Stimulus control over induction produced by upcoming food-pellet reinforcement.

Research has demonstrated that rats increase their rate of lever pressing for sucrose reinforcement when food-pellet reinforcement will soon be available within the session. Recent results suggest that this increase occurs because stimuli in the session come to signal different levels of overall reinforcement. The present experiment tested this idea by having rats respond in two types of session. In one, they pressed lever A for 1% sucrose reinforcers in the first half of the session and lever B for food-pellet reinforcers in the second half (the opposing lever was retracted in the respective half). In the other, they pressed lever B for 1% sucrose reinforcers in the first half and lever A for 1% sucrose reinforcers in the second. Thus, the presence of lever A in the first half of the session was predictive of upcoming food-pellet reinforcement, but only lever B was ever used to obtain food pellets. Subjects responded at a higher rate on lever B in the first half of the session than on lever A, despite food-pellet reinforcement being unavailable in such sessions. Furthermore, they responded at a higher rate on lever B during probe sessions in which both levers were available. These results demonstrate that stimulus control over induction occurs when a stimulus becomes differentially associated with a heightened level of reinforcement. However, questions remain as to whether this association is Pavlovian or whether stimuli with "predictive value" may ever lead to induction.