Schizophrenic cognition: regression or yielding to normal biases.

Abstract

Previous studies have shown that nonparanoid schizophrenics are more likely than normals to select word associates on forced choice tests that are more characteristic of children than they are of adults. Both the view that schizophrenics are regressed in their thinking and Chapman and Chapman's theory that schizophrenics are deficient in evaluating the strengths of competing responses can account for these findings. To test the merits of each theory, a four alternative, forced choice, word association task was constructed and validated with samples of undergraduates and children. This task and a free association task requiring self-generation of response were administered to nonparanoid and paranoid schizophrenics and to normal controls. The results for the forced choice task indicated that the associative hierarchy of nonparanoid schizophrenics approximated that of children, supporting the regression hypothesis and contradicting Chapman and Chapman's theory. The findings with the free association task were less clear, and the discrepancy in results for the two tasks is discussed.