Reinforcement-rate maximizing in the Vaughan and Miller (1984) procedure

Four food deprived pigeons responded on a multiple schedule in which components switched after reinforcement. In one component (subtractive), reinforcers were assigned and stored by a continuously operating variable-interval 30-second schedule. Additionally, every 20th response debited the reinforcement store by one. Thus, if responding was frequent, reinforcement was infrequent. The second component (yoked) was a variable interval, the intervals of which were equated with the subtractive component that occurred five components earlier. The results reproduced Vaughan and Miller's (1984) major findings. However, response rates were higher in the yoked component relative to the subtractive, indicating a maximizing strategy.