Introduction to the Best Practice Recommendations for Longitudinal Latent Transition Analysis

Abstract

The use of finite mixture models to identify a limited number of mutually exclusive latent (i.e., unknown) subgroups (i.e., classes or profiles) based on individuals' survey responses, observational and/or physiological values. Finite mixture modelling is also an active area of quantitative methodological research and advancement, which means best practices continue to evolve. This tutorial on latent transition analysis (LTA) is written to facilitate researchers using this modelling approach to answer research questions about transitions from membership in one latent class or profile construct to another latent class or profile construct rigorously and following best practices. This tutorial will cover the purpose, analysis steps, interpretation and recommended reporting practices for LTA. To increase the applicability and approachability of this LTA tutorial article example, three timepoint sport psychology data on the health behaviours of collegiate student-athletes, the Mplus syntax for the analysis and the decision process along with the results, tables and figures are included in the supplemental online materials.