“…Outcomes within class can mutually depend on continuously distributed latent factor(s) to accommodate associations among responses within class and represent systematic individual variability within class. This extension has appeared under separate names; for LPA it is called a factor mixture model (e.g., Lubke & Muthén, 2005;Yung, 1997); for GBT, a growth mixture model (e.g., Muthén & Shedden, 1999;Verbeke & Lesaffre, 1996); for LCA, a categorical-item factor mixture or IRT mixture model (Lubke & Neale, 2008;Mislevy & Verhelst, 1990;Muthén & Asparouhov, 2006;Rost, 1990); and for LTA, LTA with a categorical-item factor/IRT measurement model (Cho, Cohen, Kim, & Bottge, 2010;Nylund, 2007). Here we highlight the similarity of this extension across multivariate mixtures by reviewing the inclusion of one factor per class/state .q D 1/ in LPA, LCA, GBT, and LTA.…”