Abstract
The present experiments investigated the relationship between changes in the relative reinforced interresponse-time distributions and the occurrence of positive and negative contrast in multiple variable-interval-variable-interval and multiple variable-interval-extinction schedules of reinforcement. Experiment I demonstrated that changes in the interresponse-time distributions were consistently correlated with response-rate changes referred to as positive and negative contrast. Corresponding changes in the reinforced interresponse-time distributions suggested that negative contrast resulted as an inductive effect of selectively reinforcing long interresponse times in the altered component at the moment the baseline schedule was reintroduced. Experiment II demonstrated that the magnitude of the negative-contrast effect could be significantly decreased if the altered component schedule was modified in order to prevent the reinforcement of these interresponse times during the first few sessions of baseline recovery. The results supported a proposal that interresponse time-reinforcer relations may act as amplifiers or attenuators of negative contrast.