Rate of Bouts as a Measure of Response Strength

University Of North Carolina Greensboro, Greensboro NC

Investigators

Abstract

Rate of Bouts as a Measure of Response Strength Richard L. Shull University of North Carolina at Greensboro It seems almost self-evident that the tendency to engage in an activity should be directly related to the size of the reward for the activity (incentive value) and to the motivation level. Yet response rate in the laboratory has often turned out to be surprisingly insensitive to such factors. For example, response rate is an increasing, negatively accelerated function of the rate of its reinforcement (well described by a rectangular hyperbolic function). And sometimes response rate even decreases as the size of the reward is increased. The proposed research is based on the possibility that response rate is insensitive because, as typically calculated, it is a mixture of two different kinds of responding, a sensitive type and an insensitive type. There is reason to suspect that rate of initiating bouts of an activity might be highly sensitive to incentive and motivational variables whereas responding within bouts might not be. If so, bout-initiation rate should be a better candidate than total (composite) response rate for determining the fundamental relation between responding and reinforcement. Despite a considerable amount of suggestive evidence, however, this possibility has not been adequately developed or evaluated--mainly because researchers have been unsure how to determine unambiguously and nonarbitrarily when bouts begin and end and thus how to measure bout-initiation rate. Recently Shull and his colleagues described and validated a technique for determining bout-initiation rate and bout length--based on fits to limbs of log survivor plots of interresponse times. Using this technique, the proposed research will examine the possibility that the function relating bout-initiation rate to reinforcement variables (such as rate of reinforcement) is invariant over different response types and over variables that alter response rate by changing bout length. The proposed research will also determine if the relation between bout-initiation rate and reinforcement rate is approximately linear over a wide range. If these possibilities are confirmed, the results would favor a surprisingly simple and general principle of response strength due to incentive and motivational variables.