Schizophrenics' use of perceptual grouping in visual stimuli

This study investigated Place and Gilmore's (1980) proposal that schizophrenics are unable to perform the first stage in perceptual processing (perceptual grouping). An alternative hypothesis (the Selective Processing hypothesis) was tested which states that, when explicit instructions are given, schizophrenics can respond to the organizational features in a display.

Twelve schizophrenics and twelve normal subjects briefly viewed a series of target displays of lines. Next, subjects saw two alternative displays presented simultaneously in a forced-choice format; they chose the display that was most similar to the target. Forced-choice displays showed one alternative that was identical to the target in number (a feature perceived by local analysis) and another alternative that was identical to the target in organization (a feature perceived by global analysis). The subject's selection of one alternative indicated whether he had applied global or local perceptual analysis to the stimuli. Subjects were tested under three bias conditions: (1) an induced global bias to attend to the grouping of stimulus elements; (2) an induced local bias to attend to the number of stimulus elements; (3) a neutral bias.

The Place and Gilmore hypothesis predicted that schizophrenics would consistently apply a local bias to the stimuli, and would be unable to adopt a global bias. The Selective Processing hypothesis predicted that schizophrenics would apply a local bias under the local and neutral bias conditions, but would apply a global bias when induced to do so. Both hypotheses predicted that normals would consistently apply a global bias.

The results failed to support the Selective Processing hypothesis. They partially supported Place and Gilmore's (1980) hypothesis in that schizophrenics failed to respond to the induction of a global bias; schizophrenics did respond to the induction of a local bias. Contrary to both hypotheses, normals could be induced to apply a local bias as well as a global bias. The findings also suggested that whether subjects use global or local processing is influenced by the total number of elements in a display, and possibly, by the way these elements are configured in space.