Construct Validity of Three Measures of Depersonalization in College-Exposed College Students

Abstract

Depersonalization is a type of dissociation characterized by feelings of unreality and detachment from one’s sense of self. Despite a history rich in clinical description, depersonalization has proven difficult to define and thus measure. Not surprisingly, available measures of depersonalization have limited psychometric support. The present study examined the construct validity of three self-report measures of depersonalization using a sample of trauma-exposed college students. Depersonalization measures included the Dissociative Experiences Scale (DES; Bernstein & Putnam, 1986), the Cambridge Depersonalization Scale (CDS; Sierra & Berrios, 2000), and the Multiscale Dissociation Inventory (MDI; Briere, 2002). These three measures were compared with respect to their pattern of correlations with a range of theoretically relevant self-report measures of psychopathology. Using Westen and Rosenthal’s (2003) procedure for evaluating a pattern of convergent-discriminant relationships, all three measures demonstrated good construct validity. The CDS and MDI demonstrated the best convergent, discriminant, and content validity, and the results strongly supported the use of the CDS and MDI for the assessment of depersonalization in this population. Implications for further understanding and refining the construct of depersonalization are discussed.