Is working more costly than waiting in monkeys?

Takafumi Minamimoto,Yukiko Hori,Barry J Richmond

doi:10.1371/journal.pone.0048434

Abstract

We studied how value for instrumental action is discounted by predicted effort and delay. The monkeys were trained to perform instrumental trials that required a bar release when a visual target changed from red-to-green. There were two trial conditions. In delay trials, after the monkeys performed one instrumental trial correctly a reward was delivered 0–7 seconds later. In work trials, the monkeys had to perform 0, 1, or 2 additional instrumental trials to obtain a reward. The lengths of trials in delay matched the time it took to complete work trials. The length of delay or number of trials was indicated by a visual cue presented throughout the trial. Our hypothesis was that the monkeys would all show temporal discounting of reward in the delay trials, and that in the work trials the monkeys’ performance might reflect an additional cost due to working. The error rate increased linearly as remaining cost increased for all 8 monkeys. For 4 monkeys the error rate was significantly larger in work trials than in delay trials (effort sensitive monkeys). For the other 4 monkeys there was no significant difference in error rate (effort insensitive monkeys). Since the error rate has an inverse relation with value for action, these results suggest that value is discounted hyperbolically by effort as well as by delay. Error rates generally increased as the testing sessions progressed and the total reward accumulated (i.e., effect of reward devaluation). The relative impact of delay and effort on error rates was reasonably stable within subjects. Thus, within the monkey population there seems to be a significant dichotomy in the sensitivity governing whether working is more costly than waiting, possibly arising from a constitutional or genetic trait.

Highlights

Studies in economics, psychology and behavioral ecology show that the performance of rewarded tasks is affected by the anticipated or predicted physical and/or mental cost, of obtaining the reward
This can be extended to discounting by workloads; when monkeys have to perform a sequence of simple instrumental trials the reward value is discounted as a function of the number of trials remaining before reward [10]
The monkeys had experienced the time from release to reward delivery in the delay trial (1 remaining cost unit (CU), 3.6660.01 s; 2 CU, 7.2560.02 s) and in the work trials (1 CU, 3.7460.06 s; 2 CU, 7.1660.16 s), respectively

Summary

Introduction

Psychology and behavioral ecology show that the performance of rewarded tasks is affected by the anticipated or predicted physical and/or mental cost, of obtaining the reward. When offered a choice between a small reward that is available sooner and a larger reward that is available in the more distant future, rats [1], pigeons [2], and humans [3] frequently choose the smaller reward in such a situation. In these studies, the choice can be predicted by discounting the reward’s intrinsic value by the duration of the expected delay, an effect designated as ‘‘delay discounting’’ [1,4]. This can be extended to discounting by workloads; when monkeys have to perform a sequence of simple instrumental trials (a reward schedule) the reward value is discounted as a function of the number of trials remaining before reward [10]

Methods

Results

Conclusion