the amount of supportive
interactions and affect across 5 days. For the path analysis, the AMOS 19.0 program was utilized
to obtain maximum-likelihood (ML) estimates of the model parameters (Arbuckle, 2010). Table
2 shows the correlation matrix for estimating the recursive model hypothesized in this study. The
number of SNS friends and the frequency of SNS use were treated as exogenous variables, and
the other seven variables – the amount of supportive interaction, positive affect, perceived social
support (appraisal, esteem, and companionship), sense of community, and life-satisfaction –
were endogenous variables. Incomplete survey responses were list-wise deleted, which resulted
in the final sample size of 295. Due to its low correlations with other key variables, the number
of close SNS friends was excluded from the path model.

Table 2.

Zero-order correlation among the key variables.

1 2 3 4 5 6 7 8 9 10 11
1. Total SNS friends 1.00
2. Close SNS friends .48 1.00
3. Frequency of SNS
.37⁎⁎ .30 1.00
use
4. Supportive
.21⁎⁎⁎ .05 .14⁎ 1.00
interactiona
5. Positive affecta .12⁎ −.02 .03 .32⁎⁎⁎ 1.00
6. Negative affecta .14⁎ −.07 .12⁎ .10 .25⁎⁎⁎ 1.00
7. Appraisal support −.05 −.05 −.07 .10 .15⁎⁎ −.22⁎⁎⁎ 1.00
8. Companionship .11 .02 .01 .05 .21⁎⁎⁎ −.28⁎⁎⁎ .46⁎⁎⁎ 1.00
9. Esteem support −.01 −.09 −.03 .07 .08 −.40⁎⁎⁎ .36⁎⁎⁎ .58⁎⁎⁎ 1.00
10. Sense of
−.02 −.03 −.01 .08 .14⁎ −.25⁎⁎⁎ .65⁎⁎⁎ .44⁎⁎⁎ .41⁎⁎⁎ 1.00
community
11. Life satisfaction .09 .06 −.05 .07 .30⁎ −.04 .37⁎⁎⁎ .41⁎⁎⁎ .29⁎⁎⁎ .48⁎⁎⁎ 1.00

Note: Listwise deletion, N = 295.

p < .05.

⁎⁎

p < .01.

⁎⁎⁎

p < .001.

a
Weekly average.

Table options

The initial path model included the frequency of SNS use and the number of SNS friends (total
number of SNS friends) as exogenous variables. Although the initial model showed almost a
perfect fit, χ2 (17) = 11.67, p = .82, CFI = 1.00, and RMSEA = .00, the frequency of SNS use
was not significantly associated with the amount of supportive interaction. A model without the
frequency of SNS use still showed a good model fit, χ2 (12) = 17.27, p = .14, CFI = .99, and
RMSEA = .04. The chi-square difference testing indicated no significant change in the model fit
between the two models (χ2Δ = 5.6, p = .35). To obtain a parsimonious model, the second model
was retained. From this model, 32% of the variance in life satisfaction were explained, and 46%
of the variance in the sense of community were explained. Multicollinearity or normality
problems were not found in the data. Fig. 2 presents our final model with standardized
coefficients and their statistical significance.

Fig. 2. 

Final model with standardized coefficient for each path.

