Traditional mediation analysis assumes that a study population is homogeneous and the mediation effect is constant over time, which may not hold in some applications. Motivated by smoking cessation data, we propose a latent class dynamic mediation model that explicitly accounts for the fact that the study population may consist of different subgroups and the mediation effect may vary over time. We use a proportional odds model to accommodate the subject heterogeneities and identify latent subgroups. Conditional on the subgroups, we employ a Bayesian hierarchical nonparametric time-varying coefficient model to capture the time-varying mediation process, while allowing each subgroup to have its individual dynamic mediation process. A simulation study shows that the proposed method has good performance in estimating the mediation effect. We illustrate the proposed methodology by applying it to analyze smoking cessation data.
Relative to previous research concerning the positive association between high-performance work systems (HPWS) and employees’ voice and helping, we examined a wider range of mediators reflecting employees’ ability, motivation, and opportunity to expand their citizenship-based role definitions. Trust in the supervisor was also investigated as a boundary condition on the relationships in question. Multisource data, collected in 4 waves, from 208 supervisor–employee dyads showed that employees’ efficacy, instrumentality, and autonomy perceptions concerning voice mediated the association between employee-experienced HPWS and expanded role definition for voice. Instrumentality mediated the relationship between employee-experienced HPWS and expanded role definition for helping. The positive links between employee-experienced HPWS and both supervisor-rated helping and voice were mediated by employees’ role definitions. Trust in the supervisor positively moderated the mediated effects. 相似文献