Effects of different reinforcer probabilities and delays on choice as a function of income level

In Experiment 1, 3 rats earned their daily food ration by choosing on a trials basis between a lever that delivered one food pellet immediately and another that delivered 3 pellets after a delay that was adjusted to ensure responding to both alternatives. Choice of the delayed reinforcer increased when the number of trials/session was reduced. This change in preference defended food intake. In Experiment 2, 3 rats earned their daily ration by choosing on a trials basis between a "risky" and a "risk-free" lever. The risky lever produced either 15 pellets with p = 0.33 or no pellets with p = 0.67, while the risk-free lever always produced 3 pellets. Risky-lever choices decreased when trials/session were reduced. This preference change was incompatible with defending food intake: The demonstration in these experiments that reducing trials/session resulted in increased preference for the alternative that had the greater delay (Experiment 1) and the higher probability of reinforcement (Experiment 2) is incompatible with models that posit that reinforcer probability and immediacy (1/delay) function equivalently in choice (e.g., Rachlin, Logue, Gibbon, & Frankel, 1986).