RUI: Collaborative Research: Sample Size Bias in Judgments of Averages

这个项目关注的是一种被称为样本大小偏差的定量认知现象，这种现象发生在人们对平均值做出判断时。具体来说，随着集合中项目数量的增加，人们会判断平均值更大。例如，一组十个数字的平均值被判断为大于五个数字的平均值-即使两组数字的平均值相同。但样本量偏差也发生在风险等概念维度上。例如，十个人的平均心脏病发作风险被判断为大于五个人的平均心脏病发作风险。这个项目的核心是一系列的实证研究，其中一个横截面的人判断在各种条件下的几种平均值。这些研究有两个目的。第一个是测试一个理论，该理论假设集合大小自动激活大脑中的大小模拟表征，这反过来又影响平均值的表征和判断。第二个问题是，样本量偏差是否会影响对平均值的其他判断，包括对个人经验和行为的判断。首先，被测试的理论有可能帮助整合样本量偏差与不同研究领域的类似现象，包括判断和决策，感知和定量认知的神经科学。其次，这项研究将有助于确定人们对平均值的判断可能反映样本大小偏差的真实环境。例如，调查通常要求人们对他们或他们的同龄人的平均或典型情况做出判断（例如，他们每年使用保健服务的平均次数）。重要的是要知道这种判断是否可能是系统性的偏见，以及如何纠正这种偏见。最后，本项目将为加州州立大学、弗雷斯诺和阿巴拉契亚州立大学的本科生和硕士生提供高质量的研究培训和经验。这反过来又将使他们准备在各种领域，认知和社会心理学，判断和决策以及管理科学的任何博士研究取得成功。

This project focuses on a phenomenon of quantitative cognition called sample size bias, which occurs when people make judgments of averages. Specifically, as the number of items in a set increases, people judge the average to be greater. For example, the average of a set of ten numbers is judged to be greater than the average of five numbers -- even when the two sets have the same average. But sample size bias also occurs for conceptual dimensions such as risk. For example, the average heart-attack risk of ten people is judged to be greater than the average heart-attack risk of five people. The core of this project is a series of empirical studies in which a cross-section of people judge several kinds of averages under a variety of conditions. These studies have two purposes. The first is to test a theory that assumes that set size automatically activates an analog representation of magnitude in the brain, which in turn affects the representation and judgment of the average. The second is to see whether the sample size bias affects other kinds of judgments about averages, including judgments about one's own experience and behavior.This work is important for a number of reasons. First, the theory being tested has the potential to help integrate sample size bias with similar phenomena from distinct areas of research including judgment and decision making, perception, and the neuroscience of quantitative cognition. Second, this research will help identify real-world contexts in which people's judgments of averages are likely to reflect sample size bias. For example, surveys often ask people to make judgments about what is average or typical for them or their peers (e.g., the average number of times they use health services per year). It is important to know whether such judgments are likely to be systematically biased and how such biases might be corrected. Finally, this project will provide high quality research training and experience to a diverse group of undergraduate and master's-level students at California State University, Fresno and Appalachian State University. This in turn will prepare them to succeed in doctoral studies in any of a variety of areas, cognitive and social psychology, judgment and decision making, and management science.