Latent Class and Finite Mixture Models

Abstract

Abstract Latent class analysis and, more generally, finite mixture models have seen increased use in marketing since the early 1980s. The popularity of latent class and finite mixture models can, in large measure, be traced to the important role individual differences (i.e., heterogeneity) play in understanding marketing and consumer behavior phenomenon. At the most basic level, latent class and finite mixture models have allowed marketing researchers to account for individual heterogeneity when estimating model parameters. Our objectives in this article are twofold: First, we provide an overview of latent class and finite mixture models, which considers basic model forms, estimation, and special cases. Second, we provide a brief discussion of the various model forms and application settings that have relied on latent class and finite mixture models to account for individual‐level heterogeneity and, in so doing, provide both a historical perspective and a treatment of some recent developments.