Date of Award:


Document Type:


Degree Name:

Doctor of Philosophy (PhD)



Committee Chair(s)

Edward K. Crossman


Edward K. Crossman


J. Grayson Osborne


Richard B. Powers


Frank R. Ascione


Carl D. Cheney


Donald V. Sisson


Generally human titration performance under schedules of reinforcement has not been investigated. In an attempt to examine the variables which control titration, an interlocking progressive-ratio schedule was devised. Under an interlocking progressive-ratio schedule, the number of responses required for reinforcement increases by a constant (the increment value) after every ratio, but during each ratio the response requirement can be lowered (the decrement value) by emitting pauses of a specified duration (the stepdown duration).

The first experiment sought to determine if children would titrate when exposed to interlocking progressive-ratio schedules. Although three of the four subjects did not show evidence of titration initially, through a series of manipulations all came under schedule control.

Experiment 2 was conducted to determine the effects of a series of increment value manipulations on the level of titration. The series of increment values was tested under large and small decrement conditions. The results showed that increment value was an important determiner of titration level only when the decrement value was small. When the decrement value was large, changes in increment value had no effect upon titration.

In Experiment 3 the effects of a series of decrement manipulations on titration level were examined. The decrement manipulations were investigated under two increment values. The results indicated that as the decrement value was decreased the titration level tended to increase under both increment conditions.

In all of the experiments, rate of responding, pausing, titration variability, and patterns of responding were examined. Generally rate of responding, pausing, and response patterning were found to be related to changes in increment and decrement values. Titration variability showed no systematic changes across manipulations.

Analysis of subjects' behavior in terms of preference indicated that the subjects tended to minimize number of responses rather than maximize reinforcement frequency. Subjects would pause to bring down the response requirement, and thus increase the time between reinforcements, rather than emit responses on a lever, which would have resulted in more reinforcements per unit of time.



Included in

Psychology Commons