A dynamic model for ranking-based conjoint analysis with no-choice options
A dynamic model for ranking-based conjoint analysis with no-choice options
复制标题
具有无选择选项的基于排名的联合分析的动态模型
DOI：
10.1007/s41237-022-00178-8
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Takeuchi Makito
中科院分区：
文献类型：
--
作者：
Igari Ryosuke;Takeuchi Makito
在排序数据分析中，通常在每次试验中获得所有偏好等级。然而，如果有一个没有选择的选项，如联合配置文件的替代品，排名结束，并部分排名数据，结构取决于个人和试验，在重复选择任务中观察。此外，在重复测量中，通常假设每个消费者的部分价值效用在所有试验中是恒定的，但由于受访者的疲劳，学习和调查经验，偏好可能会在试验期间发生变化。我们提出了一个贝叶斯动态排序logit模型，没有选择的选项，并将其应用到智能手机的排名为基础的联合获得的数据分析。结果表明，部分价值效用的动态和静态模型之间的差异，部分价值效用和属性的相对重要性随着试验的进展而变化。
In ranking data analysis, it is common for all preference ranks to be obtained in each trial. However, if there is a no-choice option in alternatives such as conjoint profiles, the ranking ends there, and partial ranking data, with a structure that depends on the individuals and trials, in repeated choice tasks are observed. Moreover, in repeated measurements, it is often assumed that the part-worth utilities of each consumer are constant throughout all trials, but preferences may change over the duration of the trials due to the respondents’ fatigue, learning, and experience with the survey. We propose a Bayesian dynamic rank-ordered logit model with no-choice options and apply it to an analysis of data obtained by a ranking-based conjoint for smartphones. The results show that the part-worth utilities differ between the dynamic and static models, and that the part-worth utilities and relative importance of attributes change as trials progress.