Mixture Latent Markov Modeling: Identifying and Predicting Unobserved Heterogeneity in Longitudinal Qualitative Status Change
Mixture Latent Markov Modeling: Identifying and Predicting Unobserved Heterogeneity in Longitudinal Qualitative Status Change
复制标题
DOI：
10.1177/1094428109357107
复制
发表时间：
2011-07
影响因子：
9.5
通讯作者：
Mo Wang;D. Chan
中科院分区：
文献类型：
--
作者：
Mo Wang;D. Chan
在组织研究的许多领域中，我们可能会关注状态变化配置文件中的子组差异。本文的目的是使用一个关于退休人员退休后就业状态的真实数据集，通过对纵向质变中未观察到的异质性进行建模，说明混合潜在马尔可夫模型如何应用于组织环境中的实质性研究，以识别具有不同状态变化特征的人口亚组并检验其相关性。重点介绍了建模过程中的步骤，并讨论了该技术的限制、注意事项、建议和扩展。
There are many areas of organizational research where we may be concerned with subgroup differences in status change profiles. The purpose of this article is to illustrate, using a real data set on retirees’ postretirement employment statuses (PES), how mixture latent Markov modeling may be applied to substantive research in organizational settings to identify population subgroups with varying status change profiles and examine their correlates, by modeling unobserved heterogeneity in longitudinal qualitative changes. Steps in the modeling process are highlighted and limitations, cautions, recommendations, and extensions of the technique are discussed.