Figure options

3.4.2. Multilevel model using momentary variables

The direct relationship between the amount of supportive interaction and affect following the
interaction that participants reported each day were also analyzed using multilevel analysis
(SPSS 19.0) with an estimation of restricted maximum likelihood. The diary design yielded a
nested data structure within subjects, such that their responses across 5 days were not
independent from each other. The multilevel modeling allows multiple levels of analysis with a
series of regression equations in order to account for nesting of assessments within subjects. The
data structure in this study required two levels. The first level variables consist of momentary
measures, the amount of supportive interaction and positive/negative affect. To test the within-
individual hypotheses, the relationships between affect (positive/negative) and supportive
interaction were tested at level 1, and the aggregated parameters at level 2 across all the
participants were then estimated. For the cross-level effect, slopes and intercepts were estimated
with the momentary variable (i.e. supportive interaction), the person-level variables (i.e., number
of SNS friends, frequency of SNS use, and history of SNS use), and the cross-level interactions
between the two levels. As suggested by Hofmann, Griffin, and Gavin (2000), all the predictors
were centered at each individual’s mean in order to remove any between-individual variance in
the estimates of the relationships among the variables. Hence, the intra-individual estimates
obtained from multilevel modeling were not confounded by individual differences. Considering
the sensitive and complex nature of multilevel modeling, this study proceeded from a null model
to a fixed coefficient model step-by-step with level-1 and level-2 variables. At each step, the
results were inspected to see which parameters are significant, and how much of the residual
errors were left unexplained at each of the two levels. After fitting the model, the intercept and
slope were explored using all person-level factors, the momentary factor, and cross-level
interactions. Days with no data entry were deleted resulting in the total sample size of 1322
social networking sessions.

4. Results
4.1. Path model: Outcomes of supportive interaction

The single exogenous variable, the number of SNS friends, was found to be a significant
predictor of the amount of supportive interaction (γ = .21, p < .001), supporting hypothesis 1.
However, hypothesis 2 was not supported as the initial model showed no significant relationship
between the frequency of SNS use and the amount of supportive interaction. The amount of
supportive interaction was significantly related to positive affect felt after the interaction
(β = .32, p < .001), supporting hypothesis 3. Positive affect then was a significant predictor of
perceived companionship support (β = .20, p < .01), appraisal support (β = .15, p < .01), and life
satisfaction (β = .21, p < .001). This result partially supported hypothesis 4, but fully supported
hypothesis 5. Among the three perceived social support dimensions, perceived companionship
support was a positive predictor of life satisfaction (β = .20, p < .001), whereas perceived
appraisal and esteem support were positively related to sense of community (β = .55, p < .001
and β = .16, p < .001 respectively), but not to life satisfaction. Therefore, hypotheses 6 and 7
were partially supported. The findings also address research questions 1–3, by showing
differential relationships of each social support dimension to affect, sense of community, and life
satisfaction. Finally, the sense of community was positively associated with life satisfaction
(β = .36, p < .001), which supports hypothesis 8.

4.2. Multilevel model: Relationship between supportive interaction and affect

Before testing hypothesis 9 and research questions associated with the multi-level analyses, a
null model without level-1 and level-2 predictors was tested in order to ensure that the proposed
variables had sufficient variance to influence momentary affect and the relationship between the
amount of supportive interactions and affect. The null model yielded an ICC score of .53,
indicating that the effect of the between-individual variance was significant enough to influence
the within-individual variance (Bliese, 2000). Table 3 provides the results obtained from a series
of multilevel modeling estimated to predict affect. Since including the interactions between
momentary and person-level variables significantly worsened the model (Δχ2 = 33.50, p < .001),
we only estimated the main effects of the person-level variables, but not the cross-level
interactions between momentary and person-level variables.

Table 3.
Estimates of fixed effects on positive affect.

Within/between
Unconditional Within- Within/between
individual interaction
means model individual model individual model
model

Coefficient S.E. Coefficient S.E. Coefficient S.E. Coefficient S.E.

Mean intercept 4.09⁎⁎⁎ .06 4.09⁎⁎⁎ .06 4.11⁎⁎⁎ .06 4.12⁎⁎⁎ .06
Level 1
Supportive
.09∗∗∗ .01 .10∗∗∗ .01 .10∗∗∗ .01
interaction
Level 2
Total SNS friend .00 .00 .00 .00
Close SNS friend −.28 .17 −.27 .17
Frequency of
−.10 .06 .06 .08
SNS use
History of SNS
.22⁎ .08 .23∗ .09
use
Cross-level interaction
Total SNS
.00 .00
friends * SI
Close SNS
−.03 .04
friends * SI
Frequency * SI .00 .12
History * SI −.03 .02
Δχ2 127.79∗∗∗ 998.28∗∗∗ −33.50∗∗∗
Δdf 1 5 4

Note: Significance test follows t-ratio distribution; SI = supportive interaction.

 p < .01.
∗∗

p < .05.

