In one of these studies the authors examined millions of putts from professional golfers and suggested that the par score of a hole served as a reference point for players; with putts being less accurate when attempting shots below par (Pope and Schweitzer, 2011). Another study examined penalty

kick shootouts of soccer games (Apesteguia and Palacios, 2010). This study proposed that the score of the shootout served as a player’s reference point and leading or lagging in score had an influence on performance; with those lagging in score performing worse than those leading. These studies provide interesting insights into the SB431542 datasheet possibility of an endogenous reference point of value influencing skilled task performance. However, these hypotheses are difficult to generalize because the contexts in which the sports are played are highly variable and the data lack a degree of experimental control. Furthermore, it was impossible to directly isolate players’ endogenous reference point of value because psychological and physiological measures were not available in these data sets. Instead, these studies only infer possible mechanisms used to define reference points during task performance. Our study provides direct behavioral and neural evidence of the mechanism responsible for encoding

an endogenous reference point during skilled task performance for incentives. It is important to realize that the hypothesis of a reference dependent encoding of value, and exactly how this reference point is defined, was informed and driven by our initial imaging analysis (experiment 1). Without Carnitine palmitoyltransferase II Decitabine cost this fMRI analysis, one would simply expect, as we did initially, that increasing incentives for task performance are encoded solely as increasing potential gains. In contrast, our fMRI analysis informed the hypothesis that the brain encodes increasing potential gains

when the amount of incentive is initially presented, however when actually performing the task potential gains are reframed in terms of losses. This neurally informed hypothesis was confirmed using a separate experiment (experiment 2) in which we related a behavioral measure of loss aversion to task performance. In this way we were able to uncover the specific mechanism involved in encoding an endogenous reference point of value, and shed light on how it influences skilled task performance. This study highlights how neuroscience methods can provide insights of economic behavior: one of the major goals of the burgeoning field of neuroeconomics. Economists have long pondered the question of how best to design incentive contracts that pay workers just enough to fully maximize their performance (Smith, 1776). Standard models of these contracts assume that both the principal (manager) and agent (worker) act rationally and in a fashion that maximizes their individual utility (Laffont and Martimort, 2001). Under this assumption it follows that a worker’s performance should monotonically increase with pay.