The Effects of Conditioned Reinforcers on Extinction When Delivered on Schedules of Extinction

The purpose of the present research was to examine extinction of responding with regard to the rapidity and thoroughness of the process when conditioned reinforcement was available on one of five schedules during extinction. Forty-five mixed-breed pigeons served as subjects with 15 in each of three experiments. Reinforcement training schedules were as follows: Experiment 1, continuous; Experiment 2, fixed ratio 15; Experiment 3, variable-interval one-minute. After training, subjects experienced one of five extinction procedures (here called schedules of extinction) which were as follows: traditional schedule without keylight did not provide conditioned reinforcement; traditional with keylight had the keylight on continuously but withheld other conditioned reinforcement (no schedule, per se, was used); the remaining three schedules (i.e., continuous, fixed ratio 15, and variable-interval one-minute) provided the following four conditioned reinforcers: the sound of the food magazine, the hopper light, the sight of food, and the keylight. Predictions for responding were based on the discrimination hypothesis which states that the more alike training and extinction conditions are, the slower the process of extinction. In order to compare response rates among subjects, a percentage of baseline responding was computed. Four spontaneous recovery tests were conducted to measure the thoroughness of the extinction procedures. Results did not support predictions based on the discrimination hypothesis; that is, subject response rates did not appear to be affected by the similarity of the extinction condition to previous training history. The second finding was that the most rapid and thorough extinction was obtained when the extinction schedule was traditional without keylight. When conditioned reinforcement was available, the continuous extinction schedule produced the most rapid and thorough extinction. The third major finding was that the schedule of unconditioned reinforcement was more predictive of extinction responding than was the conditioned reinforcement schedule during extinction. The last finding was that a subject\u27s pattern of responding was typical of the schedule whether it was on an unconditioned or a conditioned reinforcement schedule. It is suggested that extinction-of-a-human-intervention strategies might be more effective if conditioned reinforcement was identified and controlled