An experimental analysis of steady-state response rate components on variable ratio and variable interval schedules of reinforcement.

Three experiments explored a novel approach to analyzing the different components of response rate that are produced by exposure to free-operant schedules of reinforcement. It has been suggested that overall response rate comprises a tendency to initiate responding, and to continue to respond once the bout is initiated. Previous post hoc analyses of interresponse times (IRT) data have suggested several features of these different aspects of responding that the current experimental procedure broadly confirmed. Increasing the size of a variable interval (VI) schedule decreases the number of "burst-initiation" responses, but has less effect on responding once the burst has been initiated (Experiment 1); that the major difference between a variable ratio (VR) schedule and a VI schedule, is not in the number of "burst-initiation" responses, but in the number of "within-burst" responses, with shorter interresponse times that are emitted, with greater numbers of such "within-burst" responses being emitted on a VR schedule (Experiments 2 and 3).