人力资源管理领域的纵向研究设计与数据分析

Longitudinal Design and Data Analyses in the Research Field of Human Resource Management

相对于横断面研究,纵向研究可以为建立变量之间的因果关系提供更好的证据。而在不同的纵向研究方法中,面板研究具有其独特的优势,常被用来检验变量之间复杂的动态关系。在本文中,笔者围绕纵向研究,梳理了相关的文献。笔者首先讨论了纵向研究设计中常见的问题与挑战,然后分别介绍了四种常用的面板数据的分析方法(随机系数模型、潜在增长模型、交叉滞后模型、潜变化分数模型),并对各自的优缺点和适用范围进行了分析。

Longitudinal studies can provide stronger evidence for establishing the causal relationships among variables than cross-sectional studies.Among various longitudinal designs,panel design has its unique advantages and is often used to examine the complex dynamic relationships.In this paper,we first discussed the common issues in longitudinal studies,then introduced four popular approaches to modeling change(i.e.,random coefficient modeling,latent growth modeling,cross-lagged panel modeling,latent change score modeling),and finally analyzed their respective advantages,disadvantages and scope of application.