Measuring natural source dependence
提出了一个可操作的自然源依赖定义，通过匹配不同源的主观概率函数来测量个体对不确定性来源的偏好，并提供了基于有限观测的估计方法，适用于分离态度与信念的常见实验设计。
Abstract The consequences of most economic decisions are uncertain; they are conditional on events with unknown probabilities that decision makers evaluate based on their beliefs. In addition to consequences and beliefs, the context that generates events—the source of uncertainty—can also impact preferences, a pattern called source dependence. Despite its importance, there is currently no definition of source dependence that allows for comparisons across individuals and sources. This paper presents a tractable definition of source dependence by introducing a function that matches the subjective probabilities of events generated by two sources. It also presents methods for estimating such functions from a limited number of observations that are compatible with commonly-used choice-based approaches for separating attitudes from beliefs. As an illustration, we implement these methods on three datasets, including two original experiments, and show that they consistently capture clear, albeit heterogeneous, patterns of source dependence between natural sources. Our approach provides a framework for future research to explore how source dependence varies across individuals and situations.