Effects of reinforcer density versus reinforcement schedule on human behavioral momentum
Slivinski, James G.
The essential tenet of the behavioral momentum model (BMM) is that relative response rate decreases less in the face of disruption when maintained by a higher reinforcer density. Empirical support exists based on both response-dependent and response-independent reinforcement. In the present study the BMM was tested with college students in 4 multi-element experiments, each using 2 reinforcement schedules and a disrupter. Participants performed a categorical sort (by orientation) of triangles on a computer monitor. Sorting response rates were disrupted by a concurrent task, pressing the keyboard “T” key whenever 2 displayed changing numbers were equal. Initial training established fast (under VR 4) and slow (under DRL 5-s) sorting rates, and provided practice with the disrupting task. In Experiment 1 DRL 5-s provided higher reinforcer density, while in Experiment 2 VR 4 did. In Experiment 3 the higher total reinforcer density was achieved by adding VT 6-s to DRL 5-s while in Experiment 4 it was achieved by adding VT 12-s to VR 4. In all 4 experiments, sorting rate decreased with introduction of the disrupter. In Experiments 1 and 3, relative sorting rate decreased less under DRL based schedule (greater reinforcer density), supporting the BMM. However, in Experiments 2 and 4, relative sorting also decreased less under DRL (lower reinforcer density), contrary to the BMM prediction. Taken together, these data show greater relative resistance to change under DRL (versus VR), independent of reinforcer density. Thus, contrary to the BMM, the nature of the reinforcement schedule seemed to be the principal factor determining behavioral momentum.
behavioral momentum, dynamics, resistance to change, response rate, disruption, reinforcement schedule, rules, schedule effect, reinforcer density, pavlovian contingency, relative rate, resistance to disruption