∗∗∗

p < .001.

Table options

The first multi-level model estimated to predict positive affect indicated a positive relationship
between the amount of supportive interaction and positive affect, after accounting for the person-
level variables (β = .10, p < .001). Among the person-level variables, however, only the history
of SNS use was significantly and directly related to the momentary positive affect (β = .22,
p < .05). Other person-level variables did not significantly influence positive affect or the
relationship between supportive interaction and positive affect. The relationship between
supportive interaction and negative affect was also tested using the identical procedure. The null
model suggested a significant variability both within a person and between persons. However,
when the second model included a fixed effect of supportive interaction on negative affect, there
was non-significant relationship between supportive interaction and negative affect (β = .01,
p = .18). In other words, although there was a significant variability in negative affect between
persons and within one person across 5 days, this variability cannot be attributed to the amount
of supportive interaction or other person-level variables. These results supported hypothesis 9.
Since there was no fixed effect of supportive interaction on negative effect, further models
including person-level variables were not fitted.

5. Discussion

The main purpose of this study was to explore a mechanism through which people’s online
social networking produces positive psychological outcomes. Using momentary sampling, the
direct relationship between the amount of supportive interaction and immediate affect after the
interaction each day was assessed. It is also tested whether the size of one’s online social
network (the total number of friends and close friends in particular), frequency of online social
networking, and history of networking directly influence one’s affective state or moderate the
relationship between supportive interaction and affect.

A series of multilevel models showed that the amount of supportive interaction was positively
associated with positive affect after the interaction, such that those who exchanged a greater
amount of support with other members felt a greater level of positive affect. However, the
amount of supportive interaction was not associated with negative affect. This result confirms
the prior social support literature which found a significant relationship between real world
social support and positive affect, but not between social support and negative affect (Baker,
Jodrey, & Intagliata, 1992). Existing studies acknowledge that social support seldom directly
influences negative affect, but can buffer negative influences of stressful situations on one’s
psychological health (Cohen & McKay, 1984).

The multilevel models further showed that the network size (the number of SNS friends) or other
indicators of general SNS use were not significantly associated with either positive or negative
affect after online social networking. Also, these variables failed to moderate the relationship
between the amount of supportive interaction and immediate affect after the interaction. This
result corroborates the argument that the type of activities people engage in social networking
sites better predicts psychological outcomes of SNS use (Burke et al., 2010) than do general
measures such as Facebook intensity (Ellison et al., 2007).

The path model reveals a possible mechanism whereby the network structure or the frequency of
social networking influences psychological well-being. The number of SNS friends led to
positive psychological outcomes, but only when it was accompanied with actual supportive
interaction and positive feelings after the interaction. Prior studies found inconsistent
relationships between the number of SNS friends and psychological states (Kalpidou et al.,
2011 and Kim and Lee, 2011). Some suggested that the number of friends plays a positive role
only up to a certain number (Kim & Lee, 2011). The present findings might contribute to
reconciling the mixed findings by highlighting the importance of the actual type and quality of
interaction between SNS users.
In the current study, the frequency of SNS use was not a significant predictor of supportive
interaction and other outcome variables. This result suggests that it is the quality of interaction
that matters in establishing social support and psychological well-being, but not the frequency or
amount of SNS use. Because these sites now have various solitary applications such as games
and photo editing, users can engage with the sites without interacting with other users. Relevant
research indicates that not all uses of SNS are social (Burke et al., 2010 and Ellison et al., 2011).
Findings of the present study show that general measures of Facebook use (i.e., Facebook
intensity) might not be a direct predictor of well-being. The results further suggest that clinical
researchers who have been debating over the linkage between general SNS use (i.e., frequency of
SNS use and hours spent on SNS) and clinical disorders (e.g., Jelenchick et al., 2013, O’Keeffe
and Clarke-Pearson, 2011 and Pantic et al., 2012) could have overlooked the possibility that
different elements of SNS could lead to either positive or negative clinical outcomes. SNS
researchers will benefit from combining general SNS usage measures with more specific
measures capturing what people actually do on social networking sites and how people interact
with their SNS friends.

