Predictors of Behavioral and Relationship Outcomes During Mutual Self-Disclosure: Implications for Dyadic Peer Support Interventions

Despite the popular use of peer support interventions to improve the interpersonal functioning of patients with mental illness, research into the effectiveness of such programs has returned mixed results. In order to better understand potential moderators of these interventions’ effectiveness, the present study involved the random pairing of two same-sex strangers who took turns engaging in mutual self-disclosure on two separate occasions, separated by one week. A total of 51 same-sex pairs of college students were enrolled in this study. Our aim was to examine whether social competence and existing levels of social support predicted engagement in the interactions, specifically self-disclosure and support provision. Although both social competence and social support predicted engagement in the interaction, social competence appeared to be the more robust predictor. Dyadic data analyses revealed that providing emotional and instrumental support to a partner predicted an increase in one’s liking for and plans to communicate with this partner. The association between emotional support and relationship outcomes was stronger among males than females, whereas the association between instrumental support and relationship outcomes was stronger for females than males. PREDICTING MUTUAL SELF-DISCLOSURE OUTCOMES 3 Predictors of Behavioral and Relationship Outcomes During Mutual Self-Disclosure: Implications for Dyadic Peer Support Interventions Individuals suffering from moderate to severe mental illness typically experience difficulty building and maintaining social relationships (Davidson, O’Connell, Tondora, Staeheli, & Evens 2004; Mueser & Tarrier, 1998). As a result of this interpersonal deficit, people suffering from a variety of mental disorders consistently report feelings of social isolation, loneliness, and a lack of much needed social support (Bengtsson-Tops & Hansson, 2001; Davidson, Stayner, Lambert, Smith, & Sledge, 1997; Green, Hayes, Dickinson, Whittaker, & Gilheany, 2002). Problematic social relationships have also been shown to have deleterious effects on an individuals’ physical and mental health (Cohen, 2004; Cohen, Sherrod, & Clark, 1986; Windle, Francis, & Coomber, 2011). Thus, establishing social relationships and improving interpersonal functioning is often considered a primary goal of recovery-oriented care (Davidson, O’Connell, et al., 2004). One often-utilized method to accomplish this goal is the use of dyadic or group peer support interventions. Unfortunately, research into the effectiveness of such programs has shown mixed results (Chinman et al., 2014; Davidson et al., 1999; Fuhr et al., 2014; Lloyd-Evans et al., 2014; Repper & Carter, 2011). In attempt to reconcile conflicting research, specific moderators of the effectiveness of peer support interventions have been examined to determine who does and who does not benefit from individual and group supportive interactions. The two primary moderators of interest for this study are social competence and existing levels of social support. Although specific operationalizations of social competence vary across the research literature, there is a general consensus that social competence is defined as real or perceived effectiveness in navigating social interactions (Rose-Krasnor, 1997). Higher levels of this PREDICTING MUTUAL SELF-DISCLOSURE OUTCOMES 4 construct have been linked with better social, emotional, and academic adjustment (Chen, Huang, Chang, Wang, & Li, 2010; Kupersmidt, Coie, & Dodge, 1990; Lengua, 2003; Newcomb, Bukowski, & Pattee, 1993; Parker et al., 1995) and have been positively associated with the development of social networks (Hansson, Jones, & Carpenter,1984; Heller & Swindle, 1983; Monroe & Steiner, 1986; Shaver, Furman, & Buhrmester, 1985). Additionally, social competence has been shown to prospectively predict friendship formation and increases in perceived support (Ciarrochi, Scott, Deane, & Heaven, 2003; Cohen, Sherrod, & Clark, 1986). Given these associations, social competence may play an important role in the establishment of new interpersonal relationships and supportive interactions. While research examining social competence specifically as a pre-requisite for the beneficial effects of support interventions is limited, the results of a dyadic peer support intervention that paired mentally ill patients with healthy peer supporters found that people who displayed lower levels of social functioning showed lower levels of study adherence, which was assessed by how often patients met with their peer supporters and engaged in activities together. Although instructed to meet on a regular basis and given a small monthly stipend to provide the financial opportunity to engage in community activities, patients with lower social functioning refused to continue meeting with their partner. They did, however, continue with study follow-up assessments for the duration of the experiment (Davidson et al., 2004). Thus, people who display lower levels of social competence may lack some of the personal resources needed to establish and maintain supportive relationships with others and therefore are less likely to reap the benefits of a peer