Homogeneous and heterogeneous reward of monkeys.

Publisher Website
Google Scholar
Abstract
Two general methods are distinguishable in studies that are concerned with the effect of amount of reward. These are the "invariant method," which, varying the amount among different groups of Ss, restricts each S to one amount, and the "comparison method," which requires each S to experience two or more amounts. Application of the invariant method has generally shown that both speed, or rate, of response and resistance to extinction are greater with larger rewards than with smaller ones (e.g., 3, 16), but has yielded conflicting results in experiments where error measures are obtained (e.g., 4, 12). Application of the comparison method with proximate temporal spacing of reward amounts has generally shown a positive relation between amount and performance efficiency (e.g., 1, 10). Little attention has been directed to the crucial temporal variable. Although reward amount has occasionally been found to have a significant effect when large intervals were used (6, 11), it is likely that the results were in part dependent on an inherent opportunity for reward comparison and might have been produced had E chosen the invariant method. The findings of Davis (1) indicate that interday shifts in reward are much less effective than intraday shifts. In most experiments utilizing the comparison method, amount has been confounded with habit in the sense that one reward was changed to another on a single continued task. This may be avoided by varying the size of reward between two or more tasks that are given the same S (7, 13). In the present investigation the comparison method is applied without confounding to the study of different temporal arrangements. A serial-discrimination procedure (8) permits assignment, of the same or different amounts of reward to problems learned concurrently. Therefore, comparisons can be made between temporally distant learning performance under heterogeneous and homogeneous reward conditions, between differently rewarded problems within the heterogeneous conditions, and between two different homogeneous reward conditions. (PsycINFO Database Record (c) 2006 APA, all rights reserved)

This publication has 0 references indexed in Scilit: