The differential outcomes procedure can interfere or enhance operant rule learning.

The differential outcomes effect--the enhancement of learning and memory performance by correlating distinct reinforcers with to-be-remembered events (sample stimuli)--has been stated to be one of the most robust phenomena in learning psychology. However, in this paper we demonstrate that the correlation between unique samples and unique reinforcers can either interfere with or enhance learning a spatial matching-rule, dependent on whether these two processes are trained concurrently or sequentially. If the Pavlovian conditioning (unique sample-reward pairings) occurs before the matching rule is learned (sequentially), the conditioned expectations of unique rewards will enhance the acquisition of the spatial matching-rule in rats (the differential outcomes effect will be observed). However, if rats are required to learn the Pavlovian associations and the matching-rule concurrently, they are impaired in acquiring the spatial matching-rule. Thus, employing the differential outcomes procedure can either enhance or detract from learning and remembering the task rule-dependent on the nature of the task and order of training. These data suggest that under some circumstances learning Pavlovian associations can compete with the formation of instrumental behavior.