Brown University, Providence, RI.
Georgetown University, Washington, DC.
Am Econ Rev. 2016 Jun;106(6):1562-5. doi: 10.1257/aer.106.6.1562.
The purpose of this document is to update and correct Figure 4 from "Optimal Expectations and Limited Medical Testing: Evidence from Huntington Disease" (Oster, Shoulson, and Dorsey 2013). This figure documents how perceptions about the risk of HD evolve with symptoms. It compares these perceptions with the "actual risk" of HD based on a Bayesian updating calculation described in the paper. The construction of Figure 4 is correctly described in the text of the paper and the data on perceptions are documented correctly. However, the construction of the "actual risk" series is not accurate. There are two central issues. First, there were data limitations at the time of publication which have since been relaxed and the better data now available changes the picture. Second, there was an error in the construction of Figure 4 which should have been recognized at the time. We detail the issues here and include the corrected figure. The original figure showed evidence of overoptimism at all levels of motor score. The corrected figure shows that for low symptom levels individuals are correct about their risk level, whereas those with more advanced symptoms are overly optimistic. Overall, the levels of overoptimism are lower than documented originally. We will briefly discuss the implications for the theory at the end of this document.
本文件的目的是更新和更正“最佳期望与有限医学检测：亨廷顿舞蹈病的证据”（Oster、Shoulson 和 Dorsey 2013）中的图 4。该图记录了人们对 HD 风险的看法如何随症状而变化。它将这些看法与基于论文中描述的贝叶斯更新计算的“实际风险”进行了比较。图 4 的构建在论文的正文中有正确的描述，并且感知数据有正确的记录。但是，“实际风险”系列的构建并不准确。有两个核心问题。首先，在发表时存在数据限制，这些限制现已放宽，现在可用的更好的数据改变了情况。其次，在构建图 4 时存在错误，当时就应该认识到这一点。我们在此详细说明这些问题，并包括更正后的图表。原始图表显示在所有运动评分水平都存在过度乐观的证据。更正后的图表表明，对于低症状水平的个体，他们对自己的风险水平是正确的，而那些症状更严重的个体则过于乐观。总体而言，过度乐观的程度低于最初记录的程度。我们将在本文的最后简要讨论这对理论的影响。
