Description
Latent Class/Cluster Analysis and Mixture Modeling is a five-day workshop focused on the application and interpretation of statistical techniques designed to identify subgroups within a heterogeneous population. Broadly, these techniques can be divided into: (a) cluster analysis procedures that group participants via algorithms or decision rules, and (b) latent class analysis, latent profile analysis, and other finite mixture models that discern latent subgroups of individuals using a formal statistical model. In practice, these methods are often implemented with the goal of identifying theoretically distinct subgroups (e.g., people with a liability for schizophrenia versus those without). Alternatively, they can be used as a data reduction device, to summarize prototypical patterns when working with complex multivariate data (e.g., market segmentation in consumer research). In recent years, an increasing focus has been on multivariate and longitudinal applications (e.g., growth mixture modeling). In this workshop we provide a comprehensive exploration of the foundations and uses of cluster analysis, latent class/profile analysis and finite mixture models, with topics ranging from introductory to advanced, and applications to both single-time point and longitudinal data, as detailed below.