Psychometric Properties of a Social Skills Assessment using a Virtual Environment

Abstract

Social Anxiety Disorder (SAD) is a marked and persistent fear of social and/or performance situations in which embarrassment or scrutiny from others may occur. One form of behavioral assessment is a Role Play Task (RPT). However, RPTs often are not feasible in clinical settings due to the common obstacles in implementation. Thus, the current study examines the feasibility, acceptability and psychometric properties of a virtual environment based social skills assessment compared to the traditional RPT. Forty-six children, ages 7 to 14, participated in two assessment conditions: RPT and a virtual environment behavioral assessment (VE BAT). Participants reported self-ratings of anxiety and acceptability, while blinded observers rated social skills and overall social anxiety. An ANCOVA, covarying for age, revealed (a) no significant task difference for voice volume, speech latency, number of words spoken, effectiveness, and SAM ratings; (b) that the VE BAT was more feasible to implement in terms of personnel time and costs and; (c) more overall anxiety during the RPT task than during the VE BAT task. In addition, the VE BAT demonstrated moderate concurrent validity when correlating the self-report ratings of anxiety with the Social Phobia and Anxiety Inventory for Children (SPAI-C). The current study addresses many of the impediments to conducting RPTs in a clinical setting and, overall, supports the utilization of VE BATS as a viable alternative for the behavioral assessment of social skills in children. Further implications address the potential for the armamentarium for social skills training with children with SAD.