The Three Systems Model of Fear and Anxiety: Implications for Assessment of Social Anxiety

It has been argued that verbal reports and questionnaire data may be constrained by cultural and social demands to the extent that they are unlikely to show concordance with more direct measures of behavioural and physiological anxiety. The present report investigates this hypothesis with reference to social anxiety. It was found that physiological and cognitive questionnaires were good predictors of physiological anxiety and cognitions in a social interaction test. A behavioural questionnaire, however, was a poor predictor of actual behaviour. In the social interaction test, which was designed to produce anxiety, there was a high degree of concordance on all the measures except non-verbal behaviour and heart rate. Because behaviour ratings showed least concordance and the poorest predictive validity, it was suggested that a behavioural assessment alone may give an erroneous picture of a person's overall social anxiety.