Abstract
In each of 2 experiments, 16 rats were trained on a brightness discrimination (with or without overtraining), and then extinguished, either by withdrawing all reward, or by training on a new discrimination with the cues of the 1st present but irrelevant. The results of relearning tests showed that the extent to which habit strengths to S+ and S- had been equalized by extinction trials was directly related to the amount of original training: nonovertrained Ss showed a significantly greater tendency than overtrained Ss to respond in the direction in which they were originally trained. The results can be explained by assuming that discrimination learning involves 2 stages, and that these stages are differentially affected by overtraining. (16 ref.) (PsycINFO Database Record (c) 2006 APA, all rights reserved)