Response latency at zero drive after varying numbers of reinforcements.

Abstract

A light-aversion expt. was run with white rats to find the functional relationship between the strength of performance under zero drive and the no. of previous reinforcements under max. drive. Strength of performance under each drive condition was measured by the difference between the median log latency of response after training and the unconditioned level before training. The response may be specified as a "cross-the-box-through-the-swinging-door" response. Two zero drive conventions were considered, one a response-definition, as that level incapable of producing learning, the other a stimulus specification as the total absence of an aversive stimulus. The low level of illumination employed as zero drive met the 1st criterion and approached the 2d. With the primary light drive at zero, the strength of performance was found an increasing, negatively accelerated function of the no. of previous reinforcements under light drive. The performance strength under no-drive was approx. 50% as high as that under high-drive, when measured by the post- and pre-training log latency differences, and approx. 20% when measured by the no. of trials to reach a performance criterion. The lack of adequate conventions relating theory to data in this problem area is noted with a justification for the conventions used in the present study. The results are interpreted as confirming the hypothesis (Hull''s Corollary II) under test.