Positive affect felt by SNS users after online social networking was positively associated with
perceived companionship support, appraisal support, and life satisfaction. Among the three
support dimensions, companionship support was the only predictor of life satisfaction, whereas
perceived appraisal and esteem support were indirectly associated with life satisfaction through
the enhanced sense of community. Prior work combined dimensions of social support in a
unitary construct. However, a major limitation with such application is overseeing the
differences in the gravity of each factor on enhancing a sense of community and life satisfaction.
Researchers have highlighted a need for distinguishing between different types of support
(Richmond et al., 2007). The current study supports this claim. Specifically, the companionship
dimension of social support had a direct and positive relationship with life satisfaction, whereas
appraisal and esteem support had indirect effects on life satisfaction through the sense of
community.

This finding can be due to the difference in our participants’ relationship-orientation. Studies on
the age and social relationship suggest that young adults tend to be more future-oriented, and
thus value an expansion of one’s network, whereas older people engage more in emotional
interaction with their close ties (Lang, 2001). Since more than 80% of our study participants
were younger than 30 years old (M = 27.56, SD = 11.77), it is possible that their life evaluation
was directly associated with whether they have friends who they can hang out with, share similar
interest, and praise each other. Social support researchers have also suggested that characteristics
of support recipients might influence how they interpret acquired social support ( Collins and
Feeney, 2004 and Lakey and Cassady, 1990). Future study should explore further how each
dimension of social support engenders differential outcomes across different groups of people.

6. Conclusion

Getting social support has long been a key to life satisfaction. Tools to fulfill such needs have
advanced with the development of new communication tools, and the Internet and social media
have entirely changed the ways people network. Although social networking sites do help people
expand and manage their network easily, this network might not be meaningful if it fails to
provide social support that recipients need. The present study highlights the importance of
looking at specific supportive elements of social networking sites for better understanding the
functions and benefits of those sites.
 6.1. Limitation and future research

First, this research used correlational data, preventing inference of causal direction. Although the
proposed mode achieved almost a perfect fit, it is also very likely that rival models can fit the
data equally well. With the path model proposed, the positive effects of supportive interaction on
perceived social support, sense of community, and life satisfaction were tested. However, it is
also likely that a greater sense of community leads people to actively engage in supportive
interaction with others. Some SNS researchers note that the relationship between psychological
states and SNS use can be reciprocal, such that those with lower levels of life satisfaction
increase their use of social networking sites to increase their well-being (Ellison et al.,
2007 and LaRose et al., 2010). Future studies could conduct a longitudinal study for a robust test
on the causal relationships proposed here, and to rule out rival explanations (e.g., LaRose et al.,
2011).

Second, the current study focuses on the amount of positive interactions on social networking
sites. However, some studies have reported negative outcomes SNS use. For example,
Valkenburg et al. (2006) found a negative impact of appraisal support on adolescents’ self-
esteem and well-being when the feedback from peers was negative. We should understand the
complexities of supportive communications, rather than assuming that all support intentions
result in success. Future studies might benefit from testing the relationship between negative
interactions and psychological outcomes, or unintended negative consequences of users’
supportive communication on social networking sites. With this respect, it is interesting to note
that positive affect and negative affect were positively correlated (r = .25, p < .001). This result
indicates that interactions on social networking sites can generate mixed emotions and, thereby,
lead to differential psychological outcomes

