Choice with imprecise information: an experimental approach
Choice with imprecise information: an experimental approach
复制标题
DOI：
10.1007/s11238-008-9119-x
复制
发表时间：
2010-09-01
影响因子：
0.8
通讯作者：
Wada, Ryoko
中科院分区：
文献类型：
--
作者：
Hayashi, Takashi;Wada, Ryoko
本文提供了一个实验分析的态度不精确和可变的信息。不精确的信息以一组可能的概率值的形式提供，使得受试者几乎不可能猜测或估计该组中的哪一个是真的或更可能是真的。我们研究如何几何特征的信息片影响的选择。我们发现，主题关心更多的功能比对最好的情况下，最坏的情况下，这是一个反证著名的模型，极大极小和α-极大极小。我们发现，集合中非极值点的存在会影响选择，这表明人们对不精确性的态度是“非线性的”。“我们还得到了一个观察结果，尽管不重要，但信息片段具有互补性，这可能无法用贝叶斯观点来解释。
This article provides an experimental analysis of attitude toward imprecise and variable information. Imprecise information is provided in the form of a set of possible probability values, such that it is virtually impossible for the subjects to guess or estimate, which one in the set is true or more likely to be true. We investigate how geometric features of such information pieces affect choices. We find that the subjects care about more features than the pairs of best-case and worst-case, which is a counter-evidence to the well-known models, maximin and alpha-maximin. We find that presence of nonextreme points in the set affects choice, which suggests that attitude toward imprecision is 'nonlinear.' We also obtain an observation, though not significant, that information pieces have a complementarity that may not be explained by the Bayesian view.