آیا سرمایه اجتماعی بر استفاده SNS تاثیر می گذارد؟ نگاهی به نقش بهزیستن ذهنی و هویت اجتماعی

Abstract This study views social capital as a precursor of SNS (Social Network Service) use, which departs from the previous thoughts that considered social capital as the outcome of SNS activities. Drawing upon the theoretical premises regarding network-based social capital, this study examines the roles of subjective well-being and social identity in terms of their moderating as well as mediating influences on SNS use. This study sought to sub-categorize social capital and SNS use with a view to providing more refined theoretical and practical implications. The study’s main objectives are three-fold: First, the study verifies whether social capital, categorized into bridging and bonding capital, influences one’s SNS use as measured by qualitative use and quantitative use. Second, the study aims to confirm whether subjective well-being mediates between social capital and SNS use. Third, it examines whether social identity moderates the relationship between social capital and SNS use. The study result indicates that bridging capital only had a significant impact on qualitative use. However, subjective well-being did not mediate the relationship between social capital and SNS use. Finally, the cognitive identity caused significant difference in the effect of social capital on quantitative SNS use, whereas cognitive and affective identities caused significant differences in regards to the effects of bridging capital on qualitative SNS use.

. Introduction The review of literature on social capital reveals that the majority of research concentrated on social network, trust, civic participation, life satisfaction and others (Coleman, 1988, Newton, 2006 and Putnam, 2000). The core concept of social capital centers around availability of resources that are obtained through social interactions (Putnam, 2000), and a predominant view has been that people with strong ties and diverse networks possess greater social capital than those without them. To maintain a desired level of social capital, one needs a social mechanism that links one with others to build relationships. Today, social network service (here-in-after SNS) provides a social mechanism that allows people to create rational or emotional connections and share knowledge and information online. Thus, the primary motives for joining SNS may be to make social contacts and strengthen ties with friends or acquaintances. According to a study which researched motives for taking part in SNS, people desire to maintain and promote social network and they make investment in social network to increase trust and reciprocal norm, which facilitates group-oriented behavior (Ellison, Steinfield, & Lampe, 2007). Thus, from this viewpoint, social network may be considered a prerequisite to collective behavior through collaborative engagement in issues requiring reciprocal trust (Putnam, 2000). In sum, the human capital which is obtained through trust and network plays a salient role in making people participate in collective behavior. Contrary to the rationale aforementioned, the past research on the relationship between social capital and SNS activity has taken a view that SNS is only one of the tools to promote social capital (Ellison et al., 2007, Ryberg and Larsen, 2008 and Koh et al., 2010), and no studies so far have conceived social capital as a precursor of SNS usage, The previous literature, by and large, conceived that SNS promotes ties among the members with little cost and much convenience so as to help nourish social networks that underpin social capital. And a few studies which viewed social capital as a causal agent to elicit collective behavior mostly focused on civic or political activities that take place in local communities (Howard & Gilbert, 2008). Therefore, it is the main thrust of this study to view SNS not as an instrument to generate social capital but as one of the behavioral manifestations of social capital, This view is consistent with the core premises of social capital theory which viewed civic or political activities as behavioral outcome of social capital (Coleman, 1988, Newton, 2006 and Putnam, 2000). Hence, this research rationale seems tenable since the use of SNS, just like participation in civic activities or political activities, requires voluntary decision which is expedited when one maintains strong ties built on mutual trust, a critical component of social capital. Thus it is a logical extension of this rationale to investigate the role of SNS as a social media by which people demonstrate social capital. Another objective of this study is to explore the role of subjective well-being as a mediator between social capital and SNS use. Past research has demonstrated that social capital produces subjective (or psychological) well-being, and that people with high degree of subjective well-being tend to partake in online community activities through reciprocal adaptation (Helliwell and Putnam, 2004 and Ellison et al., 2007). But no previous research so far reported on whether subjective well-being actually mediates between social capital and SNS use. Verifying the mediating role of subjective well-being will help us to understand whether people with high level of social capital take part in SNS activities more when they are emotionally charged. This emotional linkage has not been fully explored in past research on the outcome of social capital, which makes this study a meaningful addition to current literature. Next, drawing upon the theory of social identity proposed by Dholokia, Bagozzi, and Pearo (2004), this study aims to examine whether collective identity at individual level affects SNS use. When one contemplates on joining a new community, either online or offline, or when one reevaluates the value of continuing current community membership, one normally embarks on mental accounting where one evaluates the value of joining the new community against current membership. Hence, understanding how social identity operates is critical in determining one’s membership into SNS community. Based on this rationale, this study aims to uncover whether social identity actually moderates the relationship between social capital and SNS use. This research focus is expected to elevate the current understanding of the role of social capital that relates to identity-based community membership, which has not been previously researched. The result will also help us to better understand the specific roles of self-identity with regards to SNS use, incorporating three dimensions of social identity (cognitive, affective, and evaluative). The finding will alert social media firms to the importance of understanding specific roles of self-identity as facilitator or inhibitor of SNS use. In sum, it is the main objective of this study to examine the relationship between social capital and SNS use drawing upon existing theories pertaining to the network-based social capital. Also, the study seeks to sub-categorize social capital and SNS use in an effort to provide more specific theoretical as well as practical implications regarding the topical issue. The study has the following three research objectives. First, it adopts the network-based typology of social capital which dichotomizes social capital into bridging and bonding dimensions, and it aims to determine whether each type of social capital has significant relationship with SNS use. In this study, to differentiate behavioral intensity of SNS use, SNS use is further categorized into qualitative use and quantitative use. Second, with a view to integrating previous findings on the relationships among social capital, subjective well-being, and SNS use, this study aims to ascertain whether subjective well-being mediates between social capital and SNS use. Finally, the study seeks to verify whether social identity (cognitive, affective, and evaluative identity) moderates between social capital and SNS use.

