Pseudolearners in discrimination reversal.

This research investigated the shape of the acquisition and reversal functions in retardates' discrimination learning, with safeguards against suspected artifacts in other studies. A modified Wisconsin General Test Apparatus was used with 60 retardates in a 2 X 2 factorial design. Another 11 were classified as pseudolearners. The variables were the prereversal criteria and the number of irrelevant dimensions. After meeting criterion subjects were given 30 more trials for the detection of pseudolearners, but all received reversal training. Performance was stationary in both acquisition and reversal, suggesting all-or-none learning. For learners, differences of midreversal plateaus appeared which were opposite to predictions of the Zeaman and House (1963) theory. When pseudolearners were included, however, the plateau effect partially conformed to the theory.