Evaluating Probabilistic Alternatives for Unique and Multiple Occurring Events

实验表明，人们对概率结果的评估并不符合规范理论，特别是当他们被要求提供单个孤立事件发生的概率时。然而，有证据表明，当概率指的是唯一而不是多次发生的事件时，人们的反应会有所不同。这个项目探讨观察到的差异是否与概率只发生一次的事件的情况有关。这项工作很重要，因为它将提供决策者在做出实际选择时通常违反预期效用理论的结论的概括性信息。该研究将调查将风险前景视为多次发生的事件是否会导致长期观点，从而导致概率信息的处理更符合规范原则。与这一假设相一致的是，研究者和其他人最近发现的证据表明，当受试者评估多重发生的前景时，一些系统性违反规范理论的行为会减少。将进行一系列实验，以调查人们对单一和多次发生的事件的反应。前两个实验检验了多重玩法选项对主观效用和概率的影响。另一个实验着眼于多人游戏是否会减少框架效应。最后的实验将检验经验丰富的实验对象如何感知财务审计背景下的概率事件。

Experiments have shown that people's evaluations of probabilistic outcomes do not conform to normative theory, especiallywhen they are asked to provide the probability that a single, isolated event will occur. However, there is evidence that people responddifferently when probabilities refer to unique rather than multiply occurring events. This project explores whether the observed discrepancies are tied to situations in which probabilities apply to events that occur only once. This work is important because it will provide information on the generalizability of the conclusion that decision makers generally violate expected utility theory when making actual choices. The study will investigate whether framing risky prospects as multiply occurring events induces a long run perspective that results in the processing of probabilistic information more in accordance with normative principles. Consistent with this hypothesis is recent evidence found by the investigators and others that some of the systematic violations of normative theory are reduced when subjects evaluate multiply occurring prospects. A series of experiments will beconducted to investigate people's responses to single and multiply occurring events. The first two experiments examine the impact of the option of multiple play on subjective utilities and probabilities. Another experiment looks at the question of whether opportunity for multiple play reduces framing effects. A final experiment will examine how experienced experimental subjects perceive probabilistic events ina financial audit context.