References

1.
o Aanes et al., 2010
o M.M. Aanes, M.B. Mittelmark, J. Hetland

o Interpersonal stress and poor health: The mediating role of loneliness

o European Psychologist, 15 (2010), pp. 3–11

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (11)

2.
o Arbuckle, 2010
o J.L. Arbuckle
o Amos 19.0 user’s guide Chicago

o SPSS, Chicago, IL (2010)

o
3.
o Baker et al., 1992
o F. Baker, D. Jodrey, J. Intagliata

o Social support and quality of life of community support clients

o Community Mental Health Journal, 28 (1992), pp. 397–411

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (37)

4.
o Bliese, 2000
o P.D. Bliese

o Within-group agreement, non-independence, and reliability: Implications for data

aggregation and analysis

o K.J. Klein, S.W.J. Kozlowski (Eds.), Multilevel theory, research, and methods in
organizations: Foundations, extensions, and new directions, Jossey-Bass, San Francisco, CA
(2000), pp. 349–381
o View Record in Scopus

| Cited By in Scopus (1)

5.
o Boyd, 2006
o D. Boyd

o Friends, friendsters, and top 8: Writing community into being on social network
sites

o First Monday, 11 (2006)

o
6.
o Braithwaite et al., 1999
o D.O. Braithwaite, V.R. Waldron, J. Finn

o Communication of social support in computer-mediated groups for people with

disabilities
o Health Communication, 11 (1999), pp. 123–151
o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (137)

7.
o Burke et al., 2010
o Burke, M., Marlow, C., & Lento, T. (2010). Social network activity and social well-being.
In Proceedings of the 2010 ACM conference on human factors in computing systems (pp. 1909–1912).
New York, NY: ACM.
o
8.
o Carpentier et al., 2008
o F.R.D. Carpentier, J.D. Brown, M. Bertocci, J.S. Silk, E.E. Forbes, R.E. Dahl

o Sad kids, sad media? Applying mood management theory to depressed

adolescents’ use of media

o Media Psychology, 11 (2008), pp. 143–166

o View Record in Scopus

| Cited By in Scopus (9)

9.
o Cohen and Hoberman, 1983
o S. Cohen, H.M. Hoberman

o Positive events and social supports as buffers of life change stress1

o Journal of Applied Social Psychology, 13 (1983), pp. 99–125

o Full Text via CrossRef
10.
o Cohen and McKay, 1984
o S. Cohen, G. McKay

o Social support, stress, and the buffering hypothesis: A theoretical analysis

o A. Baum, S.E. Taylor, J.E. Singer (Eds.), Handbook of psychology and health,
Hillsdale, NY (1984), pp. 253–267
o View Record in Scopus

| Cited By in Scopus (5)

11.
o Cohen et al., 1985
o S. Cohen, R. Mermelstein, T. Kamarck, H. Hoberman
o Measuring the functional components of social support

o I.G. Sarason, B.R. Sarason (Eds.), Social support: Theory, research and
applications, Martinus Nijhoff, The Hague, The Netherlands (1985), pp. 73–94
o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (1)

12.
o Collins and Feeney, 2004
o N.L. Collins, B.C. Feeney

o Working models of attachment shape perceptions of social support: Evidence

from experimental and observational studies

o Journal of Personality and Social Psychology, 87 (2004), pp. 363–383

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (122)

13.
o comScore, 2011
o comScore. (2011). Social networking: Total time spent online.
<http://blog.comscore.com/2011/06/> Retrieved 28.03.12.
o
14.
o Diener et al., 1985
o E. Diener, R.A. Emmons, R.J. Larsen, S. Griffin

o The satisfaction with life scale

o Journal of Personality Assessment, 49 (1985), pp. 71–75

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (4087)

