Integrated Cognitive and Decision Models with Applications to Aging and Numeracy

项目总结/摘要 我们的主要方法/理论目标是开发模型，整合信息如何被 用决策模型来表示。应用领域将是计算能力：如何计算 是在认知系统中表现出来的，以及它是如何用于决策的。计算能力的具体目标 理解计算技能所涉及的基本过程，并制定措施， 了解基本流程和高级技能之间的关系。我们将检查统计数据 综合模型的属性，并与公开可用的扩散模型拟合的作者合作 软件包，使集成模型广泛可用。 我们将使用的决策模型是公认的扩散决策模型，它提供了一个统一的 解释我们将使用的简单两项选择任务的全部数据，包括分布 正确和不正确响应的响应时间，分布和准确性如何随 独立变量的操作，以及它们如何随着个体和群体之间的差异而变化。 个体我们将测试的表示模型假设数字上的分布式表示， 将产生驱动决策过程的计算能力的测量。一个成功的代表性模型将是 一个是当通过扩散模型转换时，测量准确预测响应时间， 准确度数据 拟议中的研究将通过将模型应用于简单的算术来衡量低水平的技能 辨别和记忆任务。然后，将研究这些技能、工作记忆和 处理速度以及技能和高级数学标准化考试成绩之间的关系 能力，财务计算能力和健康计算能力。所有这些任务和措施， 在大学年龄的成年人、轻度认知障碍的成年人、早期和中度阿尔茨海默氏症的成年人中进行了测试， 疾病和认知未受损的年龄匹配的成年人。问题将包括计算能力如何下降 以及他们如何受到认知障碍的影响。可能是计算能力的缺陷 能力先于记忆障碍，如果是这样，导致早期治疗和早期干预健康， 财务决策。 1

Project Summary/Abstract Our primary methodological/theoretical goal is to develop models that integrate models of how information is represented with models of how decisions are made. The domain of application will be numeracy: how numeracy is represented in the cognitive system and how it is used in making decisions. The specific goals for numeracy are to understand the basic processes involved in numeracy skills and to produce measures with which to understand the relationships between basic processes and higher-level skills. We will examine statistical properties of the integrated models and work with the authors of publicly available diffusion-model fitting packages to make the integrated models widely available. The decision model we will use is the well-established diffusion decision model, which provides a unified explanation of the full range of data for simple two-choice tasks of the kinds we will use, including the distributions of response times for correct and incorrect responses, how the distributions and accuracy change with manipulations of independent variables, and how they change with differences among individuals and groups of individuals. The representation models we will test assume distributed representations over numerosity and they will produce the measures of numeracy that drive decision processes. A successful representation model will be one for which the measures accurately predict, when translated through the diffusion model, response time and accuracy data. The proposed research will examine low-level skills, as measured by applying the models to simple numeracy discrimination and memory tasks. Relations will then be examined between these skills, working memory, and speed-of-processing and also between the skills and performance on standardized tests of higher-level math abilities, financial numeracy abilities, and health numeracy abilities. All of these tasks and measures will be tested with college-age adults, adults with Mild Cognitive Impairment, adults with early and moderate Alzheimer's disease, and cognitively unimpaired age-matched adults. Questions will include how numeracy abilities decline with normal aging and how they are impacted by cognitive impairment. It may be that impairments in numeracy abilities precede memory impairments and if so, lead to earlier treatments and earlier interventions in health and financial decision making. 1