The Anxiety Sensitivity Index-3: analyses of dimensions, reliability estimates, and correlates in nonclinical samples.

We investigated the factor structure, reliability estimates, and correlates of the Anxiety Sensitivity Index-3 (ASI-3; Taylor et al., 2007) in 2 studies. We established a bifactor model in Study 1 as an alternative representation of the structure of the ASI-3. Analyses of gender differences on the total ASI-3 and subscale scores were not statistically significant (Study 1, N = 462). In Study 2 (N = 293), results of a series of confirmatory factor analyses provided stronger support for the fit of the bifactor model compared with 2 alternative models. Estimates of scale reliability were adequate (all rho values > or = .80) and not "p" (as in italic p for significance). in the 2 studies. In addition, using simultaneous regression analyses, we found anxiety-specific correlates for the total ASI-3 and subscale scores to include responses on self-report measures of interpersonal sensitivity, obsessive-compulsive anxiety, paranoid ideation, and phobic anxiety.