Summer Camps As Social Performance Intervention for Adolescents with ASD and Their Peers

Background:   For adolescents with ASD who have developed the requisite social skills, there are limited opportunities to improve social performance(Koegel et al., 2012).  Koegel et al. (2012) tested the effectiveness of a social performance approach with elementary school-age children in their school setting with typically developing (TD) peers.  This approach was unique from other peer-mediated models in that the participating peers were not selected or trained to work with children with ASD.  Rather, the success of the approach relied on the shared interest of the children to facilitate natural social interactions.  Whether such approach could successfully be adapted to adolescents with ASD is yet unclear.

Objectives:   We provided a series of technology-related summer camps, which offered an engaging and supportive environment in which adolescents with ASD could practice appropriate social and collaborative skills with TD peers. Shared common interest was in one of two topics: robotics or computer game programming (CGP). The primary goal of the camp was to improve social anxiety and social skills.

Methods:  We recruited 33 individuals with ASD and 36 TD peers, ages 10-17 years, who received general education science instruction at school, and expressed interest in either robotics or CGP.  Participants completed a weeklong summer camp, during which they were trained in programming either a humanoid robot or computer games, depending on the type of camp in which they were participating.  In all camps, participants had the opportunity to program a robot or computer game while collaborating in pairs (1 ASD: 1 TD).  Participants were not labeled as having ASD.  Social skills were taught indirectly through assigned technology-related exercises, such as programming the robot to make appropriate eye gaze and gestures to interact naturally with the audience.  Pre- and post-treatment data were collected on participant-reported levels of social anxiety (Social Anxiety Scale) and parent-reported social skills (Social Skills Improvement System).

Results:   The data from all camps were combined and a series of paired samples t-test were conducted to compare pre- and post-intervention data. The ASD group experienced a significant improvement on social skills and social anxiety from pre- to post-test. It should be noted that averages for the ASD never reached the standardized norm on the measures.  While the TD group did experience some improvements in both measures, they never reached significance. This is to be expected considering their already desirable levels of social skills and social anxiety at the start of the camps (see Table 2).  When the sample was analyzed separately by the type of camp, the ASD sample in the robotics camp showed improvements in both social skills and social anxiety while the ASD sample in the CGP camp showed improvements in only social anxiety.  TD children, regardless of the type of camp, did not show any improvements.

Conclusions:   These results provide a strong support for the effectiveness of social performance-based camps for adolescents with ASD.  Whether there is something unique about programming humanoid robots that is more conducive to social skills learning needs to be explored further.