In three separate experiments, we examined the reliability of and relationships between self-report measures and behavioral response time measures of reward sensitivity. Using a rewarded-Stroop task we showed that reward-associated, but task-irrelevant, information interfered with task performance (MIRA) in all three experiments, but individual differences in MIRA were unreliable both within-session and over a period of approximately 4 weeks, providing clear evidence that it is not a good individual differences measure. In contrast, when the task-relevant information was rewarded, individual differences in performance benefits were remarkably reliable, even when examining performance one year later, and with a different version of a rewarded Stroop task. Despite the high reliability of the behavioral measure of reward responsiveness, behavioral reward responsiveness was not associated with self-reported reward responsiveness scores using validated questionnaires but was associated with greater self-reported self-control. Results are discussed in terms of what is actually being measured in the rewarded Stroop task.

在三个独立的实验中，我们检查了自我报告措施和奖励敏感性的行为反应时间措施之间的可靠性和关系。使用奖励 Stroop 任务，我们发现与奖励相关但与任务无关的信息在所有三个实验中都会干扰任务绩效 (MIRA)，但 MIRA 中的个体差异在会话内和大约 4 周的时间内都不可靠, 提供了明确的证据表明它不是一个好的个体差异度量。相比之下，当与任务相关的信息得到奖励时，绩效收益的个体差异非常可靠，即使在一年后检查绩效时，并且使用不同版本的奖励 Stroop 任务也是如此。尽管奖励响应的行为测量具有很高的可靠性，行为奖赏反应性与使用经过验证的问卷的自我报告奖赏反应性分数无关，但与自我报告的自我控制能力更强相关。结果根据奖励 Stroop 任务中实际测量的内容进行讨论。