Experimental examinations and a modeling study of multiattribute decision making

关于上下文依赖选择的三个被广泛研究的发现值得特别关注，因为它们构成了对理性选择至关重要的公理的违反。这些发现包括吸引效应、相似效应和妥协效应。在多属性决策中，所有这些影响都是在两个选项集合中增加第三个选项时发生的。144名大学生参加了一项实验，其中提出了20个假设购买问题。这些问题包括按照两个属性描述的三个备选方案。操纵选择比例、置信度和反应时间的第三种选择会产生显著的影响。这些结果支持我们的多属性、多选择选择过程的随机比较分组模型。我们提出了多选择决策的随机比较分组理论来解释RATI…的三个上下文诱导的违反更多的民族选择。用随机比较分组来解释吸引和相似效应，根据随机比较分组，相似的方案比不相似的方案更容易被比较。除了比较分组机制外，折衷效应还被认为是根据基本的心理物理功能来感知善的假设。为了进一步检验我们模型的理论假设，我们设计了一个实验来研究眼动在多属性、多选择过程中的作用。我们使用眼动记录器来考察这些背景效应在多属性决策中的潜在认知过程。17名大学生参加了一项实验，他们被呈现了12个假想的购买问题，其中包括三个沿着两个属性描述的备选方案。对选择比例、注视持续时间和扫视频率的第三种选择进行操控会产生显著的影响。这些结果支持我们的多属性决策过程的随机比较分组模型。较少

Three much-studied findings regarding context-dependent choice warrant special attention since they constitute violations of axioms that are fundamental to rational choice. These findings include the attraction effect, the similarity effect, and the compromise effect. In multiattribute decision-making, all these effects occur with the addition of a third alternative to a two-alternative choice set. One hundred and forty-four undergraduates participated in an experiment wherein twenty hypothetical purchase problems were presented. The problems comprised three alternatives described along two attributes. Significant effects resulted from manipulating the third alternative for choice proportions, confidence ratings, and reaction time. These results support our stochastic comparison-grouping model of multiattribute, multialternative choice processes.We propose a stochastic comparison-grouping theory of multialternative decision-making to explain the three context-induced violations of rati … More onal choice. The attraction and similarity effects are explained by stochastic comparison grouping, according to which similar alternatives are compared more often than dissimilar ones. The compromise effect is explained by the assumption that goodness is perceived according to a basic psychophysical function, in addition to the comparison grouping mechanism.In order to further test the theoretical assumptions of our model, we formulated an experiment to study the role of eye movement in multiattribute, multialternative processes. We used an eye movement recorder to examine the underlying cognitive process of these context effects in multiattribute decision-making. Seventeen undergraduates participated in an experiment where they were presented with twelve hypothetical purchase problems comprising three alternatives described along two attributes. Significant effects resulted from manipulating the third alternative for choice proportions, fixation duration, and frequency of saccade. These results support our stochastic comparison-grouping model of multiattribute decision-making processes. Less