Bayesian estimation of random effects models for multivariate responses of mixed data

• Published in: Computational statistics & data analysis Vol. 54; no. 5; pp. 1206 - 1218
• Main Authors: Wagner, Helga, Tüchler, Regina
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.05.2010; Elsevier
• Series: Computational Statistics & Data Analysis
• Subjects: Auxiliary mixture sampling; Auxiliary mixture sampling Generalized linear models MCMC Random effects model Variable selection; Exact sciences and technology; General topics; Generalized linear models; Mathematics; MCMC; Multivariate analysis; Numerical analysis; Numerical analysis. Scientific computation; Numerical methods in probability and statistics; Parametric inference; Probability and statistics; Random effects model; Sciences and techniques of general use; Statistics; Variable selection; MCMC; Auxiliary mixture sampling; Generalized linear models; Random effects model; Variable selection; Bayes estimation; Mixed distribution; Monte Carlo method; Data analysis; Generalized linear model; Binary data; Response model; Multivariate analysis; Gibbs sampling; Implementation; Markov chain; Statistical computation; Model specification; Random effect; Cohort data; Selection method; Multinomial distribution
• ISSN: 0167-9473; 1872-7352
• DOI: 10.1016/j.csda.2009.12.007

A random effects model is presented to estimate multivariate data of mixed data types. Such data typically appear in studies where different response variables are measured repeatedly for one subject. It is possible to relate normal, binary, multinomial and count data by our joint model. Further flexibility with respect to model specification is obtained by including modern variable selection techniques. Auxiliary mixture sampling leads to a Gibbs sampling type scheme which is easy to implement since no additional tuning is needed. The method is illustrated by transaction data of a costumer cohort acquired by an apparel retailer.