Generalization effects of training on the Wisconsin Card Sorting Test for schizophrenia patients.

Several studies have demonstrated that schizophrenia patients can improve their performance on the Wisconsin Card Sorting Test if given appropriate training. However, it is not clear if they are learning a generalizable problem-solving strategy rather than a circumscribed rule. We examined this question by training patients on one of two related problem-solving tests and assessing transfer of training effects across the two instruments; a third group received practice on both tests, but no training. Subjects were 27 schizophrenia patients in the latter stages of an acute hospital admission. While subjects in both training conditions exhibited marked improvement on the trained test, the results failed to reveal any evidence of generalization. Subjects trained on one test performed no better on the other instrument than subjects who received practice only. These results suggest that gains in WCST performance reported in the literature after brief training or instructions may not reflect any meaningful changes in problem solving capacity or in any underlying neurological function. Implications of the findings for rehabilitation programs are discussed, along with limitations of the data and recommendations for future research.