Rapid Assessment of Delay-of-Reinforcement Effects on Key Pecking Maintained by Fixed-Ratio Schedules

Previous research on delay of reinforcement has focused primarily on variable-interval schedules. A fair amount is known about effects of delay duration, signaled versus unsignaled delays, and interval length in the context of these schedules. Less is known about these effects on other schedules of reinforcement. For example, little is known about the effects of delay duration, signals versus unsignaled delays, or the way that delays interact with ratio requirement when behavior is maintained by fixed-ratio schedules. Furthermore, the steady-state designs (wherein stable responding is obtained at each delay duration) used to investigate delay-of-reinforcement effects are often time consuming. The present experiments used delay-of-reinforcement procedures to rapidly examine effects of delay duration and signals on responding under different fixed-ratio requirements. Experiments 1-3 used a procedure wherein the delay of reinforcement increased each session until the pigeon paused for 300 s to examine effects of delay duration (Experiments 1-3), signaled versus unsignaled delays (Experiments 1-3), the influence of decreased rates of reinforcement (Experiment 2) and the role of interval versus ratio baseline schedules (Experiment 3). Experiment 4 examined the interaction between signaled versus unsignaled delays of different durations and fixed-ratio requirement using a procedure wherein the fixed-ratio requirement increased each session until the pigeon paused for 300 s. These procedures yielded negative relations between delay duration and response rate that were consistent with previous findings using steady-state designs with variable-interval baselines, and were consistent across schedule type (Experiment 3) and fixed-ratio requirement (Experiment 4). These findings suggest generality in this negative relation across experimental design, schedule type, and schedule parameters. A similar decrease in response rates was not obtained when the rate of immediate reinforcement was similarly decreased (Experiment 2), suggesting that these findings cannot be wholly accounted for by changes in rates of reinforcement. Response rates were similar with signaled and unsignaled delays during Experiments 1-3, which rapidly changed delay durations, but not during Experiment 4, which held delays constant across a range of fixed-ratio requirements. This finding suggests that the higher response rates often seen with signaled (relative to unsignaled) delays may require several sessions to be established