Results 4.1. Reliability and validity As a result of explanatory factor analysis based on principal component analysis method, two items from bonding capital were deleted because they produced less than .5 in commonality scores. Cronbach alpha indicated that the main measures had an acceptable level of internal consistency, as alphas were all bigger than .7. Also, factor loadings and AVEs all exceeded .5, indicating that the study measures had both construct validity and convergent validity. The three dimensions of social identity were factor analyzed, resulting in one item from the cognitive dimension being deleted. According to the result, cognitive dimension had eigenvalue (2.54), variance explained (18.42), alpha (.772), and AVE (.798), Affective dimension had three items with eigenvalue (1.82), variance explained (16.34), alpha (.892), and AVE (.860). Evaluative dimension had three items with eigenvalue (1.56), variance explained (15.42), alpha (.875), and AVE (.833) (see Table 1). Table 1. Result of exploratory factor analysis. Eigen-value Var. exp. (cum) Alpha AVE Bridging 2 .823 .096 .014 .033 4.314 12.68 .868 .842 Bridging 4 .775 .098 −.037 −.140 Bridging 3 .738 .119 −.031 −.074 Bridging 5 .721 .163 .198 .208 Bridging 1 .699 .111 .022 .021 Bridging 6 .667 .169 .230 .175 Bridging 8 .599 .165 .215 .046 Bridging 7 .596 .137 .159 .068 Bonding 2 .218 .751 −.051 .027 3.440 22.81 .832 .837 Bonding 7 .109 .719 .030 .108 Bonding 1 .270 .714 −.129 −.010 Bonding 6 .075 .700 .153 .216 Bonding 5 .149 .688 .237 .207 Bonding 3 .177 .673 −.054 .085 SNS use 4 .145 .042 .836 .014 3.220 32.27 .829 .822 SNS use 5 .076 .090 .820 −.099 SNS use 3 .068 −.020 .810 −.034 SNS use 2 .119 −.025 .756 .066 SNS use 1 .093 .013 .609 .051 Sub WB 3 .116 .170 −.117 .810 3.159 41.56 .806 .813 Sub WB 1 .037 .115 .146 .774 Sub WB 4 .174 .188 −.057 .732 Sub WB 5 −.098 .028 .012 .711 Sub WB 2 .022 .064 .010 .696 The bold-faced values represent factor loadings that belong to each extracted factor. Table options The result of correlations analysis with the purpose of testing discriminant validity indicates that bridging capital and bonding capital both had significant correlations with subjective well-being, and subjective well-being had a high correlation with SNS use as well. To test discriminant validity, the square roots of AVE’s were compared with correlation coefficients, The result indicates that AVE’s squared rooted were all greater than coefficients that are located horizontally and vertically, affirming discriminant validity. The result is shown in Table 2. Table 2. Correlations and AVE. Bridging Bonding Sub. well-being SNS use (qualitative) Bridging capital .842 Bonding capital Coeff .421⁎⁎ Sig (p) .000 .837 N 242 Sub. well-being Coeff .170⁎⁎ .316⁎⁎ Sig (p) .008 .000 .813 N 242 242 SNS use (qualitative) Coeff .243⁎⁎ .111 .160⁎ Sig (p) .000 .090 .036 .822 N 242 242 242 ⁎ Sig (p) < .05. ⁎⁎ Sig (p) < .01, Numbers on diagonal line represent AVEs. Table options 4.2. Hypothesis tests 4.2.1. H1 test To test the hypothesis (H1) on the relationship between two types of social capital and SNS use classified into quantitative and qualitative uses, multiple regression analyses were performed. The result shown in Table 3 indicates that bridging capital only has a significant relationship with qualitative use (Beta = .238; t = 3.401; p = .001). Therefore, this result partially supports H1. This finding is not in line with hypothesized relationship which predicted that bridging capital will have positive influence on quantitative SNS use rather than qualitative use, because bridging capital is built on weak ties among members with emphasis on width rather than depth of relations. Hence, this finding may provide new insight on the role of bridging capital on SNS use. Table 3. The effects of social capital on SNS use. Model Dep var: quantitative SNS use Dep var: qualitative SNS use Non-std coeff Beta t Sig (p) Non-std coeff Beta t Sig (p) B SE B SE (Constant) 4.007 .857 4.687 .000 1.578 .604 2.613 .010 Bridging Capital .036 .155 .017 .231 .817 .372 .109 .238 3.401 .001 Bonding Capital .126 .153 .060 .825 .419 .017 .108 .011 .157 .875 Table options 4.2.2. H2 test In order to test the hypothesis (H2) that bonding capital has a more positive relationship with subjective well-being than does bridging capital, multiple regression analysis was performed. The result indicates that bonding capital has a significant relation with subjective well-being (β = .296; t = 4.374; p = .000), whereas bridging capital does not have a significant