Subjective Beliefs about Social Skills Importance, but Not about Social Skills, Predict Peer Interactions in Adolescents with Autism Spectrum Disorder

Background: Adolescents with autism spectrum disorder (ASD) exhibit frequent deficits in peer interactions (Rao et al., 2008). Recent work has highlighted that such adolescents’ beliefs about their own social skills may importantly differentiate them from their typically developing peers (Lerner et al., 2012). Such beliefs can be broken down into two components: beliefs about what skills one actually uses (social skills behaviors; Gresham, 1986), and beliefs about whether those skills are important (social skills importance; Gresham & Elliot, 1990). Adolescents with ASD often believe social skills are important (Rankin et al., 2015; McMahon & Solomon, 2015), and over-report their use of social skills behaviors (Lerner et al., 2012). However, little work has examined the relative contribution of these beliefs in predicting actual peer interaction in adolescents with ASD.

Objectives: This study sought to investigate the relationship between self-reported social skills importance ratings, self-reported social skill behaviors, and observed unstructured peer interactions. It was hypothesized that high levels of self-reported beliefs about social skills behaviors would exhibit an additive relationship in predicting peer interactions among adolescents with ASD.

Methods: 25 Adolescents (M_age= 14.984, SD_age= 1.480, 19 male) with ADOS-2 (Lord et al., 2012) confirmed ASD diagnoses were assigned to social groups of 5-9 peers for one video recorded twenty-minute free interaction session (e.g., Lerner & Mikami, 2012). Peer interaction duration was coded by a team of blinded and reliable (ICC = .79) coders via the Social Interaction Observation System (SIOS; Bauminger, 2002). Participants completed a self-report measure of both social skill behaviors and social skill importance (SSIS; Gresham and Elliot, 2008).

Results: Multiple linear regression determined social skills importance ratings negatively predicted the duration of time spent interacting with peers, even when controlling for self-reported overall social skill behavior (b = -0.162, p = 0.006). This relationship was driven by high importance ratings for communication (b = -1.505, p = 0.002), cooperation (b= -.871, p = .015), engagement (b = -8.55, p = 0.005), and self-control (b = -0.794, p = 0.019). Self-reported social skill behavior was not a significant predictor of peer interactions (p = 0.251), nor was the interaction between such behavior and social skills importance (p = .73).

Conclusions: Contrary to hypotheses, high social skills importance ratings predicted fewer social interactions with peers, regardless of self-reported social behavior. It is possible that adolescents with ASD may place higher importance on areas where they need the most improvement (McMahon & Solomon, 2015) – that is, they may in fact identify deficits by labeling them as important. The four significant subdomains of importance ratings (e.g. communication, cooperation, engagement, and self-control) are all specific to interactions (Meier, DiPerna, & Oster, 2006), which may provide a window into how youth with ASD perceive their social behavior, even as they rate themselves to have greater social skills (Rankin et al., 2015). This finding reveals a complex and unexpected way in which social cognition may relate deferentially to social behavior in adolescents with ASD.