Archetypes and Prototypes of Decisions under Uncertainty

本项目通过一系列决策实验来收集数据，以理解不确定性下的选择。该项目还开发了不确定性下选择的理论分析工具，以及混合参数和非参数计量经济学技术，用于用新数据测试理论。该项目的理论和实验方面提供了对风险偏好和实施它们的选择的更丰富的理解。实验提供了在不确定环境下的选择的积极解释，允许对个人行为模式进行一般表征。先前的数据表明，尽管在选择行为中存在相当大的异质性，但参与者的行为方式确实是一致的，因为选择最大化了一组一致的偏好。然而，受试者也使用可识别的通用启发式（经验法则）来做决定。这两种不同的行为反映了两种不同的理性：实体理性和程序理性。新理论将这两种不同的观点结合到一个模型中，研究团队将使用新理论来开发一种方法，以确定人们何时使用程序理性或实体理性做出决策。了解人们如何在危险的情况下做出决定是一个重要的跨学科研究问题，对公共政策也有重要的影响。研究人员正在开发的实验平台和分析技术将对研究界广泛有用。

This project uses a series of decision making experiments to gather data for understanding choice under uncertainty. The project also develops tools for the theoretical analysis of choice under uncertainty and a blend of parametric and non-parametric econometric techniques for testing the theory with the new data. The theoretical and experimental aspects of the project provide a richer understanding of risk preferences and the choices that implement them.The experiments provide a positive account of choice under uncertainty in an environment that allows for a general characterization of the patterns of individual behavior. Previous data show that although there is considerable heterogeneity in choce behavior, participants do behave in a way that is consistent in the sense that choices maximize a consistent set of preferences. However, subjects also use identifiable common heuristics (rules of thumb) to make decisions. These two different kinds of behavior reflect two different kinds of rationality: substantive rationality and procedural rationality. The new theory combines these two different ideas into a single model and the research team will use the new theory to develop a method to identify when people make decisions using procedural or substantive rationality.Understanding how people make decisions in risky situations is an important and interdisciplinary research question that also has important implications for public policy. The experimental platform and the analytical techniques that the researchers are developing will be broadly useful to the research community.