Effect of Locus of Reinforcement Control on a Concept-Identification Task

Abstract

One subject from each of 20 triads matched on age and sex was randomly assigned to yoked contingent external, noncontingent external, and self-reinforcement groups. Differential reinforcement was evaluated in terms of accuracy and output of behavior during acquisition, maintenance, and extinction on a concept-identification task. Performances of the contingent self- and external reinforcement groups were equivalent during acquisition and maintenance, while superior to the noncontingent control group. During maintenance the noncontingent group subjects continued at their same level of output and accuracy in contrast to both contingent groups who improved. Groups did not differ in their accuracy during maintenance nor during the extinction phase when all subjects, as expected, were less accurate. Contrary to previous reports, the performances (output and accuracy) of subjects who administered their own reinforcers were not more resistant to withdrawal of an available external reinforcer than subjects who had received externally controlled consequences.