Relationships among premorbid adjustment, prognosis, and choice of symptomatic behaviors in schizophrenic patients

A comparison was made among three separate criterion measures of premorbid adjustment for effectiveness in predicting types of symptoms manifested during schizophrenic episodes. The instruments under investigation were Section I of the Prognostic Rating Scale, PRS-H, (Phillips - 1953), Social Competence Index, SC-I, (Zigler and Phillips, 1960), and the abbreviated version of the Social Competence Index, SC-II, (Phillips, Broverman & Zigler, 1966). The sample included 60 patients (30 males, 30 females) with a clearly-established diagnosis of schizophrenia. It was drawn from a larger study in which a major concern was to obtain a broad range of individual differences in 'variables figuring in socio-economic class status (education, intelligence, occupational level), but a restricted range in terms of age (18-45) and length of hospitalization (less than six months). Case histories of the Ss were examined first to abstract material germane to the ratings of premorbid adjustment on the three scales. Complete inter-rater agreement was secured on each scale. The records were then reanalyzed to obtain information about symptomatic behavior, which was categorized as either 'action-oriented' or 'thought-oriented' according to a method developed by Phillips and Zigler (1961). Independent raters also classified symptoms; there were relatively few disagreements and these were arbitrated. During all of the case history analysis, identifying information was masked and coded. Correlation values derived in scale comparison were as follows: SC-I-SC-II, .66; SC-I-PRS-H, .51; SC-II-PRS-H, .11. The degree of association between the SC-I and the SC-II was attributed to item overlap; that between SC-I and PRS-H was interpreted as probably due to common content involving marital status and marital adjustment. In a previous study with Ss of quite varied diagnostic status, Phillips, Broverman, and Zigler (1966) found that the SC-II instrument was an effective predictor of symptom-occurrence. With the relatively carefully-selected sample of schizophrenic Ss employed in the present investigation, the efficiency of SC-II in predicting symptom-occurrence was no greater than chance. It appears likely that the former prediction situation involved little more than the sample identification of social class difference in the operations of symptom-identification, diagnosis, and prognosis. Good and poor prognosis groups derived from PRS-H ratings were balanced in sex distribution and did not differ significantly in mean IQ scores. From both an empirical and a theoretical standpoint, the results indicated that, among the three scales, this scale was the most effective predictor of symptom occurrence in schizophrenic patients.