Scholarlink

<p>We use a novel method (cross-lagged hidden Markov models) to identify which combinations of job demands and resources occur among workers, how often, and how these affect mental health and vice versa. Hidden Markov models (HMM) are a longitudinal extension of latent class analysis (LCA), which can be used to measure concepts that are not directly observable. As in LCA, indicator variables are used to measure such concepts. We use twelve indicators of JDR, and five indicators of mental health. HMMs group individuals with similar response patterns on the indicators in categories of the latent variable and analyse how individuals move between these categories. Additionally, predictors can be added to the model to investigate which factors influence transitions between the identified states. We used this model to study the cross-lagged relations between JDR and mental health: how JDR in time point <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(t-1\)</EquationSource> </InlineEquation> affects mental health in time point <i>t</i> and mental health in time point <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(t-1\)</EquationSource> </InlineEquation> affects JDR in time point <i>t</i>. We used yearly data from the Dutch Longitudinal Internet Social Survey (LISS) from 2016 to 2023. Our sample includes respondents who were employees in 2016 and for whom we had data on their JDR and mental health for at least four years. We identified six JDR states, ranging from ‘Tough job’ (high demands and few resources) to ‘Dream job’ (moderate demands and very high resources). We also identified three mental health states: poor, moderate, and good. Among those in moderate health, transitions to good health were more common for respondents in the ‘Dream job’ state and less common for respondents in the ‘Tough job’ state. Our results suggest a healthy worker effect: transitions from states with a disadvantageous combination of JDR to better states were more common among employees in moderate or good mental health. Our study shows how HMMs can improve our knowledge on the empirical predictions of widely studied theories such as the JDR model and its interplay with mental health. This is relevant for scholars and practitioners alike.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The interplay between job demands & resources and mental health: a novel approach using hidden Markov models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月19日

查看全文
  • Stef Bouwhuis,
  • Ceciel Pauls,
  • Mauricio Garnier-Villarreal,
  • Dimitris Pavlopoulos

摘要

We use a novel method (cross-lagged hidden Markov models) to identify which combinations of job demands and resources occur among workers, how often, and how these affect mental health and vice versa. Hidden Markov models (HMM) are a longitudinal extension of latent class analysis (LCA), which can be used to measure concepts that are not directly observable. As in LCA, indicator variables are used to measure such concepts. We use twelve indicators of JDR, and five indicators of mental health. HMMs group individuals with similar response patterns on the indicators in categories of the latent variable and analyse how individuals move between these categories. Additionally, predictors can be added to the model to investigate which factors influence transitions between the identified states. We used this model to study the cross-lagged relations between JDR and mental health: how JDR in time point \(t-1\) affects mental health in time point t and mental health in time point \(t-1\) affects JDR in time point t. We used yearly data from the Dutch Longitudinal Internet Social Survey (LISS) from 2016 to 2023. Our sample includes respondents who were employees in 2016 and for whom we had data on their JDR and mental health for at least four years. We identified six JDR states, ranging from ‘Tough job’ (high demands and few resources) to ‘Dream job’ (moderate demands and very high resources). We also identified three mental health states: poor, moderate, and good. Among those in moderate health, transitions to good health were more common for respondents in the ‘Dream job’ state and less common for respondents in the ‘Tough job’ state. Our results suggest a healthy worker effect: transitions from states with a disadvantageous combination of JDR to better states were more common among employees in moderate or good mental health. Our study shows how HMMs can improve our knowledge on the empirical predictions of widely studied theories such as the JDR model and its interplay with mental health. This is relevant for scholars and practitioners alike.