Comparer des méthodes

	Bayesian Latent Class Analysis	Analyse de Profils Latents (LPA)
Domaine	Statistique	Psychométrie
Famille	Latent structure	Latent structure
Année d'origine	1990s–2000s	2010
Auteur d'origine	Lazarsfeld (classical LCA); Bayesian formulation developed through Cheeseman & Stutz (1996) and Dunson & Xing (2009)	Lazarsfeld & Henry; Collins & Lanza
Type	Bayesian latent variable / finite mixture model	Person-centered finite mixture model
Source fondatrice	Dunson, D. B. & Xing, C. (2009). Nonparametric Bayes modeling of multivariate categorical data. Journal of the American Statistical Association, 104(487), 1042–1051.	Collins, L. M., & Lanza, S. T. (2010). Latent Class and Latent Transition Analysis. Wiley. ISBN: 978-0-470-22839-7
Alias	Bayesian LCA, BLCA, Bayesian mixture of multinomials, Bayesian finite mixture model	Continuous Latent Class Analysis, Gaussian Profile Mixture Model, Person-Centered Cluster Analysis, Gizil Profil Analizi
Apparentées	6	2
Résumé	Bayesian latent class analysis extends classical LCA by placing prior distributions on all model parameters and using posterior inference — typically via MCMC — to classify individuals into unobserved categorical groups, quantify uncertainty around class membership, and select the number of classes in a principled, probabilistic way.	Latent Profile Analysis (LPA) is a person-centered finite mixture modeling technique that identifies unobserved subgroups — called profiles — within a population based on patterns of scores across multiple continuous indicators. Rooted in Lazarsfeld and Henry's latent structure tradition and formally synthesized for applied behavioral research by Collins and Lanza (2010), LPA assumes that observed heterogeneity in continuous data arises from a discrete number of latent classes, each characterized by a unique multivariate mean profile.