Transition from variable delay of reinforcement schedules to variable-time (VT) schedules of reinforcement. A systematic investigation.

Burgess, Ian Stanley (1984). Transition from variable delay of reinforcement schedules to variable-time (VT) schedules of reinforcement. A systematic investigation. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.0000f91a

Abstract

Six experiments are reported in which rats were trained under response- dependent reinforcement schedules prior to being transferred to variable-time (VT) schedules which were matched in terms of both the overall rate and temporal distribution of reinforcement to the prior response-dependent reinforcement schedule. The response-dependent reinforcement baseline schedules varied according to whether they permitted reinforcement delays or only allowed the designated response to be contiguous with the reinforcer. In addition, some of the baseline schedules required either 'other behaviour'-response or response- 'other behaviour' sequences for reinforcement. The results indicated that delay of reinforcement schedules generally were associated with stronger resistance to response-independent reinforcement relative to immediate reinforcement schedules. Further, immediate reinforcement . schedules which demanded 'other behaviour'-response sequences for reinforcement tended to produce stronger resistance to response-independent reinforcement than those which permitted more than one response class to occur, but only required the emission of the designated operant for reinforcement. Rate of responding, level of efficiency of responding and postreinforcement pause duration during the baseline were poor predictors of response persistence during VT schedules. Response rate during the VT schedules was inversely related to mean delay of reinforcement (experiments IV and VI), while response decrement was related to the proportion of interreinforcement intervals without a response (experiments I, II,III, iv, v and VI). Reinforced and overall interresponse time (IRT) distributions during the VT condition were significantly correlated; further, changes in response rate were r e f l e c t e d in changes in the IRT distributions. It was concluded that an analysis of response maintenance under VT schedules which is framed in terms of the discrimination of response-reinforcer correlations was not supported. It was proposed that a molecular theory in which IRT-reinforcer contiguities are emphasized probably holds more promise.