The effects of delay-to-reinforcement, response requirement and schedule duration on performance under fixed-ratio schedules

Experiment 1 examined the effects of delay-to-reinforcement on between-ratio pausing in multiple fixed-ratio 5 fixed-ratio 5 schedules. In one component the reinforcer was delivered immediately on completion of the ratio, and in the other there was a delay before reinforcement. The longest between-ratio pauses always preceded the ratio in which the delay was signalled, and these pauses increased as the delay increased. Experiment 2 examined the effects of varying delay-to-reinforcement on behaviour under fixed-ratio schedules that were increased each session, keeping the time available for responding constant. At each ratio requirement, with increased delay, total consumption and response rates decreased and between-ratio pause durations increased, although this effect was small. Functions were fitted to the relation between the natural logarithms of the consumption and fixed-ratio size using Hursh, Raslear, Shurtleff, Bauman and Simmonds’s (1988) nonlinear and Hursh and Silberberg’s (2008) exponential equations. Both fitted the data well, but in neither case did any of the parameters vary systematically with delay. The only consistent finding was that initial consumption was largest when there was no delay and smallest in the condition with the largest delay. This suggests a decreasing effect of delay-to-reinforcement with increases in the response requirement. Experiment 3 used conjunctive fixed-ratio fixed-interval schedules to separate the effects of schedule duration and ratio requirement in increasing fixed-ratio schedules. The schedule-initiation immediacies (the inverse of the pause following reinforcement) decreased in conditions in which the ratio was fixed and duration increased, and remained relatively constant in conditions in which the ratio was increased and duration fixed. It is suggested that the schedule-initiation immediacy is a useful measure as it eliminates some of the confounds of measures such as total consumption or consumption rate. These experiments show that with some procedures delay-to-reinforcement affects behaviour under fixed-ratio schedules in a way similar to reduced reinforcer magnitude and an increased force requirement. However, with other procedures, delay-to-reinforcement has different effects from reinforcer magnitude and quality. Also, overall, these three experiments show that increased time to reinforcement (whether as part of schedule requirements or at the end of the response requirement) has a larger effect on behaviour than the number of responses it takes to complete the ratio