Simple reaction chains and their integration. I. Homogeneous chaining with terminal reinforcement.

Abstract

The author's purpose was a study of reaction chains in which no locomotion was involved and in which trial-and-error was kept at a minimum. Using albino rats and an apparatus which permitted serial button-pressing, the latency of reactions was measured. One group of 29 rats was reinforced with food and a second group of 17 rats was reinforced by cessation of shock. Preliminary training in the operating of the buttons was carried on until the animals had achieved a criterion of 10 successive trials with an average latency of 10 seconds or less. They were then given one chaining trial a day for 25 days. The shock reinforcement proved less effective for learning than did the food. With both types of reinforcement the latency was maximal at the first link of the chain, and minimal at the third (next to last) link. The results are interpreted in terms of deductions from the goal gradient principle and stimulus generalization. 16 references. (PsycINFO Database Record (c) 2006 APA, all rights reserved)