Effect of instrumental response pretraining in classical-to-instrumental transfer of a differential reward magnitude discrimination

Abstract

Amount of preliminary bar-press training (25 vs 125 discrete fixed-ratio trials) and amount of noncontingent pre training (0 vs 500 presentations of cues and associated reward magnitudes with the lever retracted) were manipulated factorially prior to contingent discrimination between five- and one-pellet cues. The 500-presentation groups demonstrated differential responding in start speed earlier in the contingent phase than the O-presentation groups, confirming previous findings of classical-to-instrumental transfer of the discrimination. The major effect of the lever pre training was a higher small-reward start speed in the 125-trial groups than in the 25-trial groups. Ratio speed data differed markedly from start speed data in showing large differences in favor of one-pellet ratio speeds early in training and very little evidence of five- over one-pellet discrimination late in training.