Making predictions based on data: Holistic and atomistic procedures

• 成果类型: Article
• 署名作者: Glazer, Jacob; Rubinstein, Ariel
• 署名单位: Tel Aviv University; University of Warwick; Tel Aviv University; New York University
• 刊物名称: JOURNAL OF ECONOMIC THEORY
• ISSN/ISSBN: 0022-0531
• DOI: 10.1016/j.jet.2023.105791
• 发表日期: 2024
• 关键词: Experimental economics; Non-Bayesian beliefs; Data-based predictions

摘要

Subjects were asked to predict the choice made by a hypothetical individual, based on a small sample of his past choices in identical situations. Each of the alternatives facing the individual has a number of components. When the data is presented explicitly as a distribution of past choices, most subjects use the holistic procedure according to which the prediction is the most common choice made in the past. Nevertheless, a significant minority of subjects use atomistic procedures, which relate to the various components individually. When the distribution is not explicitly presented but can be derived from the data, subjects apply both atomistic procedures and holistic rules of thumb.