This workshop covers introductory and some intermediate aspects of the increasingly popular modeling approach of latent class analysis (LCA; also referred to as finite mixture analysis). LCA has developed over the past nearly 60 years into a mainstream analytic statistical procedure for classification and clustering, which is applicable with both categorical and continuous observed measures (finite mixture modeling). LCA is currently widely applied in empirical studies across the social, behavioral, political, marketing, business, and biomedical disciplines when the concern is to find out whether there may be some evidence consistent with the existence of certain latent classes of units of analysis, which are not observed and correspond to one or more categorical latent variables. The workshop includes a thorough discussion of the fundamentals of LCA as well as the topics of model identification and model selection, and instrumentally utilizes the framework of the latent variable modeling methodology and the software Mplus. Throughout the workshop a number of empirical examples are used, with detailed discussions of pertinent Mplus code and output.
The target audience for this workshop are Faculty and Graduate Students with no or limited exposure to multivariate statistics or latent class analysis.