Psychometric Evaluation of Social Cognitive Measures for Adults with Autism

Although social cognition is frequently identified as a target in clinical trials and psychosocial interventions for adults with autism spectrum disorder (ASD), these efforts are hampered by a lack of consensus and validation of social cognitive measures. The current study provides psychometric evaluation of 11 frequently used measures encompassing different subdomains of social cognition. Adults with autism (N = 103) and typically developing controls (N = 95) completed 11 commonly used social cognitive tasks spanning the domains of emotion processing, social perception, and mentalizing/theory of mind. We examined each measure's internal reliability and sensitivity to group differences, how performance related to general intellectual ability, and alignment of measures with a proposed two-factor structure of social cognition in ASD. Controls outperformed the ASD group on 8 of the 11 social cognitive tasks, with the largest group differences occurring on two mentalizing measures, The awareness of social inference task (TASIT) and hinting task. In ASD, all tasks demonstrated strong internal consistency and avoided ceiling and floor effects. Social cognitive performance was also related to, but not redundant with, intellectual functioning. We also found support for a two-factor structure of social cognition, with basic social perception and emotional processing aligning into a lower-order social perception factor, while mentalizing tasks aligned into a higher-order social appraisal factor. In sum, eight tasks showed adequate to strong psychometric properties. The psychometric data, effect size estimates, and correlations between measures reported here can be used for study planning for social cognitive interventions in autism. Autism Research 2019, 12: 766-778. © 2019 The Authors. Autism Research published by International Society for Autism Research published by Wiley Periodicals, Inc. LAY SUMMARY: We examined 11 tasks that measure how adults with autism perceive and interpret social information. Eight of the tasks were reliable and showed lower performance in adults with autism compared to typically-developing controls. Task performance was related to but distinguishable from IQ. These measures evaluated here may be useful in assessing the effectiveness of interventions and treatments to improve social abilities in adults with autism.