From Fast Friends to Close Friends? The Impact of Interpersonal Closeness Tasks on Adolescent Friendships

• In Event: 3-043 - From Making to Influencing Friends: Using Social Network Analysis to Guide, Deliver, and Evaluate Peer Interventions

Integrative Statement

Early adolescence is a developmental period marked by the importance of peer relationships (Buhrmester & Furman, 1986). The literature clearly documents that having a friend is a protective factor for youth (Vitaro, Boivin, & Bukowski, 2009), yet little research has examined how to promote the development of friendships during this critical time. As such, the purpose of the present study was to evaluate the effectiveness of an intervention program designed to build friendships among middle school peers. Given the growing diversity of the K-12 population and the need for research that demonstrates how to foster positive intergroup relations, this study also sought to examine whether a friendship-building program would be equally effective for same- and cross-race peers.

Participants in the study were 298 seventh and eighth grade students (51% male) attending a public middle school in a small Midwestern community. The racial/ethnic composition of the sample was 51% White, 33% Latino/Mexican American, 10% Multiethnic/Biracial, 3% Asian, 1% African American/Black, and 2% Other. At the beginning of the fall semester, participants completed a survey that included sociometric ratings of their classmates (e.g., “How well do you know…?”) as well as peer nominations of their same-grade friends. Following the survey, participants were assigned a same-gender intervention partner (a.k.a. “fast friend”) and control partner who they had reported not knowing well and had not nominated as a friend. Once a month for the next three months, fast friend partners participated in tasks designed to induce interpersonal closeness (see Page-Gould, Mendoza-Denton, & Tropp, 2008). Approximately two weeks after the final task, participants completed the survey again.

Mixed ANOVAs were used to examine mean differences in sociometric ratings of same- and cross-race partners before and after the closeness tasks. Following the closeness tasks, participants rated their fast friend partners significantly higher than control partners on how well they knew them (F(1, 226=67.74, p<.001) and how much they considered them a friend (F(1, 223)=26.66, p<.001). There were no significant differences between same- and cross-race partners based on partner type (intervention or control) or across time (see Figure 1).

Exponential random graph models (ERGMs) were then used to estimate the odds of participants nominating their partner as a friend. ERGMs operate like logistic regression but account for structural features of networks (e.g., indirectly connected ties) that may influence opportunities for network ties such as friendships. Since participants were only allowed to nominate same-grade peers as friends, a separate ERGM was conducted for each grade. For seventh graders, the ERGM indicated that being fast friend partners was not associated with a significant increase in the odds of nominating a peer as a friend. For eighth graders, however, being fast friend partners was a significant predictor of a friendship nomination (η=.25, p<.05). There was no difference in odds between same- and cross-race partners, providing further support that the intervention was equally effective for promoting both same- and cross-race friendships.

Subsequent analyses will explore particular network structures amenable to friendship formation that could provide a starting point for future interventions.

Authors

• Leslie Echols, Missouri State University
  Presenting Author

• Jerreed D Ivanich, University of Nebraska at Lincoln
  Non-Presenting Author