Document Type



The present two experiments conducted a limited parametric study of the overtraining variable using a whole-partial reversal procedure in rats. Rats received two concurrent two-choice discriminations (Experiment 1) , two concurrent go-no go successive discriminations (Experiment 2), and then received reversal training under a given condition of a whole reversal or a partial reversal either immediately after criterion training, following 2, 10, or 20 days of overtraining in Experiments 1 and 2. In Experiment 1, after 20 days of overtraining, a whole reversal, in which both discrimination tasks were reversed, produced more rapid reversal than a partial condition, in which one of the two discriminations was reversed whereas the other was maintained as in original learning. Conversely, after 2 days of overtraining rats in the partial condition reversed faster than did those in the whole condition. Experiment 2 essentially replicated results of Experiment 1. Experiments 1 and 2 show that a stimulus function transfer and stimulus substitutability, respectively, are most clearly evident after 20 days of overtraining. These results indicate that additional prereversal training (i.e., overtraining) is an obvious operational precondition for stimulus classes formation in simple concurrent discriminations in rats.