ABSTRACT
In this chapter, we present an overview of latent mixture modeling. Mixture models are a special type of quantitative model in which latent variables can be used to represent mixtures of subpopulations or classes where population membership is not known but, rather, is inferred from the data. Mixture modeling is used to assign individuals to their most likely latent class and to obtain parameter estimates that explain differences between the classes identified. Mixtures can be identified among individuals within organizational units as well as among organizational units. The approach can be applied to both cross-sectional and longitudinal models and can enrich our understanding of heterogeneity among both individuals and groups. In this chapter, we provide several examples where mixture models might be applied to typical multilevel models. We then illustrate how choices of analytic method can impact the optimal investigation of the data. This overview provides some basics for further development of these issues and models.
