Retention period differentially attenuates win–shift/lose–stay relative to win–stay/lose–shift performance in the rat

Hungry rats were trained in a two-lever conditioning chamber to earn food reinforcement according to either a win–shift/lose–stay or a win–stay/lose–shift contingency. Performance on the two contingencies was similar when there was little delay between the initial, information part of the trial (i.e., win or lose) and the choice portion of the trial (i.e., stay or shift with respect to the lever presented in the information stage). However, when a delay between the information and choice portions of the trial was introduced, subjects experiencing the win–shift/lose–stay contingency performed worse than subjects experiencing the alternative contingency. In particular, the lose–stay rule was differentially negatively impacted relative to the other rules. This result is difficult for ecological or response interference accounts to explain.