Abstract: Cognitive control refers to people’s ability to adaptively employ cognitive resources and adjust cognitive processes in pursuit of goal-directed behavior. Since naturally occurring behavioral situations are constantly changing, people would mobilize their control adaptively. According to the Expected Value of Control (EVC) model, the dynamic adjustment of control can be thought of as value-based decision making, centered on the integration of rewards and costs that can be expected from a control-demanding task. Hence, reward and cost are two key factors jointly modulating people’s motivation and determining the allocation of control. Following this framework but going beyond the EVC model, the current review elucidated the role of various motivation-related factors that can act as rewards or costs in the implementation of cognitive control, and discussed how they collectively adjust cognitive control.
More specifically, money, juice, or emotional/social stimuli are extrinsic rewards that can drive cognitive control and improve task performance, albeit with a few exceptions. Considering this complexity, other factors can further modulate the beneficial effects (e.g., reward-poor vs. reward-rich task conditions, the congruity of reward and task performance, and individual reward sensitivity). Besides, in contrast to extrinsic rewards that are manipulated externally, intrinsic rewards are highly integrated into control-related tasks. It can be reflected in people’s autonomic engagement with certain tasks and the positive emotions they generated. In this sense, the investigation of the influence of intrinsic rewards on cognitive control is relatively indirect, which can be achieved by adjusting effort levels and positive emotions. Relatedly, individual differences in intrinsic motivation, as reflected by the need for cognition (NFC), are also closely tied to intrinsic rewards in driving control. That is, individuals high in need for cognition are more inclined to be involved in control demanding tasks and to persist in difficult or unprofitable cognitive tasks.
Meanwhile, due to the presence of cognitive costs associated with exerting cognitive control, individuals typically show a bias toward opting for “low-effort” tasks, while decreasing the subjective value of the expected value. When discussing the impact of costs on cognitive control, it is necessary to consider the trade-off between rewards and costs. Previous studies have demonstrated that this trade-off process may vary among individuals based on their willingness to invest effort and their capacity to exert effort, depending on whether they place a higher value on rewards or costs. Consequently, we have further delineated the control signal intensity to effort levels and introduced the concept of “Subjective Expected Value of Control”, which is determined by the difference between the Subjective Value of Reward and the Subjective Value of Cost. Furthermore, the reward-cost trade-off is inherently dynamic, with individuals adapting their cognitive control with the automaticity of task performance in a given task, or in response to feature transfer across different task situations.
Nonetheless, some unanswered questions need to be further investigated. Firstly, the mechanism underlying the reward-cost trade-off requires refinement. As individuals persistently allocate control, their instantaneous subjective evaluation of the rewards and costs expected from the current task may change dynamically. Although several theories have introduced dynamic elements to the EVC model in various ways, a fully dynamic representation of the reward-costs trade-off remains a topic of ongoing exploration. Secondly, the subjective trade-off between rewards and costs can be further modulated by additional individual factors closely related to external and internal motivations. Consequently, it is intriguing to explore how individual differences in reward sensitivity, cognitive need, intrinsic motivation, and opportunity costs may dynamically impact subjective evaluation of the rewards and costs of investing cognitive effort.

Key words: cognitive control, EVC, reward, cost, reward-cost trade-off

摘要： 根据控制的预期价值（expected value of control, EVC）模型,回报与代价通过调节动机共同影响认知控制的实施与分配（即回报-代价权衡）。在EVC模型外,还有许多探究动机相关因素如何影响认知控制的研究。文章首先总结梳理了可被视为回报或代价的因素及相关的实证证据;其次,基于回报-代价权衡简要讨论二者如何共同影响认知控制;最后提出未来研究可以从客观（优化权衡的计算模型）与主观（个体差异）两个方面更全面地解析回报-代价的权衡过程,以实现对认知控制的有效调控与精准干预。

关键词: 认知控制, EVC模型, 回报, 代价, 回报-代价权衡