Amsel's frustration effect: A Pavlovian replication with control for frequency and distribution of rewards

Amsel's frustration effect refers to a facilitation of behavior immediately after unexpected nonreward. The original experiments were plagued with problems of interpretation, mainly because the critical comparison involved conditions that were not equated in terms of reward frequency and amount. A replication of the original experiment (with a 50% partial reinforcement schedule used to induce response facilitation) is reported here with one important addition: A control group matched in terms of reward parameters, for which the reward was not signaled by a discrete cue. Matching is achieved through the use of a Pavlovian training procedure. The results show facilitation of rats' lever-contact performance after unexpected reward omissions. The control group showed no facilitation. These results provide support for the view that response facilitation is caused by increased arousal after unexpected nonreward.