Does Peer-Mediation Increase Effectiveness of the PEERS Curriculum? A Pilot Randomized Control Trial Examining Social Function, Well-Being and Classroom Social Network Centrality

Background: PEERS (Laugeson & Frankel, 2010) is one of the most well-studied social skills interventions for intellectually-able adolescents with autism (ASD). However, the potential for increasing effectiveness using peer mediation strategies remains unexamined. Also, the majority of previous research has relied on questionnaires to the exclusion of more objective outcome measures (Hood et al., 2017).

Objectives: (1) To compare change in adolescent- and parent-reported social functioning and well-being from pre-to-post intervention and at a 4-month follow-up among participants who completed PEERS, a peer-mediated PEERS model (PMP), or a delayed treatment control (DTC) group. (2) To compare classroom social network centrality (SNC; Farmer & Rodkin, 1996) among study groups at immediate and 4-month follow-ups.

Methods: Participants were 34 intellectually able (IQ ≥ 70) adolescents with ASD randomly assigned to PEERS, PMP, or DTC (see Table 1); one parent of each adolescent; 16 typically developing peer mentors, and 562 general education (GE) classmates of participants with ASD. PEERS and PMP were administered identically except for the peer-mediation strategies of proximity and peer initiation (Odom & Strain, 1984) in PMP using a 1:1 ratio of typically developing peer mentors. Questionnaires were collected pre- and post-treatment, and at a 4-month follow-up. SNC data were collected post-treatment and at a 4-month follow-up in a subsample whose schools approved research procedures. Interviews were conducted with GE classmates in one classroom of each participant with ASD to generate SNC scores (0-Isolated; 1-Peripheral; 2-Secondary; 3-Nuclear).

Results: Reported in Table 1, MANOVAs indicated greater improvement at immediate follow-up in parent-reported social skills and problem behaviors, adolescent-reported social skills knowledge, and loneliness in PMP relative to DTC. PEERS participants demonstrated greater improvements in adolescent-reported social skills knowledge, loneliness, and hosted get-togethers relative to DTC. Improvements in social skills, problem behaviors, and social skills knowledge were maintained by PMP at the 4-month follow-up, as were improvements in social skills knowledge and adolescent-reported get-togethers by PEERS participants. Notably, number of actual get-togethers did not differ between PEERS and PMP at either follow-up. Table 2 reports SNC scores. When PEERS and PMP were collapsed into one treatment group, Fisher’s exact tests indicated significantly less isolation in treatment participants relative to DTC at the immediate follow-up (p = .05), but not at the 4-month follow-up (p = 0.56).

Conclusions: Findings suggest a modest improvement parent-reported social skills and problem behaviors when PEERS was implemented using peer mediation strategies. When PEERS and PMP were collapsed into one treatment group, treatment participants were less isolated in their classrooms relative to DTC at the immediate follow-up, but these levels of engagement were not maintained. SNC findings provide initial evidence for generalization of acquired skills to the classroom that requires replication with a larger sample. Not a single participant with ASD received a nuclear rating, and very few were rated as secondary. This suggests the need for additional efforts to improve peer acceptance of high school students with ASD in GE classrooms. To our knowledge, this is the only study to examine SNC of PEERS graduates.