Latent variable mixture modeling represents a flexible approach to investigating population heterogeneity by sorting
cases into latent but non-arbitrary subgroups that are more homogeneous. The purpose of this selective review is to provide
a non-technical introduction to mixture modeling in a cross-sectional context. Latent class analysis is used to classify
individuals into homogeneous subgroups (latent classes). Factor mixture modeling represents a newer approach that
represents a fusion of latent class analysis and factor analysis. Factor mixture models are adaptable to representing categorical
and dimensional states of affairs. This article provides an overview of latent variable mixture models and illustrates
the application of these methods by applying them to the study of the latent structure of psychotic experiences. The
flexibility of latent variable mixture models makes them adaptable to the study of heterogeneity in complex psychiatric
and psychological phenomena. They also allow researchers to address research questions that directly compare the viability
of dimensional, categorical and hybrid conceptions of constructs.