relation (Beta = .046; t = .674; p = .501). Therefore, H2 is partially supported. This result is in support of previous research which verified significant relationship between strong tie and psychological well-being ( Valkenburg et al., 2006) (see Table 4). Table 4. The effects of social capital on subjective well-being. Model Non-std. coeff Beta t Sig (p) B SD (Constant) 1.940 .446 4.351 .000 Bridging capital .054 .081 .046 .674 .501 Bonding capital .351 .080 .296 4.374 .000 Dep. var = subjective well-being. Table options 4.2.3. H3 test Next, to test hypothesis (H3) which proposed the mediating role of subjective well-being between social capital and SNS use, the prerequisite requirements for the presence of mediator effect were examined as proposed by Baron and Kenny (1986). The preconditions for mediating effect require that the mediator as well as predictor variable have significant impact on outcome variable and mediator be significant in its effect on outcome variable (Baron & Kenny, 1986). But the result of testing for this relational significance revealed that some of the relationships did not fulfill the prerequisite requirements. Thus the subject well-being did not mediate between social capital and SNS use. Therefore, the result does not support H3. The result indicates that the relationships between social capital, subjective well-being, and SNS use are such that there are not coherent direct and indirect effect (which serves as a proxy for the presence of mediating effect) among the three variables. The result of testing for the mediating effects is shown in Table 5. Table 5. Test of mediating effects of subjective well-being. Predictor variable Outcome variable Quantitative SNS use Qualitative SNS use Beta t Beta t Sample size 242 242 Test 1: Bridging capital →→ subjective well-being →→ SNS use Base model Bridging capital a.090 1.396 a.243 3.824⁎⁎ Full model Bridging capital b.079 1.201 b.238 3.693⁎⁎ Subjective well-being .067 1.023 .031 .480 Reduction rate (a − b)/a 12.22% 2.54% Test 2: Bonding capital →→ subjective well-being →→ SNS use Base model Bonding capital a.074 1.145 a.111 1.700 Full model Bonding capital b.054 .793 b.099 1.441 Subjective well-being .063 b.930 .038 .554 Reduction rate (a − b)/a 27.02% 10.81% ∗ p < .05. ⁎⁎ p < .01. Table options 4.2.4. H4 test To test hypothesis (H4) proposing the moderating role of social identity in relationship between social impact and SNS use, sub-group analysis was performed, using median values as the cut-off scores for high and low levels of social identity (cognitive dimension = 4.33; affective dimension = 5.02; evaluative dimension = 5.10). The regression analysis shown in Table 6 revealed that for quantitative SNS use, the only significant differences were found in cognitive identity where bridging capital was influential on quantitative SNS use at high cognitive level, and bonding capital was influential at low cognitive level. This result provides partial support to H4 such that cognitive identity caused significant difference in the effects of bridging as well as bonding capital on quantitative SNS use. Table 6. The effects of social capital on SNS’s quantitative use by the type and level of social identity. Indep. var. Cognitive identity Affective identity Evaluative identity High Low High Low High Low Beta t Beta t Beta t Beta t Beta t Beta t (Const.) 3.88⁎ 4.01⁎⁎ 4.05⁎ 2.17⁎ 6.13⁎⁎ 2.28⁎ Bridging capital .18 2.01⁎ −.12 −1.12 .03 −.30 .17 1.43 −.01 −.17 .053 .49 Bonding capital .09 1.07 .24 2.23⁎ .02 .25 .06 .53 −.19 1.91 .177 1.64 Dependent variable = quantitative SNS use. ⁎ Sig at 0.05 level. ⁎⁎ Sig at 0.01 level. Table options With regard to qualitative SNS use, two dimensions of social identity-cognitive and affective-caused significant differences in the effects of social capital on SNS use. Specifically, at high cognitive level, bonding capital was influential on SNS use (Beta = .33, t = 3.77), whereas at high affective level, bridging capital was influential (Beta = .28; t = 2.83) as shown in Table 7. Table 7. The effects of social capital on SNS’s qualitative use by the type and level of social identity. Indep. var. Cognitive identity Affective identity Evaluative identity High Low High Low High Low Beta t Beta t Beta t Beta t Beta t Beta t (Const.) 2.74⁎⁎ 1.48⁎ .75⁎ 2.85⁎⁎ 2.11⁎ 1.90⁎ Bridging capital .33 3.77⁎⁎ .09 .86 .28 2.83⁎⁎ .09 .78 .25 2.62 .11 1.03 Bonding capital 1.12 −1.43 .14 1.28 .02 .26 .02 .22 −.12 −1.22 .16 1.51 Dependent variable = quantitative SNS use. ⁎ Sig at 0.05 level. ⁎⁎ Sig at 0.01 level.