The effects of response contingent parental reinforcement upon schizophrenics' learning

Forty schizophrenic and 40 non-schizophrenic male psychiatric inpatients, matched for age, intelligence, and social competence, were administered equivalent form nonsense syllable discrimination tasks under each of four conditions: an initial baseline information feedback condition, two response contingent reinforcement conditions, and a final baseline information feedback condition. Subjects of both diagnostic groups were further divided into eight groups of ten patients each. Half the groups received praise and half received censure during the response contingent conditions. In counterbalanced order, the reinforcing agent was the experimenter in one condition and a recorded voice of one of the parents in the other. Half received the mother's and half received the father's voice. Schizophrenics responded faster under experimenter than under parent reinforcement. Poor premorbid schizophrenics, making more errors than goods, responded faster under experimenter as opposed to parent reinforcement. Among non-schizophrenics, education, social competence, and I.Q. were all negatively correlated with response latency and number of errors. Among schizophrenics, the higher the I.Q., the fewer the errors and the shorter the response latency, except when such subjects were reinforced by their parents. Then no significant relationship existed. Marital status, however, was significantly related to performance, with unmarried status positively correlated with slower response to parents. All subjects responded more quickly and accurately in the final than in the first baseline condition regardless of diagnosis or interpolated reinforcement experiences. Data were interpreted as providing limited support for the social censure hypothesis of Garmezy and his colleagues, but as consistent with recent formulations concerning the schizophrenic patient in relation to his family.