A comparison of empirical and theoretical explanations of temporal discrimination.

The empirical goals were to describe the behavior of rats trained on multiple temporal discriminations and to use these descriptions to predict behavior observed under new training conditions. The theoretical goals were to fit a quantitative theory to behavior from one training condition, estimate parameters for the intervening perception, memory, and decision processes, and use these parameters to predict behavior observed under new conditions. Twenty-four rats were trained on a multiple-peak-interval procedure with two stimuli that were presented individually (Stimulus A and B), or in compound (Compound AB); either different responses (Experiment 1) or the same response (Experiment 2) were reinforced during the presentations of Stimulus A, Stimulus B, and Compound AB. The patterns of correct and stimulus-error responses during Stimulus A and Stimulus B (Experiment 1) were used as elements that, with summation rules, predicted behavior under new conditions (Compound AB, Experiment 1; Stimulus A, Stimulus B, and Compound AB, Experiment 2). A comparison of the success of the empirical and theoretical goals supported the use of a quantitative theory of behavior to explain the data.