Group-Delivered Video Model Intervention Package Improves Social Skills in Adults with Autism Spectrum Disorder

Saturday, May 13, 2017: 12:00 PM-1:40 PM
Golden Gate Ballroom (Marriott Marquis Hotel)
E. S. Brodkin1, A. A. Pallathra2, J. Day-Watkins3 and J. E. Connell3, (1)Department of Psychiatry, University of Pennsylvania, Philadelphia, PA, (2)Department of Psychiatry, Catholic University of America, Washington, DC, (3)Drexel University, Philadelphia, PA
Background: Video modeling (VM) is an evidenced-based intervention proven effective in improving social skills in children and adolescents with autism spectrum disorder (ASD) (Shukla-Mehta et al., 2010; Day-Watkins et al., 2014). Few reports have been published on the development of VM services suitable for adults beyond their early twenties (Cappadocia & Weiss, 2011; Williams et al., 2007). As part of a pilot study of a new cognitive behavioral treatment program, Training to Understand and Navigate Emotions and Interactions (TUNE In), we developed a VM package to train social skills to adults with ASD.

Objectives: To use VM, role-play, and feedback to teach 13 adult participants with ASD four fundamental social skills. The goal of the study was to extend the VM literature to teaching specific social skills to adults with ASD to examine the efficacy and feasibility of delivering the social skills interventions to an adult population and in community behavioral health settings.

Methods: Across eight weekly sessions, participants were taught to engage in four different social skills (e.g. approach a group (skill 1); don’t approach a group (skill 2); greet a group (skill 3); and initiate conversation (skill 4)). During sessions facilitated by behavioral coaches, participants role-played with actors (i.e. undergraduate research assistants) on five exemplars of each social skill. The actor engaged in scripted scenes (discriminative stimuli) of each social skill exemplar, setting the occasion for the participant to respond by either demonstrating or not demonstrating the appropriate skill. Baseline data on responses was collected in week 1. In week 2, skills 1 and 2 entered treatment, while skills 3 and 4 remained in baseline. During treatment, if participants responded incorrectly, they were given a video model depicting an individual correctly engaging in the target skill. After access to the video model, participants repeated role-play trials, and feedback was provided on steps completed correctly and incorrectly. In week 3, training on skill 3 began and training on skill 4 began in weeks 4. Generalization probes were interspersed between training trials. Training on each social skill was staggered across four weeks to assess for the emergence of correct responding to the generalization probes, which had not yet received training with a video model. Paired t-tests were used to compare participant scores from final baseline measurements vs. week 8 measurements.

Results: All participants’ scores increased from final baseline measurements to week 8 (Figure 1). (Skills 1-3 p<0.001; Skill 4 p<0.003).

Conclusions: Participants demonstrated a significant increase in correct responses for each social skill after receiving this VM intervention package. These data demonstrate the effectiveness of VM of social skills across multiple participants and multiple skills. This group delivery model is critical for the dissemination of behavioral and mental health services to sites where large consumer populations and limited resources don’t allow for the delivery of intensive, individualized intervention to individuals with ASD.