Effect of behavioral history on subsequent responding: experiments using rotary fixed-ratio schedules

The present study proposes a novel experimental approach to examine the effect of previous schedule history on subsequent responses. College students used a three-button device to indicate their choice of response. We performed two experiments. Each experiment consisted of three phases. In experiment I, fixed-ratio 10 (FR 10) reinforcement schedule was employed in phases 1 and 3. Phase 2 employed "Rotary-FR 10" schedule: after obtaining a point produced by pressing one of three specified buttons in a triangular array 10 times, the effective button changed to another in a clockwise direction. Optimal response pattern for Rotary-FR 10 was analyzed. Two types of behavior, referred to as "consecutive" and "scatter" were observed in phase 1. The mean number of optimal responses in consecutive type was significantly higher than scatter type in phases 2 and 3. In experiment II, two schedules were used in phase 1, and the same protocol used in experiment I was followed. We found that the reinforcement schedule used in phase 1 affected acquisition of the optimal response in phase 2. These results suggest that the earlier behavioral tendencies play an important role in determining subsequent behavior and that such behavioral tendency can be modified by specifically designed protocols.