Han Sapphire Yu, Liefbroer Aart C, Elzinga Cees H
Netherlands Interdisciplinary Demographic Institute (NIDI/KNAW), The Hague, the Netherlands; University of Groningen, the Netherlands.
Netherlands Interdisciplinary Demographic Institute (NIDI/KNAW), The Hague, the Netherlands; University Medical Centre Groningen, University of Groningen, The Netherlands; Vrije Universiteit Amsterdam, the Netherlands.
Adv Life Course Res. 2020 Mar;43:100265. doi: 10.1016/j.alcr.2019.03.001. Epub 2019 Jul 29.
Life courses consist of complex patterns of correlated events and spells. The nature and strength of these correlations is known to depend on both micro- and macro- covariates. Life-course models such as event-history analysis and sequence analysis are not well equipped to deal with the processual and latent character of the decision- making process. We argue that Hidden Markov Models satisfy the requirements of a life course model. To illustrate their usefulness, this study will use Hidden Markov chains to model trajectories of family formation. We used data from the Generations and Gender Programme to estimate Hidden Markov Models. The results show the potential of this approach to unravel the mechanisms underlying life-course decision making and how these processes differ both by gender and education.
生命历程由相互关联的事件和阶段的复杂模式组成。已知这些关联的性质和强度取决于微观和宏观协变量。诸如事件史分析和序列分析等生命历程模型并不擅长处理决策过程的过程性和潜在特征。我们认为隐马尔可夫模型满足生命历程模型的要求。为了说明其有用性，本研究将使用隐马尔可夫链对家庭形成轨迹进行建模。我们使用了代际与性别项目的数据来估计隐马尔可夫模型。结果显示了这种方法在揭示生命历程决策背后的机制以及这些过程在性别和教育方面如何不同的潜力。
