multilevLCA
multilevLCA is an R package for single-level and multilevel latent class analysis with covariates. It was developed by Roberto Di Mari and Johan Lyrvall, with contributions from Zsuzsa Bakk, Jennifer Oser and Jouni Kuha.
The package efficiently implements based on C++ integrated routines the two-step, two-stage, and one-step estimation approaches for latent class models with covariates. It performs semi-automatic sequential and simultaneous model selection, and visualizes the output in text and figures in a user-friendly way.
Documentation
Bias-adjusted three-step modeling
Bias-adjusted three-step analysis of single-level and multilevel latent class models can be implemented in multilevLCA by means of the functions in this downloadable .R file.
These functions replace the code used in Bias-Adjusted Three-Step Multilevel Latent Class Modeling with Covariates (Lyrvall et al., 2024), which is outdated after package updates.