Hierarchical structure and general factor saturation of the Anxiety Sensitivity Index: Evidence and implications.

Boston University The Anxiety Sensitivity Index (ASI) is one of the most widely used measures of the construct of anxiety sensitivity. Until the recent introduction of a hierarchical model of the ASI by S. Lilienfeld, Turner, and Jacob (1993), the factor structure of the ASI was the subject of debate, with some researchers advocating a unidimensional structure and others proposing multidimensional structures. In the present study, involving 432 outpatients seeking treatment at an anxiety disorders clinic and 32 participants with no mental disorder, the authors tested a hierarchical factor model. The results supported a hierarchical factor structure consisting of 3 lower order factors and 1 higher order factor. It is estimated that the higher order, general factor accounts for 60% of the variance in ASI total scores. The implications of these findings for the conceptualization and assessment of anxiety sensitivity are discussed. Reiss and his colleagues (Reiss, 1987; Reiss Reiss, Peterson, Gursky, Clark, 1986; Goldstein McNally, 1990; Reiss, 1991; Reiss Reiss et al., 1986). Reiss et al. (1986) created one of the most Richard E. Zinbarg, Department of Psychology, University of Oregon; David H. Barlow and Timothy A. Brown, Center for Anxiety and Related Disorders, Boston University. This work was supported in part by Grant RO1M-39096 from the National Institute of Mental Health and by a Shannon Award (1 R55 MH/OD53425-01A1 ) from the National Institute of Mental Health. We are grateful to the individuals who graciously consented to serve as judges for the matching task. The judges included seven PhD-level psy- chologists with expertise in clinical psychology (Anne Marie Albano, Janet Borden, Paul Collins, Peter M. Lewinsohn, Susan Mineka, Anne Simons, and Robert Weiss), two PhD-level psychologists with expertise in personality and factor analysis (Lewis R. Goldberg, William R. Re- velle), and three graduate students in clinical psychology who had just completed an anxiety assessment and treatment practicum with Richard E. Zinbarg (Leslie Dana, Amy Reiss, Carie Rodgers). We thank Stephen Haynes, Scott O. Lilienfeld, Janet Mohlman, and Steven Taylor for their extensive and helpful comments on earlier versions of this article. We also thank Brian Cox, Steven Taylor, and Sherry Stewart for graciously providing the factor loadings from their analyses that have not yet been published. We are also grateful to Steven Taylor for providing the factor loadings from Peterson and Heilbronner's (1987) four-factor solution because these results are unpublished. Correspondence concerning this article should be addressed to Rich- ard E. Zinbarg, Department of Psychology, 1227 University of Oregon, Eugene, Oregon 97403-1227. Electronic mall may be sent via the In- ternet to rzin@oregon.uoregon.edu. 277 widely used measures of the AS construct, known as the Anxiety Sensitivity Index (ASI). Until recently, there has been considerable debate surrounding the issue of the factor structure of the ASI. On the one hand, Reiss et al. (1986) and several other researchers (e.g., Pe- terson Reiss, Peterson, Sandin, Chorot, Stewart, Dubois-Nguyen, Taylor, Koch, Taylor, Koch, McNally, War- die, Ahmad, & Hayward, 1990) advocated a multidimensional structure including as many as four factors. The debate regarding the factor structure of the ASI appears to have been largely resolved by a hierarchical model proposed by Lilienfeld, Turner, and Jacob (1993). Their model is hierar- chical in that it contains several first-order factors that aU load on a single, higher order factor. Furthermore, it is capable of integrating many of the apparently discrepant findings in the literature by suggesting that those investigators who have advo- cated a multifactor solution were focused on the lower level of the hierarchy, whereas those who have advocated a single-factor solution were focused on the higher level. Indeed, the hierarchi- cal model appears to have been largely accepted by Taylor (1995a), who was formerly one of the main proponents of the unifactorial view. Given the relatively long history of hierarchi- cal models of intelligence, an analogy to the structure of intelli- gence may be useful. Within a hierarchical model of intelligence, the identification of separable lower order factors would not dispute the existence of a general intelligence factor. Fur exam- ple, at least two group factors (e.g., Verbal and Spatial) in addition to a general factor of intelligence are hypothesized in Vernon's (1969) model of intelligence. Similarly, Lilienfeld et al. suggested that there may be group factors (i.e., common to some but not all items) and a general factor (i.e., common to nearly all items) underlying the ASI. Although a hierarchical model has promise for resolving con- troversies in this area, appropriate tests of the hierarchical nature