15.
o Diener et al., 1991
o E. Diener, E. Sandvik, W. Pavot

o Happiness is the frequency, not the intensity, of positive versus negative affect

o F. Strack, M. Argyle, N. Schwarz (Eds.), Subjective well-being: An

interdisciplinary perspective, Pergamon Press, Oxford, England (1991), pp. 119–139
o
16.
o Diener, Suh, Lucas, & Smith, 1999
o E. Diener, E.M. Suh, R.E. Lucas, H.L. Smith

o Subjective well-being: Three decades of progress

o Psychological Bulletin, 125 (1999), pp. 276–302

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (1995)

17.
o Eastin and LaRose, 2005
o M.S. Eastin, R. LaRose

o Alt. support: Modeling social support online

o Computers in Human Behavior, 21 (2005), pp. 977–992

o Article

PDF (143 K)

View Record in Scopus

| Cited By in Scopus (41)

18.
o Ellison et al., 2007
o N.B. Ellison, C. Steinfield, C. Lampe

o The benefits of Facebook “friends”: Social capital and college students’ use of
online social network sites
o Journal of Computer-Mediated Communication, 12 (2007), pp. 1143–1168
o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (1056)

19.
o Ellison et al., 2011
o N.B. Ellison, C. Steinfield, C. Lampe

o Connection strategies: Social capital implications of Facebook-enabled

communication practices

o New Media & Society, 13 (2011), pp. 873–892

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (60)

20.
o Fredrickson and Joiner, 2002
o B.L. Fredrickson, T. Joiner

o Positive emotions trigger upward spirals toward emotional well-being

o Psychological Science, 13 (2002), pp. 172–175

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (377)

1.
o Gonzalez et al., 2004
o J.S. Gonzalez, F.J. Penedo, M.H. Antoni, R.E. Durán, S. McPherson-Baker, G. Ironson et
al.

o Social support, positive states of mind, and HIV treatment adherence in men and
women living with HIV/AIDS

o Health Psychology, 23 (2004), pp. 413–418

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (102)

2.
o Greenglass and Fiksenbaum, 2009
o E.R. Greenglass, L. Fiksenbaum

o Proactive coping, positive affect, and well-being

o European Psychologist, 14 (2009), pp. 29–39

o View Record in Scopus

| Cited By in Scopus (26)

3.
o Haber et al., 2007
o M.G. Haber, J.L. Cohen, T. Lucas, B.B. Baltes

o The relationship between self-reported received and perceived social support: A

meta-analytic review

o American Journal of Community Psychology, 39 (2007), pp. 133–144

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (75)

4.
o Hampton et al., 2011
o K. Hampton, L.S. Goulet, L. Rainie, K. Purcell

o Social networking sites and our lives, Vol. 2012Pew Research Center’s Internet &
American Life Project, Washington, DC (2011)
o
5.
o Hofmann et al., 2000
o D.A. Hofmann, M.A. Griffin, M.B. Gavin

o The application of hierarchical linear modeling to organizational research

o K.J. Klein, S.W.J. Kozlowski (Eds.), Multilevel theory, research, and methods in
organizations: Foundations, extensions, and new directions, Jossey-Bass, San Francisco, CA
(2000), pp. 467–511
o
6.
o Jelenchick et al., 2013
o L.A. Jelenchick, J.C. Eickhoff, M.A. Moreno

o “Facebook depression?” Social networking site use and depression in older

adolescents

o Journal of Adolescent Health, 52 (2013), pp. 128–130

o Article

PDF (346 K)

View Record in Scopus

| Cited By in Scopus (2)

7.
o Kalpidou et al., 2011
o M. Kalpidou, D. Costin, J. Morris

o The relationship between Facebook and the well-being of undergraduate college

students

o Cyberpsychology, Behavior, and Social Networking, 14 (2011), pp. 183–189

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (25)

8.
o Kazarian and McCabe, 1991
o S.S. Kazarian, S.B. McCabe

o Dimensions of social support in the MSPSS: Factorial structure, reliability, and

theoretical implications

o Journal of Community Psychology, 19 (1991), pp. 150–160

o View Record in Scopus
 |

Full Text via CrossRef

| Cited By in Scopus (75)

