Multi-modal measurement of anxiety and social skills in a behavioral role-play test: Generalizability and discriminant validity

The Simulated Social Interaction Test (SSIT), a behavioral procedure for assessing anxiety and social skills, was evaluated using a generalizability approach with college students. In addition to self-report and behavioral ratings, heart rate responsivity during the SSIT was employed as an index of anxiety. For the most part, results of the generalizability analysis replicated previous findings with a psychiatric population, suggesting that the SSIT is a psychometrically sound procedure for obtaining judgments of anxiety and social skills. A moderate negative relationship was found between the ratings of anxiety and social skills, providing further support for the discriminative validity of these constructs. Finally, including a third measurement condition (physiological arousal) did not result in increased generalizability. Indeed, even when a subsample of "heart rate responders" was empirically identified, the unexplained residual variance was not decreased, suggesting a lack of convergence for different methods of measuring anxiety.