WITHDRAWN: A Social Network Approach to Testing the Effects of a School-Based Lunchtime Mentoring Program on Peer Likeability of Aggressive Children

Introduction. Aggressive children have problematic peer relationships (Rodkin & Berger, 2008; Sentse, Kiuru, Veenstra, & Salmivalli, 2014; Watling Neal, 2010). Developing and maintaining positive peer relations is an important developmental task that has implications for children’s social, emotional, psychological, and academic adjustment. Selective interventions that directly target the peer ecology of children displaying disruptive behaviors may prove to be a valuable adjunct to more formal intervention efforts that target inadequate parenting practices or social skills deficits.

Evidence suggests that youth mentoring can have positive impacts on children across a wide range of psychosocial and educational outcomes (DuBois, Holloway, Valentine, & Cooper, 2002; Dubois, Portillo, Rhodes, Silverthorn, & Valentine, 2011). The current study examines whether a school-based lunchtime mentoring intervention for aggressive children can affect change in peer likeability, and includes a test of innovative likeability trajectory effects using Social Networks Analysis. In particular, we examined whether aggressive children in the Lunch Buddy (LB) Mentoring condition had a higher likelihood of (a) sending and receiving likeability nominations (i.e. nominated or were nominated by classmates as some they liked the most), (b) likeability reciprocation (whether both children named each other as liked most), and (c) likeability maintenance (e.g., does likeability persist) compared to aggressive children in a Waitlist Control condition (WC).

Method. Participants were 342 (181 girls, 159 boys) 2^nd- through 4^th-grade children (125 in 2^nd grade, 133 in 3^rd grade, and 88 in 4^th grade) between the ages of 6 and 10 years old (M age at consent = 8.49 years, SD= .88 years). Fifty-five of these children (28 = LB; 27 = waitlist control) were identified as displaying heightened levels of aggression and were randomized to either the LB condition or a WC condition. Two hundred and eighty-seven children were classroom peers of children participating in the intervention trial. A peer nomination procedure was used to assess likeability (“Who in your class do you like the most?”).

Results and Conclusion. Analyses were conducted using the Simulation Investigation for Empirical Network Analyses program, version 4.0 (RSiena 4.0; Ripley, Boitmanis, & Snijders, 2013). Children in the LB intervention evidenced higher stability in likeability ratings over the academic year compared to children in the WC condition (OR = 2.69, p < .05), and at rates slightly higher than the normative sample (i.e. children not identified as aggressive). Results suggest the LB intervention could be a promising option for improving the peer relations of aggressive children.