This chapter provides a broad introduction to a person-centered perspective of commitment research more generally, and to the various possibilities provided by mixture models more specifically. Thus, after introducing the person-centered approach to commitment research, key elements of its implementation in research are addressed. A generalized structural equation modeling framework of broad relevance to all forms of person-centered analyses is then introduced, together with user-friendly descriptions of the various types of models that can be estimated using this framework (latent profile analyses, latent profile analyses with covariates, factor mixture analysis, similarity testing for profile solutions, latent transition analyses, mixture regression analyses, growth mixture analyses). For each model, a non-technical description of the model and of its implementation is provided, and followed by a brief illustration of the model through a short description of previously published applications
Alexandre J. S. Morin, Matthew J. W. McLarnon and David Litalien
This chapter introduces mixture modeling, with a specific focus on the analytical possibilities provided by this methodological framework for cross-sectional and longitudinal organizational behavior research. We first introduce basic principles of mixture modeling, which are also broadly labeled person-centered approaches, before presenting, in a pedagogical manner, various types of mixture models that are available to researchers. These models include latent profile analyses, multi-group latent profile analyses, mixture regression analyses, and the longitudinal models of profile similarity, latent transition analyses, and growth mixture analyses. For each model, a non-technical description and recommendations for its implementation are provided, followed by brief illustrations of the model as it has been applied in previous studies. To enable readers to carry out the advanced models demonstrated here in an informed manner, we also provide an extensive set of online supplements that focus on the estimation of these models in the Mplus statistical package.