Page 119 - Never Too Far Away? The Roles of Social Network Sites in Sojourners’ Adjustment

With covariates included in the model. Next, we ran a model (Model 2) with the following covariates added to the baseline model: Gender, age, length of relationship, and each communication frequency with partner using phone, SNS, email, instant messaging, video calling, and Twitter and Instagram. The results showed poor fit to the data (see Table 2). To improve the fit, we tested a modified model (Model 3) with only video calling (i.e., Skype) as a covariate and excluded all others with non-significant paths. The ensuing model had a satisfactory fit (χ2 = 410. 74; χ2/df = 1.73; CFI = 0.94; TLI = 0.91; RMSEA = 0.06; SRMR = 0.06). The model also showed metric and scalar invariance (see Table 2).
With the inclusion of video calling in the model, the pattern of significant and non-significant direct and indirect causal paths in the model for LDRR and GCRR did not differ from Model 1 (see Table 3). The direct association between Facebook use and perceived relationship stability became more significant when video calling was accounted for (βLDRR = -.25, p = .024), distilling the negative effect of Facebook use on relationship stability in LDRR. The results also showed that the frequency of communication with partner via video calling had opposite associations with relationship quality for LDRR and GCRR. In LDRR, video calling was positively associated with relationship satisfaction (βLDRR = .22, p = .004) and stability (βLDRR = .26, p < .000). In GCRR, it was negatively associated with relationship satisfaction (βGCRR = -.14, p = .014).5
Discussion
Earlier studies using a social network perspective have shown that romantic relationships benefit from offline social network support (Cox et al., 1997; Felmlee, 2001). This current study adds to existing literature by showing that network support could be extended and experienced on SNSs. We examined the proposition that SNSs use predicts access to SNSs network support which, in turn, predicts relationship quality in LDRR and GCRR. We focused on Facebook, a very popular SNS. The results showed that Facebook use per se was negatively associated with relationship quality. However, it was when users accessed relationship support from their Facebook networks that Facebook use indirectly contributed to positive
5 Due to the relatively small sample in LDRR, we validated our findings by conducting a series of simple mediation analyses using Hayes’ PROCESS (Hayes, Montoya, & Rockwood, 2017). We found similar patterns of results.
SNS and Relationship Support 117