Overview
Jeroen K. Vermunt is a Dutch methodologist and statistician at Tilburg University, known for advancing latent class and finite-mixture modeling for the social sciences and for co-developing the Latent GOLD software. His work spans LC cluster, factor, and regression models, latent Markov (transition) models, and the "three-step" methods for relating class membership to external variables.
Key Contributions / Features
- Co-author of the unified LC cluster/factor/regression treatment in Magidson-Vermunt (2004) Latent Class Models.
- Co-developer (with Jay Magidson) of Latent GOLD, a widely used program for LC and finite-mixture modeling.
- Contributions to latent Markov models, bias-adjusted three-step LC modeling, and LC models with covariates and mixed-scale indicators.
- Multilevel LC models (Vermunt 2008): extended latent class / finite mixture models to hierarchical data via random effects (continuous or discrete) across higher-level units, yielding mixtures at two levels, with an adapted EM algorithm.
Related