Effects of varying the probability of reinforcement correlated with two neutral stimuli in a temporally defined schedule

Abstract

Two experiments were conducted to assess the effects of varying the probability of reinforcement correlated with two neutral stimuli in temporally-defined schedules. In Experiment 1, 8 white Wistar rats were employed. The rats were exposed to three experimental phases with reinforcement probabilities set at 1.0-0.0, 0.75-0.25 and 0.5-0.5 with a redetermination of the first phase, and to three probes interpolated between the experimental phases with reinforcement being provided non-contingent upon responding. For four subjects, reinforcement probabilities were correlated with diferent stimulus lights (SD and S°), while for the other four subjects a single stimulus light was present during the probability variations in the experimental phases. In Experiment 2 six white Wistar rats were employed, that were exposed to the same conditions of experiment 1, with the exception that there was a different lever for responding to each of the two stimuli (SD and S°). The results of both experiments did not support the assumption that differential rates and patterns of responding would be established in the presence of the neutral stimuli according to their correlation with different probabilities of reinforcement. Frequency of responding was very low and a high percentage of reinforcers was lost. During experimental phases responding was higher during EE, while during probes responding was higher during ED. These results are examined in relation to the properties of temporally-defined schedules and some interpretations about findings are raised.