Positive and negative behavioral contrast in the rat.

Three groups of rats received either 8, 23, or 53 sessions of multiple variable-interval variable-interval baseline training before being shifted to a multiple extinction variable-interval schedule. The rate of responding during the unaltered component was higher for the groups shifted to multiple extinction variable-interval than for control groups remaining on multiple variable-interval variable-interval (positive contrast). Furthermore, when the multiple variable-interval variable-interval schedule was re-instated, stable negative contrast was found in the groups that had received 23 or 53 baseline sessions, but not for the group that had received only eight sessions. Positive and negative contrast were also demonstrated in the eight and 23-session groups when the multiple extinction variable-interval and multiple variable-interval variable-interval schedules were re-administered in further phases of the experiment. These results suggest that both positive and negative behavioral contrast can be obtained reliably in a species other than the pigeon.