| Abstract: | The DSM-III method of diagnosing Axis II personality disorders is reviewed, and the following points are emphasized, (a) Axis II uses a categorical system for diagnosing personality disorders, but the advantages of a dimensional system appear to outweigh the categorical approach. Two alternative dimensional models are discussed (i.e., Millon's three dimensions of active-passive, self-other, and pleasure-pain and Leary's interpersonal circumplex), and research suggestions are presented, (b) Axis II is inconsistent in the extent to which it is based on a classical or a prototypic model of categorization. Implications of a prototypic typology for testing hypotheses concerning specific etiology and treatment and for determining relative diagnostic efficiency, are presented, (c) An innovative feature of DSM-III is the use of explicit and fixed rules to increase the reliability of diagnosis. Fixed rules, however, can decrease diagnostic efficiency when the cut-off points for diagnosis are not adjusted for the local base rates. Furthermore, while it is acknowledged that operational definitions increase reliability, increased reliability can result in decreased validity. An alternative to operational definitions is the use of multiple methods of measurement of constructs convergent and divergent to the hypothetical construct in question. |
