ABSTRACT
Growth mixture modeling (GMM) is informative when examining the associations between heterogeneity in a time-varying attribute (e.g., health) and its predictors and outcomes. This chapter illustrates the incorporation of predictors and outcomes (covariates) into GMMs. Like other modeling approaches, growth mixture models (including latent class growth analyses, or LCGA) can incorporate multiple covariates (predictors and distal outcomes). Two approaches can be used in Mplus to add covariates into a growth mixture model: (1) the direct specification approach (i.e., one-step approach) and (2) the three-step approach. This chapter provides Mplus syntax for the incorporation of predictors and outcomes into a GMM using both approaches and guidance for interpreting the results.
