Resistance to Extinction of the Partial Delay of Reward Concerning the Few-trial PRE

Abstract

The partial reinforcement effect (PRE) which is obtained after a small number of training trials can't be easily explained by the foregoing theories, since in these the crucial mechanisms of the PRE presuppose rather extensive trials. Thus a PRE-mechanism that is independent of the number of trials shoud be hypothesized. Capaldi interpreted the PRE in terms of the generalization of conditioned stimulus-aftereffect. The present study was conducted to determine whether this aftereffect-mechanism is independent of the number of training trials, employing a situation in which reward was partially delayed.In Exp. I, extended training (48 trials) was used with 4 groups of rats;continuous reinforcement (CR), partial reinforcement (PR), 5 sec. partial delay of reinforcement (PD5), and 40 sec. partial delay of reinforcement (PD40). In Exp. II, the same design was used, but limited training (5 trials) was given. In each experiment, the subjects received a training and an extinction in a straight alley.In two experiment, the PRE was obtained and the resistance to extinction increased as a function of the delay time of reinforcement (Fig. 1 and Fig. 3). The results of Exp. II show that CAPALDI'S prediction concerning the extinction process following the partial delay of reward is applicable even to the condition of fewtrial training, and combined with those of Exp. I, suggest that the generalization of stimulus-aftereffect was independent of the number of training trials. The applicability of CAPALDI'S stimulus-aftereffect hypothesis to the few-trial PRE was proposed.