Concreteness, overinclusion, and symptomatology in schizophrenia

In the area of conceptual functioning in schizophrenia, there is the following central controversy: Goldstein regarded schizophrenic conceptual performance as a deficit condition involving extreme 'concreteness'—an absolute loss of abstracting ability; Cameron considered the impairment in schizophrenia as one involving 'overinclusion'—unmonitored response to irrelevant internal and external stimuli in conceptual performance. .The overinclusion vs. concreteness controversy has stimulated a good deal of research. Most of this research has been of a hypothesis-testing type, which has left the original constructs unanalyzed, and led to a proliferation of tests, a broadening of criteria, and a diversity of explanations stretching, extending, and diffusing, the original meaning of the terms. The terms 'overinclusion' and 'concreteness' have been used theoretically and operationally in varied ways. The basic aim of the present study was to clarify the meaning of these terms through more precise definitions and analysis of measurement procedures. The fulfillment of this aim involved the following steps: (1) an analysis of the various descriptions, criteria, and measures of overinclusion and concreteness; (2) the gathering and construction of representative tests and measures to assess, as precisely as possible, each specific meaning assigned to the constructs; and, (3) assessment of the relationship among test measures by means of a factor analytic design. The second aim of this study was to relate differences in conceptual performance to differences in schizophrenic processes. Fulfillment of this aim involved relating conceptual performance to schizophrenic symptoms, as determined by ratings; these ratings were based upon a system devised by Zigler and Phillips to derive a Process-Reactive ratio of differences in symptomatology. A battery of seven tests containing 40 measures was gathered and constructed to reflect different aspects of concreteness and overinclusion. It was predicted that factor analysis of the battery would elicit two factors— overinclusion and concreteness—and that the test loadings on these factors would define the meaning of the terms. It was also predicted that symptoms of process schizophrenia would be related to concreteness and symptoms of reactive schizophrenia would be related to overinclusion. The battery of tests was administered to 60 male chronic schizophrenic patients at the VA Hospital, Houston, Texas. The patients were selected on the basis of established psychiatric diagnosis. The data were first analyzed by computation of Pearson r's. Several measures were dropped because they were obtained through correlated scoring procedures. A total of 24 variables were included in the final analysis. The resulting 24 x 24 intercorrelation matrix was factor analyzed by the Principal Axis method. Eight factors were obtained and rotated to simple structure by the Varimax method. Factor scores were obtained and the correlation of the Process-Reactive ratio and the Syndrome scores (groups of symptoms) with the extracted factors were computed. The predicted two general factors did not emerge from the factor analysis; instead one general and seven specific factors were extracted. The findings of the study suggested that the structure of conceptual pathology in schizophrenia involves at least four dimensions—Conceptual Deficit; Conceptual Autism; Overinclusion vs. Underinclusion and Associational Disturbance. Factor I approached the status of a general factor; however; certain measures of both concreteness and overinclusion loaded on this factor. It was concluded that a distinction between concreteness and overinclusion; at least as assessed by some of the measures in the present study; was not warranted. An analysis of the literature indicated that the validity of most measures of overinclusion and the existence of an overinclusion factor independent of measures of concreteness has not been firmly established. The underlying communality of Factor I was inferred to be a failure of the conceptualization process. The tasks loading on this factor were of two types: errors in identifying characteristics of presented categories and failures to categorize objects or stimuli. The interdependence of the two processes were discussed. Of the other factors, the most important was Factor IIB, Conceptual Autism. Factor IIB was a test specific factor with high loadings from the open-private conceptual area. This study strongly supported a distinction between the 'overinclusion' of including irrelevant elements in the definition of a concept and the overabstractions of open-private responses. No relationship was found between the Process-Reactive ratio of this study and the extracted factors. The rationale of the Process-Reactive ratio was criticized, and some evidence was adduced to indicate the possible value of a separate-symptom approach. A general critique of the process-reactive distinction was given; the need for studies demonstrating the intercriteria comparability of the different process-reactive systems was emphasized. The discussion concluded with a review and critique of current research strategies in the area of conceptual behavior, and some recommendations for methodological improvements. The thesis advanced was that future research in this area should emphasize basic methodological improvements (e.g., elaboration of scoring systems, establishing of construct validity of important constructs) and deemphasize hypothesis-testing research until more ground-work has been completed.