Partial avoidance contingencies.

Rats were trained in a discrete-trial paradigm with no intertrial interval. The first response changed an auditory stimulus for the remainder of the trial. Shocks were delivered only at the end of the trial cycle. Avoidance contingencies were defined by the conditional probability of shock, given no response (P(0)), and the conditional probability of shock given a response (P(1)). The maximal avoidance contingency was P(0)=1.0, P(1)=0, and noncontingent conditions were those for which P(0)=P(1). In Experiment I, after training on the maximal contingency, three groups of subjects experienced either P(0)=P(1)=0, P(0)=P(1)=0.5, or P(0)=P(1)=1.0. Eight of 10 subjects stopped responding under the noncontingent conditions. Experiment II studied partial contingencies by varying P(0) and P(1). For one group, P(0) was reduced holding P(1)=0. Responding decreased to zero as P(0) approached zero. A second group was studied under P(1)>0, holding P(0)=1.0. For three of the six rats in this group, responding decreased to zero with increasing P(1). The other three maintained responding as P(1) was increased up to the noncontingent, P(1)=P(0)=1.0 value. The P(0) group was also studied with P(0)=P(1)>0, and half of these subjects responded. The results demonstrated two modes of response to weakening or eliminating the avoidance contingency. Some subjects were sensitive to contingency only, and insensitive to changes in shock density. Approximately one half of the subjects were sensitive to both contingency and shock density. This shared control was observed only when P(1)> 0.