106. Adolescent Social Networks and School Engagement Across the High School Transition

ABSTRACT

Purpose: Attending and meaningfully engaging in school is a critical component of adolescence with broad implications for future health and well-being. Social network analysis offers powerful techniques to examine how the quality and structure of social relationships affect adolescent health and education. Peer and adult relationships impact adolescent health-affecting behaviors such as substance use and violence involvement, with potential for beneficial or deleterious effects. Social networks may similarly confer risk or protection for school engagement, an urgent issue of adolescent health considering COVID-19-related school disruptions. We describe adolescent social network quality and structure and examine associations between social network characteristics and school engagement during the transition from middle to high school. Methods: We analyzed data from an ongoing randomized trial of the Advancement via Individual Determination (AVID) college readiness program across 5 public high schools in Southern California. Participants completed baseline surveys at the end of 8th/beginning of 9th grade and follow-up surveys at the end of 9th grade. The Student Engagement Instrument assessed school engagement and egocentric social network surveys assessed adolescent social network quality (e.g., network composition, peer behavior) and structure (e.g., centrality, density). Multilevel mixed-effects models examined associations between baseline social network characteristics and concurrent and future school engagement. Models were adjusted for gender, ethnicity, prior AVID participation, study arm, and baseline school engagement as appropriate. Results: Baseline participants (n=431) were 60.6% female and 82.8% identified as Latinx. 418 participants completed follow-up surveys;retention rate=97.0%. Mean school engagement was 3.26 at baseline (SD=0.53;range 1-4) and 3.14 at follow-up (SD=0.52). Across all individuals in participants' social networks, 68.7% were peers, 24.5% were adults, and 55.4% were identified as sources of support. Across all peers in participants' social networks, 9.1% engaged in substance use and 82.6% were highly engaged in school. Overall social network characteristics associated with concurrent school engagement included a greater proportion of coaches (β=3.01;95%CI 0.01-6.02) and sources of support (β=0.20;95%CI 0.03-0.37). Peer social network characteristics associated with concurrent school engagement included a greater proportion of peers highly engaged in school (β=0.40;95%CI 0.20-060), a lower proportion of peers who use substances (β=-0.25;95%CI -0.50-0.00), and lower peer network density (β=-0.40;95%CI -0.72–0.08). Overall social network characteristics associated with future school engagement included a greater proportion of teachers (β=0.94;95%CI 0.05-1.82) and sources of support (β=0.28;95%CI 0.11-0.44). Peer social network characteristics associated with future school engagement included a greater proportion of peer sources of support (β=0.17;95%CI 0.03-0.32) and a lower proportion of peers who disrupt class (β=-0.29;95%CI -0.56–0.03). Conclusions: Building or bolstering connections to supportive peers and adults may serve to enhance adolescents' school engagement. Similar to other adolescent health-affecting behaviors, school engagement may be positively or negative influenced by the behaviors of peers in adolescents' social circles. Network-based interventions that target multiple associated social network-mediated behaviors, such as school engagement and substance use, may be particularly beneficial to adolescent health. Sources of Support NIH/NICHD (T32HD071834), UPMC Children's Hospital of Pittsburgh Scholar Award, NIH/NIDA (1K23DA040733-01A1), Robert Wood Johnson Foundation (E4A 74086).