Mathematical Modeling and Process Description of Group Decision Making

规范模型和描述性模型之间的冲突和解决一直是我们研究的重点。例如，比较了几种数学集结规则和简单多数规则在群决策情况下的效率。对于公共决策来说，衡量人们的实际想法的方法论是非常重要的，我们调查了通过语言表达来衡量主观概率的方法的可能性。我们的感觉是，我们的普通决策似乎达到了某种理性。从这一点出发，我们试图为受试者的典型反应找到理由，这些反应通常被认为是错误的反应。更具体地说，我们进行了三项研究，以考察这如何影响被试在顺序反应中的变化。一个是使用Wason的四张卡问题的实验。我们发现，数学规范模型预测，当先验概率很小时，规则本身是概率的，典型的错误反应是合理的，尽管真实受试者的数据并没有显示出满足这一规范预测的明确趋势。此外，我们还表明，在被称为“秘书问题”的连续决策任务中，受试者的反应变得更接近理性解决方案。在第三个任务中，达马西奥的“赌博任务”表明，受试者是否接近标准解取决于受试者的情绪反应，而情绪反应是通过皮肤电导反应来衡量的。

Conflicts and resolutions between normative models and descriptive models have been our research concern. For example, some mathematical aggregation rule and simple majority rule were compared in terms of efficiency in the group decision-making situation. For public decision making, the methodology to measure what people actually have in mind is very important and we investigated the possibility to apply the method of measuring subjective probability by means of language expression.Our feeling is that our ordinary decision making purports to reach some kind of rationality. From this point, we tried to find the justification for the subjects' typical responses, which are often regarded as the wrong responses. More concretely, we made three research efforts to investigate how this affects subjects change in the sequential responses. One is the experiment using Wason's four card problem. We showed that mathematical normative model predicts that when the prior probability is small and the rule itself is probabilistic, typical "wrong" responses can be justified, although real subjects' data did not show the clear tendency to meet this normative prediction. Also, we showed that in the sequential decision making task known as the "secretary problem", subjects' responses became closer to the rational solution. In the third task, Damasio's "gamble task", it was shown that whether subjects approached the normative solution or not depended subjects emotional responses which was measured by skin conductance responses.