Improvements to Subjective Social Impression Ratings Associated with Participation in the START Social Skills Program for Adolescents with ASD

Background: Social competence in adolescents with ASD has been implicated in a variety of favorable outcomes (Carter et al, 2014).  Current social skill group interventions have primarily relied upon parent-report surveys to gauge improvement (Miller, Vernon, Wu, Russo, 2014).  These instruments, while useful, appear to have several noteworthy limitations (including inherent reporting bias).  A multimodal assessment of social competence may be warranted to more fully capture all of the interrelated factors of interpersonal success (Reichow, Steiner, Volkmar, 2012). One promising method, subjective ratings by naïve observers, may help us to better understand the components contributing to the formation of positive social impressions.  Such data, when combined with video-coded social skill behaviors, parent and adolescent survey measures, and real-world interaction data, may yield a more robust social competence profile for adolescents participating in ongoing social skill intervention efforts.

Objectives: The current study reports on the utility of using subjective social impression ratings within a repeated-measure/randomized controlled trial (RCT) social intervention project for adolescents with ASD.

Methods:   Participants were 17 adolescents with ASD (ages 12-17) enrolled in an ongoing RCT using the Social Tools And Rules for Teens (START) 20-week social skills curriculum.  The adolescents completed a weekly 2-hour session consisting of individual therapeutic check-in sessions, an unstructured socialization period with typically developing (TD) high school and undergraduate facilitators, a structured social activity, discussion and practice of a social skills topic, and an individual check-out session with their parents. At a pre-intervention time point and after every 5 sessions of the program, adolescents participated in two 5-minute videotaped conversations with unfamiliar undergraduates who were unaffiliated with the project and unaware of its hypotheses. These video clips were subsequently shown to a group of naïve observers, who were asked to rate both of the on-screen conversational partners along a number of social dimensions (social skills, comfort, awkwardness, perceived relationships). Additionally, these video clips were coded by trained raters for the total frequency of (a) interpersonal questions asked and (b) overall verbal output. Finally, parent and self-report surveys (i.e. SSIS, SRS-2, and an unpublished social competencies and motivation scale) were collected at each time point for comparative analyses.

Results: As expected, all participants with ASD received significantly lower overall subjective social ratings than their TD conversational partners at the pre-intervention time-point.  While participants varied in their total change scores over the course of their START program enrollment, many experienced a significant improvement in overall subjective rating from pre-intervention to their post-intervention (20 session) time point.  Subjective rating scores were found to be positively correlated to both objectively coded social skill behaviors and survey measure scores.

Conclusions: Subjective ratings of social skill improvement may add a valuable new dimension to existing social competence measures already being implemented in the research literature.  Such a strategy directly taps into the immediate subjective impressions of potential social partners, which may be informed by a multitude of subtle verbal and nonverbal factors that are not adequately captured through traditional behavioral coding paradigms.