Effect of Noncontingent Exposure to Shifts in Reward Magnitude on Subsequent Instrumental Runway Performance

Abstract

This experiment extends the paradigm of nonresponse acquisition and extinction studies to the case of shifts in reward magnitude. In each of two replications, rats were first trained to traverse a runway for a reward of either 1 or 8 food pellets. Subgroups then received a series of direct placements (DP) into the goalbox with either the same or the opposite reward magnitude to that received in acquisition. This was followed by another series of running trials on which all Ss now received the reward magnitude opposite to that received in acquisition. The results indicated that on the first post-DP trial, speeds were a function only of the reward magnitude received on DP trials. Over all test trials, however, a complex pattern emerged For Ss shifted from 1 to 8 pellets, DP with 8 pellets produced faster performance changes and permanently faster running than DP with 1 pellet. For Ss shifted from 8 to 1 pellet, one replication showed no difference between groups which had received 1- or 8-pellet DPs in the rate at which performance dropped, while in the other replication the group receiving 1-pellet DPs showed no performance decrease at all. The implications of these results for two-process theories of instrumental reward learning are discussed briefly.