support intervention. A second variable that may influence social interactions and the effect of peer support interventions is existing levels of social support. Social support has been positively associated PREDICTING MUTUAL SELF-DISCLOSURE OUTCOMES 5 with better mental and physical health (Berkman, 1995; Cohen & Janicki-Deverts, 2009; Uchino 2004; Umberson & Montez 2010). The relation of existing social support to peer support intervention outcomes, however, has appeared to be paradoxical. Even though social competence is positively associated with social support, and social competence should predict benefits from peer support interventions, high levels of social support have been found to be a disadvantage in terms of benefits from peer support groups. In a study of women with breast cancer, peer discussion groups were helpful for women who lacked social support, but were associated with a deterioration in functioning for those with high levels of outside social support (Helgeson, Cohen, Schulz, & Yasko, 2000; Helgeson, Cohen, Schulz, & Yasko, 2001). Existing social support might also predict whether someone chooses to participate in a support intervention. One study found that those who attended a support group had lower levels of existing support (Taylor, Falke, Shoptaw, & Lichtman, 1986), suggesting that individuals may be motivated to participate in support interventions when they are currently lacking supportive relationships. Taken together it seems that existing social support may influence engagement in a peer support intervention, and ultimately, the outcomes of peer support interventions. Although this study does not examine the effects of support interventions on interpersonal and psychological health outcomes, this study does examine how social competence and existing social support influence relationship formation–specifically initial engagement in self-disclosure and support providing behaviors that are needed for the successful establishment of relationships. In order to better understand the links of social competence and social support to relationship formation, it is helpful to draw upon the capability, motivation, and opportunity model of behavior (COM-B; Figure 1). In this system capability (one’s psychological or physical capacity to engage in a behavior), motivation (one’s internal processes PREDICTING MUTUAL SELF-DISCLOSURE OUTCOMES 6 that energize and direct behavior), and opportunity (one’s external factors that make a behavior possible) cooperate to generate behavior (Michie, Stralen, & West 2011). In the context of the present study which involves an interaction between two strangers, the interaction itself provides the opportunity for interpersonal engagement, social competence reflects one primary aspect of an individual’s capability to engage, and existing social support influences one’s motivation to engage (Figure 2). Capability and motivation then lead to actual engagement, which is operationalized as the quantity of self-reported self-disclosure and support provision. The two engagement processes just mentioned, self-disclosure and support provision, have been shown to be vital for the initial stages of relationship formation. Self-disclosure is conducive to increased liking and closeness between partners (Sprecher & Duck, 1994) and is predictive of increased positive affect in get-acquainted interactions between same-sex strangers (Vittengl & Holt, 2000). Providing social support has also been shown to improve mood and enhance self-esteem (Williamson & Clark, 1989), and in the context of relationships, support provision is associated with higher relationship and life satisfaction (Gosnell & Gable, 2015). To test these ideas, the present study involved the random pairing of two same-sex strangers who were brought into the lab for two one-hour sessions, separated by one week. In each session, participants took turns sharing information about a personal problem with one another. Face to face interactions were used as opposed to online or telephone interactions, because they tend to produce higher affiliative outcomes (Sprecher, 2014) and are often used in real-life social support interventions. Four hypotheses were developed. First, higher social competence will predict greater amounts of self-disclosure and support provision. Second, higher levels of existing social support will predict lower levels of self-disclosure and support provision. Third, the effect of social PREDICTING MUTUAL SELF-DISCLOSURE OUTCOMES 7 competence and social support on these two engagement processes will be mediated by increased and decreased motivation, respectively. That is, people who have higher levels of social competence and lower levels of social support will be more motivated to engage in selfdisclosure and support provision. Such increases in motivation will then lead to higher selfdisclosure and greater support provision. Finally, higher levels of behavioral engagement will result in positive relationship outcomes, specifically partner liking and plans to communicate with their partner after the completion of the study. Although other relations between interpersonal variables may exist and are shown in our model in Figure 2, we are only testing the relations just mentioned in the present study. These relations are shown in bold.