Efficiency of evidence accumulation (EEA) as a higher-order, computationally defined RDoc construct

项目总结/摘要 与自我调节相关的认知结构，包括认知控制，注意力和工作记忆， 研究领域标准（RDoC）倡议的一个突出重点是表征个人 这些变异传递了精神障碍的风险。然而，这些概念中的许多都受到其模糊性的限制。 定义，与神经生物学的模糊联系，以及这些结构的假定测量具有 弱的心理测量特性，包括可靠性差和不连贯的因素结构。此外，一致 研究发现，患有多种精神疾病的人往往表现出非特异性的认知缺陷， 这一系列结构表明，与精神病理学相关的认知失常可能更好- 用高阶因子而不是离散函数来解释。我们建议评估是否有效， 证据积累（EEA）-一种认知结构，在计算 建模和神经生理学研究，但尚未与RDoC集成-可以克服许多 通过在RDoC矩阵中作为高阶因子来操作这些限制。EEA是一个核心机制， 证据积累模型（EAMs）-解释认知的主要数学框架 性能-在心理和神经生理学上都有精确的计算定义 分析水平，明确的生物相容性，以及强大的心理测量特性。先前的工作已经建立了 EEA是一个可靠的因素，可以解释各种各样的个人表现差异， 认知任务-从简单的决定到复杂的认知控制和工作记忆范式-并且 与自我调节困难相关的多种疾病中受损。我们认为EEA代表了一种更高层次的 在RDoC矩阵中占认知领域差异很大比例的因素 弱的EEA可能通过损害决策来传递多种精神病理学的风险 跨语境。EEA尚未与RDoC整合，尽管在 EEA与精神病理学有关，EEA在现实世界中的状态（受试者内）变化的相关性 上下文是未知的。我们建议评估EEA作为RDoC中候选高阶因子的作用 框架，并设置一个更大的程序，计算严格的研究EEA作为一个桥梁的阶段 神经生物学机制和现实世界的行为之间完成以下目标：1）定义 特征EEA作为RDoC矩阵中高阶认知域的结构和边界，2） 开发和试点工具，用于每日评估国家欧洲经济区及其与现实世界波动的关系 在环境因素和行为上。这个项目有可能以一种更好的方式改进RDoC， 代表精神病理学的认知风险因素（即，总任务和跨诊断关联 之间的认知和精神病理学结构），并允许它利用关键的好处， 计算模型，以增加RDoC构建和测量的精确度和生物相容性。

PROJECT SUMMARY/ABSTRACT Cognitive constructs relevant to self-regulation, including cognitive control, attention, and working memory, are a prominent focus of the Research Domain Criteria (RDoC) initiative to characterize dimensions of individual variation that convey risk for mental disorders. However, many of these constructs are limited by their vague definitions, ambiguous links to neurobiology, and evidence that putative measures of such constructs have weak psychometric properties, including poor reliability and an incoherent factor structure. Further, consistent findings that people with multiple psychiatric disorders tend to display non-specific cognitive deficits that span this array of constructs suggest that cognitive aberrations associated with psychopathology may be better- explained by a higher-order factor than by discrete functions. We propose to evaluate whether efficiency of evidence accumulation (EEA)—a cognitive construct that has been well-characterized in computational modeling and neurophysiological research but has yet to be integrated with RDoC—can overcome many of these limitations by operating as a higher-order factor within the RDoC matrix. EEA is a core mechanism of evidence accumulation models (EAMs)—a predominant mathematical framework for explaining cognitive performance— that has a precise computational definition across both psychological and neurophysiological levels of analysis, clear biological plausibility, and strong psychometric properties. Prior work has established EEA as a reliable factor that accounts for individual differences in performance across a wide variety of cognitive tasks—from simple decisions to complex cognitive control and working memory paradigms—and is impaired in multiple disorders linked to self-regulatory difficulties. We posit that EEA represents a higher-order factor that accounts for a substantial proportion of the variation across cognitive domains in the RDoC matrix and that weak EEA conveys risk for multiple psychopathologies, potentially by impairing decision making across contexts. EEA has yet to be integrated with RDoC and, although trait (between-subjects) variation in EEA is linked to psychopathology, the correlates of state (within-subjects) variation in EEA across real-world contexts are unknown. We propose to evaluate EEA’s role as a candidate higher-order factor in the RDoC framework and set the stage for a larger program of computationally rigorous research on EEA as a bridge between neurobiological mechanisms and real-world behavior by completing the following aims: 1) define the structure and boundaries of trait EEA as a higher-order cognitive domain in the RDoC matrix, 2) develop and pilot tools for daily assessment of state EEA and its relations with real-world fluctuations in contextual factors and behavior. This project has the potential to refine RDoC in a way that that better represents cognitive risk factors for psychopathology (i.e., task-general and transdiagnostic associations between cognition and psychopathology constructs) and allows it to leverage key benefits of well-established computational models to increase the precision and biological plausibility of RDoC constructs and measures.