The External Validity of Behavioral Fear Assessment

Abstract

This article reviews the literature relevant to the external validity of laboratory tests: (a) the degree to which the fear-eliciting stimuli in behavior tests adequately represent the phobia-relevant situations that subjects encounter in the natural environment, and (b) the effects of measurement reactivity on subjects' responses to these stimuli. We then review the small amount of empirical research correlating subjects' performance on laboratory tests with information about their naturalistic functioning. Our conclusion, based on both conceptual and empirical considerations, is that laboratory tests do not provide data that currently allow us to predict how subjects will function in naturalistic situations targeted by treatment; thus, research using these tests as major outcome criteria cannot be used to argue for the clinical effectiveness of change techniques. Finally, we present methodological suggestions designed to (a) encourage research cross-validating subjects' laboratory performance with information about naturalistic functioning, and (b) facilitate the development of laboratory tests with greater external validity.