Integrating context into decision making

将情境融入决策

• 批准号: RGPIN-2018-05946
• 负责人: Cunningham, William
• 金额: $ 6.85万
• 依托单位: University of Toronto
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2022
• 资助国家: 加拿大
• 起止时间: 2022-01-01 至 2023-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=764250
• 关键词: Integrating; context; into; decision; making

Adaptive behaviour in the world necessitates integrating relevant information from multiple sources and exploiting this information to inform decision processes. Importantly, how an individual optimally integrates these sources of information depends on multiple factors such as environmental factors, current need states, and considered goals. The aim of this project is to understand how the human brain represents contexts and contextually-embedded decision cues, the processes by which these sources of information are integrated to drive choice, and how outcomes are valued with respect to these contexts. In this proposal, we extend previous work by considering how people use internal (current need states) and external context cues (affordances in the environment) to modify the valuation of cues to guide choice, and how these signals unfold dynamically in time. Concretely, we examine (a) how information about changing needs/goals is represented in the brain, (b) how these needs/goals modulate valuation processes, (c) how these neural value signals predict individual choices, and (d) how people discover changes in context to guide these processes.Four fMRI studies are proposed in two projects. The first project will examine how and when people select between different goals when multiple goals must be balanced based on changing needs. People often need to select between multiple goals and adjust actions and evaluations appropriately. One way to deal with this is by organizing different types of goals into a hierarchy of goals, with the higher-order goal of maximizing some form of reward (e.g. advancing in one's career, building a larger social network). Evaluating possible actions or goals can then be based on their weighted relative importance for higher-order goals, which may also depend on factors such as environmental demands or current needs. By using state-of-the-art statistical tools that have been developed in the lab, we will examine how brain systems involved in value (ventral striatum and ventromedial prefrontal cortex) may be modulated by need states, and how neural systems underlying value processing shape the choices that people make. Further, we will investigate how prefrontal regions contribute to representing need states, the current value of holding a goal, and the selection and implementation of new goals. We will use a novel paradigm and analytic strategy to determine how information about the external state is represented as an internal neural context, and critically how these signals guide choices. Then we will leverage these results to examine how people chose to explore their environments to explicitly build mental context models. Our novel experimental design and analytic approach will enable us to explicitly dissociate the neural processes underlying the exploitation of a current context model from those associated with the exploration of contexts.

世界上的适应行为需要整合来自多个来源的相关信息，并利用这些信息为决策过程提供信息。重要的是，一个人如何最佳地整合这些信息来源取决于多种因素，如环境因素，当前的需求状态和考虑的目标。该项目的目的是了解人类大脑如何表示上下文和上下文嵌入的决策线索，这些信息源被整合以驱动选择的过程，以及如何根据这些上下文来评估结果。在这个建议中，我们扩展了以前的工作，考虑人们如何使用内部（当前的需求状态）和外部环境线索（在环境中的启示），以修改线索的估值，以指导选择，以及这些信号如何动态地展开的时间。具体而言，我们研究了（a）关于不断变化的需求/目标的信息如何在大脑中呈现，（B）这些需求/目标如何调节评价过程，（c）这些神经价值信号如何预测个人选择，以及（d）人们如何发现环境的变化来指导这些过程。第一个项目将研究当多个目标必须根据不断变化的需求进行平衡时，人们如何以及何时在不同的目标之间进行选择。人们经常需要在多个目标之间进行选择，并适当地调整行动和评估。解决这个问题的一种方法是将不同类型的目标组织成一个目标层次，其中更高层次的目标是最大化某种形式的奖励（例如，在职业生涯中取得进步，建立更大的社交网络）。评估可能的行动或目标可以基于它们对高阶目标的加权相对重要性，这也可能取决于环境需求或当前需求等因素。通过使用实验室开发的最先进的统计工具，我们将研究涉及价值的大脑系统（腹侧纹状体和腹内侧前额叶皮层）如何受到需求状态的调节，以及价值处理的神经系统如何塑造人们做出的选择。此外，我们将研究前额叶区域如何有助于代表需要状态，持有目标的当前价值，以及新目标的选择和实施。我们将使用一种新的范式和分析策略来确定外部状态的信息如何被表示为内部神经环境，以及这些信号如何引导选择。然后，我们将利用这些结果来研究人们如何选择探索他们的环境，以明确地建立心理背景模型。我们新颖的实验设计和分析方法将使我们能够明确地将利用当前上下文模型的神经过程与探索上下文的神经过程分离开来。