The Effects of Social Skill Interventions for Adolescents With Autism: A Meta-Analysis

Abstract

Communicating with peers often poses challenges for individuals with autism spectrum disorder (ASD); these social interactions, however, can be especially difficult during the teenage years. For many adolescents with ASD, peer interaction is often limited or nonexistent, even for those taught in general education settings. The purpose of this meta-analysis was to investigate social interaction interventions for adolescents with ASD within public school settings, with a focus on studies that utilized single-case design. Studies were evaluated per the What Works Clearinghouse guidelines. This meta-analysis includes 24 studies with participants with ASD ranging in ages from 13 to 21 years. The effects of social skill interventions were evaluated using visual analysis, Tau- U, and a parametric effect size ( d-Hedges–Pustejovsky–Shadish [DHPS]). Combined effect sizes were calculated and compared. Results provide evidence that interventions can produce positive change in social behavior of adolescents with ASD. Effect sizes were generally large. Interpretations and implications are discussed.

Meta-Analytic Methods to Detect Publication Bias in Behavior Science Research

“…Methods used to detect asymmetry and publication bias are routinely employed in meta-analyses of group-design research. Although used less often in behavior analysis research compared to other fields, these methods are used in metaanalyses of behavioral research (see e.g., Amlung et al, 2016;McCormack et al, 2019) and meta-analyses of SCED studies (see e.g., Babb et al, 2020;Bowman-Perrot et al, 2014;Garwood et al, 2021;Shin et al, 2020). When attempting to aggregate behavior science research using frequentist statistical analysis, heterogeneity should be accounted for in the model minimally at the participant level (level 1) and the study level (level 2), referred to as a multilevellevel model meta-analysis (Becraft et al, 2020;Moeyaert et al, 2020).…”

Between-case standardized mean differences: Flexible methods for single-case designs

“…Thus, use of BC-SMDs can enrich research syntheses and evidence-based practice reviews by allowing for inclusion of evidence from SCDs that might otherwise be omitted (Shadish et al, 2015). Due to these advantages, BC-SMD effect sizes have been applied in several recent meta-analyses of single-case research (e.g., Babb et al, 2021;Losinski et al, 2017;Shin et al, 2020). Furthermore, the What Works Clearinghouse recently adopted the BC-SMD as a metric for summarizing findings from SCD studies (What Works Clearinghouse, 2020).…”

Peer Interactions of School-Aged Girls on the Autism Spectrum: A Scoping Review

“…They also encountered significant levels of victimisation and worries associated with victimisation (Ashburner et al, 2019; Sterzing et al, 2012). The importance of peer interactions for school-aged students on the autism spectrum is reflected in the attention given to skills for enhancing peer interactions within evidence-based social skills and peer-mediated interventions (Babb et al, 2021; Bene et al, 2014).…”