9.
o Kim and Lee, 2011
o J. Kim, J.E.R. Lee

o The Facebook paths to happiness: Effects of the number of Facebook friends and
self-presentation on subjective well-being

o Cyberpsychology, Behavior, and Social Networking, 14 (2011), pp. 359–364

o
10.
o Klingensmith, 2010
o Klingensmith, C. L. (2010). 500 friends and still friending: The relationship between
Facebook and college students’ social experiences. Honors Projects.
<http://digitalcommons.macalester.edu/psychology_honors/22> Retrieved 08.08.12.
o
11.
o Köbler et al., 2010
o Köbler, F., Riedl, C., Vetter, C., Leimeister, J. M., & Krcmar, H. (2010). Social
connectedness on Facebook – An explorative study on status message usage. In Proceedings of 16th
Americas conference on information systems (Vol. 247).
o
12.
o Kutek et al., 2011
o S.M. Kutek, D. Turnbull, A.K. Fairweather-Schmidt

o Rural men’s subjective well-being and the role of social support and sense of
community: Evidence for the potential benefit of enhancing informal networks

o Australian Journal of Rural Health, 19 (2011), pp. 20–26

o
13.
o Kwak et al., 2004
o N. Kwak, D.V. Shah, R.L. Holbert

o Connecting, trusting, and participating: The direct and interactive effects of social
associations

o Political Research Quarterly, 57 (2004), pp. 643–652

o
14.
o Lakey and Cassady, 1990
o B. Lakey, P.B. Cassady
o Cognitive processes in perceived social support

o Journal of Personality and Social Psychology, 59 (1990), pp. 337–343

o
15.
o Lang, 2001
o F.R. Lang

o Regulation of social relationships in later adulthood

o The Journals of Gerontology Series B: Psychological Sciences and Social

Sciences, 56 (2001), pp. 321–326
o
16.
o LaRose et al., 2011
o LaRose, R., Wohn, D. Y., Ellison, N., & Steinfield, C. (2011). Facebook fiends: Compulsive
social networking and adjustment to college. In Proceedings of the international association for the
development of the information society.
o
17.
o LaRose et al., 2010
o R. LaRose, J.H. Kim, W. Peng

o Social networking: Addictive, compulsive, problematic, or just another media

habit?

o Z. Pappacharissi (Ed.), A networked self: Identity, community, and culture on

social network sites, Routledge, New York, NY (2010), pp. 59–81
o
18.
o Magid, 2011
o Magid, L. (2011, March). Is there really “Facebook Depression?” Cnet.
<http://news.cnet.com/8301-19518_3-20048148-238.html> Retrieved from March 2011.
o
19.
o Manago et al., 2012
o A.M. Manago, T. Taylor, P.M. Greenfield

o Me and my 400 friends: The anatomy of college students’ Facebook networks,

their communication patterns, and well-being

o Developmental Psychology, 48 (2012), pp. 369–380

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (12)

20.
o O’Keeffe and Clarke-Pearson, 2011
o G.S. O’Keeffe, K. Clarke-Pearson

o The impact of social media on children, adolescents, and families

o Pediatrics, 127 (2011), pp. 800–804

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (55)

1.
o Pantic et al., 2012
o I. Pantic, A. Damjanovic, J. Todorovic, D. Topalovic, D. Bojovic-Jovic, S. Ristic et al.

o Association between online social networking and depression in high school

students: behavioral physiology viewpoint

o Psychiatria Danubina, 24 (2012), pp. 90–93

o View Record in Scopus

| Cited By in Scopus (4)

2.
o Park et al., 2009
o N. Park, K.F. Kee, S. Valenzuela

o Being immersed in social networking environment: Facebook groups, uses and

gratifications, and social outcomes

o CyberPsychology & Behavior, 12 (2009), pp. 729–733

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (85)

