Journal of Abnormal Psychology 1991, Vol. 100. No. 3.308-315

Copyright 1991 by the American Psychological Association, Inc. 0021-843X/91/S3.00

DSM-IV: Empirical Guidelines From Psychometrics Rosemery O. Nelson-Gray

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of North Carolina at Greensboro

This commentary addresses the use of psychometric theory and methodology in the development of the 4th edition of the Diagnostic and Statistical Manual of Mental Disonlers(DSM-lV).R.e\iability issues include interdiagnostician reliability, temporally consistent diagnoses, and the relations of diagnostic criteria within categories. Validity issues include content validity of the diagnostic criteria, criterion-related validity (the relation between different criterion sets or their algorithms and alternative diagnostic criteria), and construct validity (the relation between diagnostic categories and external validators). Specific questions and methodology to investigate its utility vary with the different uses proposed for the diagnostic system. Specific psychometric methodologies that may be useful in developing the DSM-IV are noted, as are the limitations of psychometrics and their applicability to DSM-IV.

Literally, millions of words evaluating the current psychiatric diagnostic systems, the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IH; American Psychiatric Association, 1980) and its revision (AW-///-R; American Psychiatric Association, 1987) have been written. (For a historical review of psychiatric diagnosis, see Gray, 1978.) These evaluations have included abundant suggestions for improvement directed at the in-process DSM-IV (e.g., Last & Hersen, 1987; Millon & Klerman, 1986). Many of these suggestions have focused on greater utilization of empirical methodologies, in response to what was seen as the largely political and consensual nature of DSM-III and DSM-UI-R. Empirical methodologies recommended for DSM-IV ate often derived from psychometric theory. Indeed, Blashfield and Livesley (1991) explicitly detailed the ways in which psychiatric classification may metaphorically be viewed as a psychological test, to justify the use of psychometrics. Other authors, including the present one, have made and are making recommendations for the DSM-IVbased on psychometric theory, implicitly assuming the utility of this metaphor. Psychometric theory and methodology are also not without their critics. Criticisms range from the messiness of the various subtypes of reliability and validity (which resulted in a neater organization, known as genemtizability theory; Cronbach, Gleser, Nanda, & Rajaratnam, 1972) to the questionability of the psychometric assumption that measurement is a composite of accurate measurement of immutable entities plus measurement error (Hayes, Nelson, & Jarrett, 1986). This commentary uses psychometric concepts as a basis for enhancing the emp irical underpinningsof the DSM-IV, but with occasional recognition of the shortcomings of psychometrics. Most of this commentary is devoted to Axes I and II, with the last portion di-

reeled at other axes of the multiaxial format of the current and planned DSMs. Reliability Issues Reliability is a determination of the consistency or replicability of the products of an assessment device (Anastasi, 1982). As applied to the DSM-IV, reliability is an examination of the consistency with which diagnoses are assigned. Interdiagnostician Reliability The main empirical data collected during the DSM-III field trials were on interdiagnostician reliability, with moderately good results

DSM-IV: empirical guidelines from psychometrics.

This commentary addresses the use of psychometric theory and methodology in the development of the 4th edition of the Diagnostic and Statistical Manua...
851KB Sizes 0 Downloads 0 Views