Schizophrenic performance has been interpreted as being the result of ability loss and of impression management. An experiment is reported which allows a direct test of the ability loss assumption as opposed to one form of the impression management hypothesis. 20 healthy presenter and 20 sick presenter schizophrenics were given differential “normative” information concerning the meaning of word association performance. Healthy presenters shifted their performance in the direction predicted by the impression management hypothesis but sick presenters did not. Results suggest that the likelihood of occurrence of impression management depends heavily on the degree of pathology displayed by S. (PsycINFO Database Record (c) 2006 APA, all rights reserved)