3.
o Pinquart and Sörensen, 2000
o M. Pinquart, S. Sörensen

o Influences of socioeconomic status, social network, and competence on subjective

well-being in later life: A meta-analysis
o Psychology and Aging, 15 (2000), pp. 187–224
o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (206)

4.
o Richmond et al., 2007
o C.A.M. Richmond, N.A. Ross, G.M. Egeland

o Social support and thriving health: A new approach to understanding the health of
indigenous Canadians

o American Journal of Public Health, 97 (2007), pp. 1827–1833

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (27)

5.
o Sarason, 1974
o S.B. Sarason

o The psychological sense of community: Prospects for a community psychology

o Jossey-Bass, Oxford, England (1974)

o
6.
o Schaefer et al., 1981
o C. Schaefer, J.C. Coyne, R.S. Lazarus

o The health-related functions of social support

o Journal of Behavioral Medicine, 4 (1981), pp. 381–406

o
7.
o Scheufele and Shah, 2000
o D.A. Scheufele, D.V. Shah

o Personality strength and social capital: The role of dispositional and informational
variables in the production of civic participation

o Communication Research, 27 (2000), pp. 107–131

o
8.
o Sheldon & Gunz, 2009
o K.M. Sheldon, A. Gunz

o Psychological needs as basic motives, not just experiential requirements

o Journal of Personality, 77 (2009), pp. 1467–1492

o
9.
o Simmons, 2011
o A. Simmons

o Social media consumer trend and benchmark report

o Simmons Market Research Bureau, New York, NY (2011)

o
10.
o Steinfield et al., 2008
o C. Steinfield, N.B. Ellison, C. Lampe

o Social capital, self-esteem, and use of online social network sites: A longitudinal
analysis

o Journal of Applied Developmental Psychology, 29 (2008), pp. 434–445

o
11.
o Valenzuela et al., 2009
o S. Valenzuela, N. Park, K.F. Kee

o Is there social capital in a social network site?: Facebook use and college
students’ life satisfaction, trust, and participation

o Journal of Computer-Mediated Communication, 14 (2009), pp. 875–901

o
12.
o Valkenburg et al., 2006
o P.M. Valkenburg, J. Peter, A.P. Schouten

o Friend networking sites and their relationship to adolescents’ well-being and

social self-esteem

o CyberPsychology & Behavior, 9 (2006), pp. 584–590

o View Record in Scopus

Full Text via CrossRef

 | Cited By in Scopus (119)

13.
o Vieno et al., 2007
o A. Vieno, M. Santinello, M. Pastore, D.D. Perkins

o Social support, sense of community in school, and self-efficacy as resources

during early adolescence: an integrative model

o American Journal of Community Psychology, 39 (2007), pp. 177–190

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (24)

14.
o Vitak et al., 2011
o Vitak, J., Ellison, N. B., & Steinfield, C. (2011). The ties that bond: Re-examining the
relationship between Facebook use and bonding social capital. In Proceedings of the 2011 44th Hawaii
international conference on system sciences (pp. 1–10). Washington, DC: IEEE Computer Society.
o
15.
o Watson et al., 1988
o D. Watson, L.A. Clark, A. Tellegen

o Development and validation of brief measures of positive and negative affect: the
PANAS scales

o Journal of Personality and Social Psychology, 54 (1988), pp. 1063–1070

o View Record in Scopus

Full Text via CrossRef

| Cited By in Scopus (8478)

16.
o Wood et al., 2002
o W. Wood, J.M. Quinn, D.A. Kashy

o Habits in everyday life: Thought, emotion, and action

o Journal of Personality and Social Psychology, 83 (2002), pp. 1281–1297

o View Record in Scopus
 |

Full Text via CrossRef

| Cited By in Scopus (169)

17.
o Young, 2006
o K.W. Young

o Social support and life satisfaction

o International Journal of Psychosocial Rehabilitation, 10 (2006), pp. 155–164

o View Record in Scopus

| Cited By in Scopus (10)