Improved Social Motivation in Adolescents with ASD Following a Social Skills Intervention

Background:

Despite increases in research examining the efficacy of social skills interventions for individuals with autism spectrum disorders (ASDs), few studies have targeted adolescents. This is particularly problematic as adolescence is a time of significant social change with growing emphasis placed on peer relationships.

 Objectives:

To gain increased understanding of the clinical features of individuals who might benefit most from a given intervention will have obvious benefit to families and providers. It was hypothesized that individuals who were more mildly impaired in regards to ASD symptoms would show more significant treatment effects compared to moderately and/or severely impaired individuals. Finally, additional differential treatment effects by internalizing and externalizing symptoms were explored in post hoc analyses.

Methods:

 The Multi-Media Social Skills Project was developed to help teens with ASDs acquire better social conversational abilities. The model utilized components of social skills interventions which have strong empirical support as established through previous work, namely group instruction and video modeling. Twenty-one adolescents with an ASD were recruited for this study. Participants were grouped into four cohorts and each received a manual-based, 12-week social skills intervention. Each session lasted three hours. The first 90 minutes were devoted to group instruction of new skills using video modeling. The remaining time was spent participating in a digital photography class with typically developing peers where the participants and peers worked collaboratively. Videotaped samples of five minute unstructured conversations between each participant and a novel peer were obtained pre- and post-intervention, as well as at a three month follow-up. These samples were coded for social behaviors and fluencies (e.g., amount of silence, eye contact). Additional measures were obtained at each evaluation point including the Social Responsiveness Scale (SRS) and the Strengths and Difficulties Questionnaire (SDQ).

 Results:

 One-factor repeated measures ANCOVAs were conducted to compare the effect of the intervention on behavioral observations over three time points. Cohort was entered as a covariate in all analyses. Post hoc Bonferroni t-tests for significant effects were conducted to identify differences between pre- and post-test means, as well as post-test and follow-up means. There was a significant between-subjects effect for level of social impairment (F = 7/34, p < .01). The first hypothesis was supported; between-subject effects for severity of ASD symptoms were found in relation to significant treatment effects for amount of silence. Additionally, although externalizing symptoms did not have a moderating effect on outcome, anxiety was found to significantly moderate behavioral outcomes.

 Conclusions:

Degree of social impairment, as well as level of anxiety, needs to be considered when designing and implementing social skills interventions. Individuals with moderate social skills impairments and anxiety demonstrated the most significant change with this intervention. Individuals possessing more awareness of their social skill deficits and experiencing higher anxiety and discomfort in social situations might be better able to learn from video modeling and more likely to try new skills with peers during generalization experiences. Anxiety in social situations may be an important determining factor in overall efficacy of social skills interventions.