| Abstract: | Three experiments were conducted to explore the specificity of the contrast effect in judgments of psychopathology. In the first two studies, respondents initially attempted to infer whether each item in a series of behavior samples (vocabulary definitions in one study, handwriting samples in the second) came from a schizophrenic or a nonschizophrenic patient. Some respondents were presented with highly pathological samples in this induction series, while others were presented with relatively nonpathological samples. These divergent experiences led to marked contrast effects in evaluating test stimuli from the same behavioral domain (e.g., additional vocabulary definitions), but had significantly less impact on the respondent's reactions to stimuli from a different domain (e.g., handwriting samples). A third experiment yielded similar results, using a paired-comparison methodology. In this study subjects first judged a series of high- vs low-pathology definitions. They were then presented with a series of matched stimulus pairs, each including one vocabulary definition and one hand-writing sample. Respondents were to indicate the member of each pair that seemed more indicative of schizophrenia. People assigned to the high-pathology induction group typically chose the handwriting samples as being more indicative of schizophrenia, compared with respondents who were assigned to the low-pathology conditions. These results were interpreted as supporting a representational (perceptual) theory of contrast. |
