Alternative factor models of the Personal Disturbance Scale (DSSI/sAD)

This study examined the factor structure of the Personal Disturbance Scale (DSSI/sAD) [Bedford, A. & Foulds, G. A. (1978). Manual of the Personal Disturbance Scale (DSSI/sAD). Windsor: NFER-Nelson] on data from a sample of 979 elderly respondents. A series of four alternative factor models were specified and estimated using weighted least squares in LISREL8 [Joreskog, K, G. & Sorbom, D. (1993a). Nee features in LISREL8, Chicago: Scientific Software]. Contrary to previous findings, and to the theoretical underpinning of the scale, a one factor model was judged to provide an adequate and parsimonious explanation of the data. The criteria on which this judgement was based were levels of overall fit, levels of component fit and model parsimony. The issue of parsimony has not been addressed in the previous examinations of the scale's structure. The findings of this study suggest that the DSSI/sAD is measuring a single factor, previously labelled general psychological distress [Bedford, A. gr Deary, I. J. (1997), The Personal Disturbance Scale (DDSI/sAD): development use and structure, Personality and individual Differences, 22(4), 493-510], rather than two distinct constructs of anxiety and depression. (C) 1998 Elsevier Science Ltd. All rights reserved.