SENSITIVITY TO REINFORCEMENT AND CHANGEOVER REQUEREMENTSIN DYNAMIC AND QUASI-STABLE ENVIRONMENTS

Abstract

An ABA design served to assess sensitivity to reinforcement in dynamic and quasi-stable environments. In phase A, eight rats responded on two levers that provided food according to seven reinforcement ratios (27:1, 9:1, 3:1, 1:1, 1 :3, 1 :9, and 1 :27), one per day. To alternate between the levers, a changeover lever required 1, 4, 8, 16, o 32 responses. Over 21 days all reinforcement ratios were studied with the same changeover requirement (COR); then, a different COR was arranged. This cycle was repeated until all CORs were assessed in ascending and descending orders. Phase B used a dynamic environment where all reinforcement ratios were assessed in random order, one per day. Within the same session, all five CORs were scheduled to occur in random order; each COR lasted for ten reinforcers and was separated from the next COR by a 60-s blackout. Phase B lasted 70 days, ten sessions for each reinforcement ratio. The last phase was a replication of phase A, with the exception that the CORs were only assessed in ascending arder. The results ~howed that distributions of responses favored the lever associated with the highest reinforcement probability. On both levers, the changeover rates decreased with increasing changeover requirements. Generally, sensitivity to reinforcement increased with increasing changeover requirement.

These findings are consistent with models of choice that predict overmatching when a cost is imposed on the behavior of switching from one alternative to the other.