NATURE AND MEASUREMENT OF ANXIETY SENSITIVITY - REPLY TO LILIENFELD, TURNER, AND JACOB (1996)

The present article has two related aims. First to reply to the recent claims made by Lilienfeld, Turner, and Jacob. Second, as part of my r eply I will present an integrative review of the current status of res earch on the nature and measurement of anxiety sensitivity (AS). In co ntrast to the largely unsubstantiated claims by Lilienfeld et al., the re are empirical, methodological, and conceptual bases for drawing the following conclusions: (a) there are theoretical grounds for proposin g that AS predicts fear-proneness; (b) correlations between the Anxiet y Sensitivity Index (ASI) and measures of common fears are not artifac ts of overlapping content; (c) the ASI is not a measure of panic sympt oms; it measures the fear of anxiety- and panic-related sensations; (d ) the factorial structure of the ASI is highly relevant to the constru ct validity of this instrument; (e) a unifactorial ASI is the most rel iable (replicable) factor structure. However, several studies have fou nd support for multidimensional structures. There is growing evidence that the construct of AS is probably multidimensional, consisting of a t least three dimensions: fear of somatic sensations, fear of cognitiv e sensations, and fear of publicly observable sensations; (f) The hypo thesis that the AS-by-trait anxiety interaction predicts fear pronenes s received partial support from a methodologically flawed study. The h ypothesis was not supported by methodologically sound research. These conclusions are supported by the available literature and by new findi ngs presented here for the first time.