Social Role-based Training Using a Computer Graphics Avatar in a Real-world Commercial Setting for Individuals with Autism Spectrum Disorder: A Quasi-experimental Study.

Abstract

The lack of social role-based behaviors, such as guiding and assisting others, is a major barrier to employment success and community participation for individuals with autism spectrum disorder (ASD). Enhancing self-efficacy is essential for improving these behaviors. To enhance the self-efficacy of social role-based behaviors, we developed a training system that enables users to perform social role-based behaviors, including guiding and assisting others, using a computer graphics avatar in a real bustling commercial facility (TCR). We aimed to evaluate the feasibility of TCR and its potential to reduce anxiety and enhance self-confidence, motivation, metacognition, and self-efficacy in performing social role-based behaviors in real-world interactions among individuals with ASD. Thirty-eight adults with ASD were assigned to either the TCR group (n=20), which received TCR alongside communication guidance from teachers (CGT), or the control group (n=18), which received only CGT, based on the chronological order of enrollment. Both groups underwent a 5-session intervention (days 14, 21, 28, 35, and 42). Anxiety, self-confidence, motivation, metacognition, and self-efficacy related to social role-based behaviors were assessed at days 0, 14, 28, 42, and 56 using self-report and observer ratings. Linear mixed model analyses showed group × time interactions for self-reported self-efficacy at day 56 (t₁₄₄.₀₁₆ = 2.670, P=.008, 95% CI 0.349-2.340). Nominally significant group × time interactions were also observed for self-reported anxiety, self-confidence, motivation, and metacognition, suggesting greater improvements in the TCR group than in the control group. Observer-reported outcomes demonstrated significant group × time interactions for self-efficacy at day 56 (t₁₄₄.₀₁₆ = 4.500, P<.001, 95% CI 0.405-1.039), anxiety at day 56 (t₁₄₄.₀₁₆ = -3.593, P<.001, 95% CI -1.425 to -0.414), and self-confidence at day 56 (t₁₄₄.₀₁₆ = 3.321, P=.001, 95% CI 0.269-1.059), indicating greater improvement in the TCR group than in the control group. Nominally significant group × time interactions were additionally observed for observer-reported motivation at day 56 (t₁₄₄.₀₁₆ = 2.471, P=.015, 95% CI 0.103-0.930). Possible group × time interactions were also observed for observer-reported metacognition at day 56 (t₁₄₄.₀₁₆ = 1.858, P=.065, 95% CI -0.023 to 0.751). This study is innovative in that TCR enables users to perform social role-based behaviors. Unlike previous studies, TCR integrates opportunities for observing peers and role models, along with real-time encouragement, feedback, and verbal persuasion from a trainer, thereby incorporating multiple sources of self-efficacy. By demonstrating the feasibility of TCR and its potential to enhance self-efficacy in social role-based behaviors, this study contributes to the development of more effective interventions targeting functional social outcomes in individuals with ASD. TCR may support the development of skills relevant to employment and social participation in real-world settings and represents a potential scalable approach for community-based ASD interventions.　.