Confirmatory factor analysis of single- and multiple-factor competing models of the dissociative experiences scale in a nonclinical sample

Previous research on the Dissociative Experiences Scale (DES) has demonstrated that (a) dissociation is quantifiable in both clinical and nonclinical samples and (b) a three-factor structure (amnesia, depersonalization, and absorption) is tenable for clinical samples. The factor structurefor nonclinical samples is less clear, with one- and multiple-factor solutions proposed. To clarify the DESfactor structure in nonclinical samples, confirmatory factor analyses were conducted on (a) one-, two-, three-, and four-factorfirst-order models and (b) two bifactor (hierarchical) models of DES scoresfor two samples of nonclinical university students. Results of delta(chi2) and goodness-of-fit indices support the three-factor (first-order) model as bestfitting of the datafor these samples. The utility of this DES model for screening both dissociative pathology and elevated normal dissociative behavior in clinical and nonclinical populations is discussed.