[Computerized analysis of the verbal behavior of patients and therapists in the 1st diagnostic interview]

From 282 psychoanalysis-oriented First Interview existing in the Computarized Text-Bank of the University of Ulm, Germany, a 68-sample was selected in order to assess and evaluate the following points: (a) Describing the talking behavior of both patients and therapists, according to specific variables; (b) Constructing a sample for the future development of a "Computarized First-Interview Archive", and (c) Testing some hypotheses connected to the verbal style of psychosomatic patients, compared with psychoneurotic patients; female patients compared with male patients; and older patients compared with younger patients. Using computarized dictionaries and the Electronic-Verbal-Analysis System (EVA) already stored and programmed at Ulm, formal, grammatical, and content analysis were performed. As a result, the determinant influence of diagnosis, and the heavy influence of age and sex on verbal style were shown. As regards the text measurements discriminating capacity, it was found out that grammatical analysis, with its high reliability, and validity, proved to be the best discriminative tool.