A word association analysis of schizophrenic thought processes

This study was a comparison of word association patterns of matched schizophrenic and nonpsychiatric patients. Patients were matched in age, sex, race, education, and verbal fluency. A series of ten word association conditions was administered which varied in instructions. The conditions were free association, repeated free association, five continuous associations, popular association, definition of the stimulus, and final free association. Response patterns consisted of (a) response quality (judged by the degree to which a response was a quick, meaningful, single word), (b) response commonality (judged by response rank among responses of a reference group to the same stimulus words), (c) response latency in tenths of seconds, (d) semantic relationship of stimulus-response pairs, and (e) temporal series of two or more responses. These patterns were more compatible with several of the many postulated causes of the schizophrenic condition. It was found that the basic cognitive responses of schizophrenics were not disrupted. Rather, schizophrenic patients tended to change more from one semantic set to another than did nonpsychiatric patients. Schizophrenic responses were less common and less quick, especially after the first continuous association response was produced. The acceptable responses of schizophrenic patients varied more than those of nonpsychiatric patients. Response faults were not fully explained by perceptual mistakes, lack of stimulus word knowledge, motivation, or a calculated effort to appear "sick." The results were more compatible with fluctuating attention and editing deficit postulates.