Durability of Secondary Reinforcement as a Function of Manipulations in the Training and Testing Situations

Author

Lee E. Doerries, University of Rhode Island

Date of Award

1974

Degree Type

Dissertation

Degree Name

Doctor of Philosophy in Psychology

Department

Psychology

First Advisor

Nelson F. Smith

Abstract

Two experiments studied durability of secondary reinforcement as a function of manipulations in the training (Experiment I), and training and testing situations (Experiment II). In Experiment I, a total of 40 rats were trained on a single or four different tasks over four consecutive days. Rats received 120 pairings of white noise and food in either a simultaneous or backward conditioning procedure. Testing occurred on a new learning task (lever box) one wee following was consequated with white noise, the other having no consequence, were present during testing. The results showed that the simultaneous conditioning procedure was an efficient method for establishing white noise as a conditioned reinforcer when tested in a new learning paradigm. The backward conditioning control performed consistently at a low operant level of bar press responding. The hypothesis that animals trained on multiple tasks would emit more responses during testing than animals trained on a single task was not supported.

Experiment II repeated the conditions studied in Experiment I but omitted the backward conditioning control. A total of 160 animals were trained on one or four different tasks. Testing occurred either 0-, 1-, 2-, or 4-weeks following the end of training. Test sessions were either proportionately distributed over the four delay conditions (four 15 min. sessions) or were massed for one 60 min. session 0-, 1-, 2-, or 4-weeks following the end of training. This design produced 16 experimental conditions with 10 animals assigned to each condition. The results of Experiment II confirm the findings reported in Experiment I. Simultaneous conditioning was an efficient procedure for establishing a durable secondary reinforcer. Training on multiple tasks failed to increase the number of responses made in testing over response rates of animals that received training on a single task. A significant interaction resulted between 0- and 1- week for subjects tested under proportionately distributed trials. Bar pressing of subjects tested under massed conditions had extinguished at the end of the 60 min. test. Rats tested under distributed conditions 0- to 1-week following the end of training, continued to respond at high levels at the end of the fourth 15 min. session.

The results indicate that the type of test (massed or proportionately distributed) is a critical factor in assessing strength of secondary reinforcement. The response rates of rats tested under proportionately distributed trials (N=80) showed significantly less loss of original learning than animals tested under massed conditions. For distributed test subjects, extinction of the new learning response occurred on the final 15 min. distributed trial, two weeks following the end of training. The results indicate that durable secondary reinforcement can be reliably demonstrated without presenting primary reinforcement in the test situation. The sizable secondary reinforcement effects observed in the present study should renew interest in the new learning paradigm as an effective procedure for assessing the strength of secondary reinforcement.

Recommended Citation

Doerries, Lee E., "Durability of Secondary Reinforcement as a Function of Manipulations in the Training and Testing Situations" (1974). Open Access Dissertations. Paper 911.
https://digitalcommons.uri.edu/oa_diss/911