1145 - ENHANCING EMOTIONAL COMPETENCE OF CHILDREN WITH AUTISM THROUGH SOCIAL ROBOT-ASSISTED INTERVENTIONS

Session: D06S016 - Digital Mental Health 2
AUTHORS:
Liu Xiaocen (Capital Normal University ~ Beijing ~ China) , Bi Yuwei (Capital Normal University ~ Beijing ~ China) , Zhang Jiayi (Capital Normal University ~ Beijing ~ China) , Dou Donghui (Central University of Finance and Economics ~ Beijing ~ China)
Abstract text:
Introduction: Children with autism spectrum disorder (ASD) often show deficits in emotional understanding and expression, which limit their social interactions and overall adaptation. Social robots, with their predictable and structured interactions, provide an innovative avenue for emotional interventions, yet evidence from controlled comparative studies remains limited.


Purpose: This study aimed to evaluate the effectiveness of social robot-assisted interventions, compared with therapist-only interventions, in improving the emotional competence of children with ASD.


Method: A randomized controlled trial was conducted with 13 children with ASD, assigned either to a robot-assisted rehabilitation group (n = 7) or a therapist-only group (n = 6). Interventions targeted emotion understanding (naming, re-recognition, and perspective-taking) and emotion expression (facial expressions in structured tasks and natural contexts). Assessments included behavioral tasks and observational measures across pre- and post-test sessions.


Results: Both groups showed significant improvements in emotional competence after intervention. However, the robot-assisted group demonstrated greater gains. Specifically, post-test scores for emotion naming, re-recognition, and perspective-taking were significantly higher in the robot-assisted group than in controls. Regarding emotion expression, children in the robot-assisted group exhibited significant improvements in happy and sad expressions under task conditions and sustained enhancement of happy expressions across later stages of the intervention. No significant improvements were found for anger and fear expressions, which may represent more threat-related emotions.


Conclusions: Findings indicate that social robot-assisted interventions can effectively enhance emotional understanding and the expression of non-threatening emotions (happy and sad) in children with ASD, with outcomes surpassing those of therapist-only approaches. These results highlight the potential of human-robot collaborative models in applied psychology and suggest promising directions for integrating emerging technologies into educational and clinical practices to support children's socio-emotional development.