Date of Award:


Document Type:


Degree Name:

Master of Science (MS)



Committee Chair(s)

Richard B. Powers


Richard B. Powers


Carl D. Cheney


J. Grayson Osborne


This experiment examined the role of delays of reinforcement in producing the response pattern generated by fixed interval (FI) schedules. Conjunctive schedules which required one response and maintained a 100 sec interreinforcement interval were used. These schedules permitted a variable delay of reinforcement to separate the last response in each interval from reinforcement. The maximum delay was specified, however, by specifying the percentage of the interval which must elapse before responses counted toward reinforcement. Maximum delays of 100, 95, 75, 50, 25, 5 and 0 seconds were examined and their effects on response rate, response distribution, and post-reinforcement pause were measured.

The study generated the following results: (1) Although the maximum delay possible was specified, this delay was seldom taken; interim responses were emitted and the delay was shortened. Longer median delay of reinforcement produced lower response rates. (2) Fixed interval schedules produced the shortest post-reinforcement pauses. When the schedule allowed longer delays the pause became longer. These changes are seen in overall distributions and only slightly affect the mean pause length. (3) When no delays were possible, e.g. on FI 100 sec, the distribution of pauses within the interval was positively accelerated. With schedules which permitted delays the distributions became less positively accelerated. Generally, the longer the delay of reinforcement, the more negatively accelerated the response distribution within the interval.

These data indicate that the delay of reinforcement disrupts the FI schedule performance. Longer delays lower the rate, lengthen the pause and eliminate the pause. Dews, in "The effects of multiple SΔ periods on responding on a fixed-interval schedule," Journal of the Experimental Analysis of Behavior, 1962, 5, 369-374, suggests that the cumulation of delays of reinforcement produces the scalloped pattern. This experiment suggests that this is not the case. Even short delays, e.g. less than 8 sec, disrupt the scalloped pattern. These data suggest that the behavior occurring at the moment of reinforcement, rather than merely the delay of reinforcement per se, determines FI schedule performance.



Included in

Psychology Commons