Abstract The standard decision-theoretic model separates information and preferences. We argue that some behavioral phenomena in information processing (e.g., polarization) are closely related to this separation. We propose a model in which agents receive vague information, which they refine when they are required to make probabilistic judgements, or to take an action. By doing so, they also refine their beliefs about some traits or states (e.g., their information-processing ability), which affects their utility.

中文翻译：

---

模糊信息下的偏好与信息处理

摘要 标准的决策理论模型将信息和偏好分开。我们认为信息处理中的一些行为现象（例如极化）与这种分离密切相关。我们提出了一个模型，其中代理接收模糊信息，当需要他们进行概率判断或采取行动时，他们会对其进行改进。通过这样做，他们还改进了他们对某些特征或状态（例如，他们的信息处理能力）的信念，这会影响他们的效用。