On Diagnostic Categories of Archetypal Psychiatric Patients

Title:	On Diagnostic Categories of Archetypal Psychiatric Patients
Authors:	Ivo Düntsch , Dept of Computer Science , Brock University , St Catherines, Ontario, L2S 3A1, Canada Günther Gediga , Institut für Evaluation und Marktanalysen; Brinkstr. 19; D-49143 Jeggen; Germany Josef Rogner, Fachbereich Psychologie , Universität Osnabrück (Equal authorship implied)
Status:	Manuscript
Abstract:	The archetypal diagnostic approach (Cantor et al., 1980) is based on defining features which not all members of the class must possess. In this study, we re-analyse data of the diagnostic categories of archetypal psychiatric patients given in Mezzich & Solomon (1980) with the method of enhanced rough set data analysis (RSDA) (Düntsch & Gediga, 1997; Pawlak, 1982). RSDA has recently come to the fore as a means of data analysis in the medical sciences (e.g. Browne, 1997; Düntsch & Gediga, 1995; Paszek & Wakulicz-Deja, 1996; Pawlak et al., 1986; Tanaka et al., 1996; Woolery & Grzymala-Busse, 1994); it is a non - invasive method which uses only internal data information and does not rely on external parameters as most other taxonomic methods do. Even though the data set has been used in connection with DSM-II, our results show that the findings are compatible with the outcome of more recent studies. We also show that some of our results are significant in terms of highlighting dependencies among diagnostic categories which have not been noticed before.