The effects of amount and variance of reinforcement on risky choice

In Experiment 1, four rats were given a choice between 3 pellets with p = 1.00 and 15 pellets with p = 0.33 or zero pellets with p = 0.67. Reinforcement amount was manipulated by varying number of free-choice trials per session across conditions. When number of trials per session was decreased, rats preferred the risk-free alternatives. The results were not consistent with the prediction of a risk-sensitive foraging model. Experiment 2 varied total amount of reinforcement by keeping the number of free-choice trials per session constant while reinforcement size was increased across conditions. In lower-reinforcement conditions, four rats were given a choice similar to that used in Experiment 1. In higher-reinforcement conditions, reinforcer size was tripled. When reinforcement amount was decreased, rats preferred the risky alternatives. The results of Experiment 2 supported the prediction of a risk-sensitive foraging model.