Matching to Reinforcer Value: Human Concurrent Variable-Interval Performance

Abstract

A male paedophile offender chose between pairs of sexual stimulus classes in a two-operandum procedure with reinforcement arranged according to concurrent variable-interval variable-interval schedules. Condition 1 involved choice between slides of women and slides of men; condition 2, slides of men vs. slides of children; and condition 3, slides of women vs. slides of children. On the assumption that the subject matched his ratios of responses and times on the two operanda to reinforcer value, estimates were obtained from the data in conditions 1 and 2 of the values of slides of women and of children relative to slides of men. These yielded predictions of the relative allocations of responses and of time in condition 3. The predicted and obtained outcomes were similar. The results support the use of matching-based scaling of reinforcer values in the prediction of choice between qualitatively different reinforcers and extend the application to a human subject, the two-operandum procedure, and sexual stimuli.