The Influence of Monetary Rewards on Performance and Attentional Effort Mobilization in a Visual Selective Attention Task

Abstract

Intuitively, most people assume that offering monetary rewards is a good way to motivate others to increase their performance. In contrast to this assumption, however, in analyses that examine the effects of monetary rewards on behavior, mixed results turn up: Although rewards do indeed often increase performance, they also decrease performance just as often. The most common result, however, is that these rewards simply do not have any measurable effect at all. Newer theories attempt to explain these mixed findings by stating that a host of mediators influence the relations between rewards and effort on the one hand and effort and performance on the other hand. For the sake of comprehension, the most important of these mediators are named and their effects are described. Following this, a bunch of newer studies that deal with the interactions between motivation and cognition is summarized. All of these studies report a modulation of behavior by monetary rewards. However, one question that these studies do not answer is whether monetary rewards also increase performance, and if so, by which mechanism they do so. Furthermore, almost all of these studies use very high rewards. Given the previous discussion of the various mediators the question arises if smaller rewards are capable of producing similar results, given that the various mediators are adjusted adequately. In order to investigate these questions, 3 studies are conducted. Study I is to be regarded as a pilot study that serves to determine the optimal methodology for the examination of these issues. Study II confirms that even small rewards are capable of increasing performance, and that this increase in performance is moderated by an increase in the quality of the sensory coding of the stimuli. In Study III, the main focus is on two important mediators of the relations between rewards, effort, and performance, namely, the ordering of the response deadlines and the rewarded aspect of behavior. As it turns out, a descending deadline order produces the largest increase in performance. Regarding the rewarded dimension of behavior, it turns out that it is easier for people to control their response speed rather than their accuracy, and consequently, a larger increase in performance is to be expected if slowness is punished harder than committing errors.