| Abstract: | An experiment was performed to assess the role of expectations in psychosocial contrast effects. Subjects were initially presented with a series of word definitions that were either high or low pathology; their task was to indicate those definitions that had been produced by schizophrenic (vs nonschizophrenic) patients. After this “induction” series, subjects were presented with a common “test” series that included both definitions and handwriting samples; they evaluated the degree of psychopathology that was implied by each item. Some subjects were required to state their expectations in connection with the test items. First they attempted to guess the percentage of schizophrenics in the overall test series; second, before being presented with each test item, they guessed whether the patient in question was (or was not) schizophrenic. Subjects' expectations were consistent with the induction series to which they were assigned; those assigned to the high-pathology induction expected the test series to include more schizophrenic patients than those in the low-pathology group. Nonetheless, despite these expectations, subjects showed a contrast effect in their evaluations of the test definitions (but not in evaluating the handwriting samples). That is, subjects from the low-pathology induction group rated the test definitions as being more pathological than did those in the high-pathology condition. Within-group correlations indicated that the subjects' expectations were directly related to subsequent judgments. A two-path model is discussed, relating (a) the respondents' induction group assignment, (b) his/her expectations, and (c) subsequent test judgments. |
