CAREER: Neural and Computational Basis of Guilt in Decision-Making

情绪是人类经历的一个重要方面，它直接影响我们如何做出决定，如何建立人际关系，以及我们更广泛的心理和身体健康。当我们考虑到未来的目标、过去的经历和现在的状态时，对世界的评估就会产生情绪。例如，人类有与他人联系的基本需求，内疚等情绪可以提供信号，帮助指导我们的行为，通过最大限度地减少我们的行为对他人的负面影响来实现这些更广泛的社会目标。然而，这些内在体验目前只能通过内省来主观评估，这限制了我们理解大脑如何产生这些独特感觉以及这些体验在个体之间的整体一致性的能力。通过一系列的研究，我们将：(a)激发参与者的内疚感，让他们做出减少他人痛苦的选择；(b)利用面部表情和大脑活动的计算模型，制定这些内疚体验的客观衡量标准；(c)评估这些内疚经历与其他类型的心理过程的关系，如感知他人的内疚，回忆过去的内疚经历，以及做出伤害他人的决定。这项工作对许多重要的、情绪化的、代价高昂的决策具有重要意义，这些决策可能会影响医疗保健、政治和商业等众多领域的其他人。该提案利用计算技术，通过大脑模式和面部表情来确定内疚体验的客观衡量标准。在目标1中，我们将阐明内疚厌恶是如何成为决策的核心动机，以尽量减少对他人的伤害，并可以通过模拟大脑活动模式客观地衡量。在目标2中，我们使用这些客观测量来更好地理解内疚体验与其他类型的心理过程（感知、记忆和想象内疚）之间的关系。在目标3中，我们将研究大脑信号是否能够独特地捕捉内疚的特征，或者这些经历是否也可以通过面部表情模式来测量。这项工作提供了一种独特的跨学科方法来提高我们对内疚的理解。首先，我们展示了通过自然的社会互动成功引发内疚感的能力，并在大脑活动模式中识别出一种共同的表征。这将允许我们对这一测量进行结构验证，以表征内疚体验与其他相关心理体验（如观察他人痛苦、回忆以前的内疚体验或做出伤害他人的决定）之间的关系。第三，该提案将评估内疚是否在大脑信号中唯一编码，或者是否可能在其他下游信号中表现出来，如面部表情模式。如果成功，这项工作可能会通过克服自我报告在研究情感体验方面的局限性，对情感领域产生变革性的影响。这一建议也可以为评估通过面部表情捕捉心理体验的神经成像替代低成本测量的效用提供重要的见解。最后，这项工作可能对更广泛的决策产生影响；我们对最小化罪恶感的偏好可能是如此强烈，以至于压倒了与决策结果相关的其他成本。该建议还纳入了协同教育培训和外联工作，以扩大工作范围，包括一个新的强化夏季培训计划；拓展社区服务组织；为本科生和医学院学生开设新课程。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Emotions are a critical aspect of the human experience that directly impact how we make decisions, how we form interpersonal relationships, and our broader mental and physical health. Emotions result from making evaluations about the world while considering our future goals, past experiences, and current states. For example, humans have a basic need to connect with others, and emotions such as guilt can provide signals that help guide our behavior to meet these broader social goals by minimizing the negative impact of our actions on others. However, these internal experiences can currently only be subjectively assessed via introspection, which has limited our ability to understand how our brain generates these unique feelings and the overall consistency of these experiences across individuals. Through a series of studies, we will: (a) elicit feelings of guilt in participants tasked with making choices to reduce another person's suffering; (b) develop objective measures of these guilt experiences using computational models of facial expressions and brain activity; and (c) evaluate how these guilt experiences relate to other types of psychological processes such as perceiving guilt in others, remembering past experiences of guilt, and making decisions to harm others. This work has important implications for a number of consequential, emotional, and costly decisions that might impact others in a multitude of domains such as healthcare, politics, and business.This proposal leverages computational techniques to identify objective measures of guilt experiences using brain patterns and facial expressions. In Aim 1, we will elucidate how guilt-aversion is a central motivation in making decisions to minimize harm to others and can be objectively measured by modelling patterns of brain activity. In Aim 2, we use these objective measures to better understand how guilt experiences relate to other types of psychological processes (perceiving, remembering, and imagining guilt). In Aim 3, we will examine if brain signals are uniquely able to capture signatures of guilt, or if these experiences can also be measured using patterns of facial expressions. This work provides a unique interdisciplinary approach to improve our understanding of guilt. First, we demonstrate our ability to successfully elicit guilt using a naturalistic social interaction and identify a common representation in a pattern of brain activity. This will allow us to perform construct validation on this measure to characterize how the experience of guilt relates to other related psychological experiences such as observing others in pain, recalling a previous guilty experience, or making decisions to harm others. Third, this proposal will evaluate whether guilt is uniquely encoded in brain signals, or if it might be manifested in other downstream signals such as patterns of facial expressions. If successful, this work might have a transformative impact on the field of emotion by overcoming the limitations of self-report in studying emotional experiences. This proposal can also provide important insight into evaluating the utility of alternative low-cost measurements to neuroimaging in capturing psychological experiences via facial expressions. Finally, this work could have implications for decision-making more broadly; our preferences for minimizing guilt may be so strong that they overwhelm the other costs associated with decision outcomes. This proposal also incorporates synergistic educational training and outreach to broaden the work, including a new intensive summer training program; outreach to community service organizations; and new curriculum for undergraduate and medical students.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.