Open Access
2014 A note on BIC in mixed-effects models
Maud Delattre, Marc Lavielle, Marie-Anne Poursat
Electron. J. Statist. 8(1): 456-475 (2014). DOI: 10.1214/14-EJS890

Abstract

The Bayesian Information Criterion (BIC) is widely used for variable selection in mixed effects models. However, its expression is unclear in typical situations of mixed effects models, where simple definition of the sample size is not meaningful. We derive an appropriate BIC expression that is consistent with the random effect structure of the mixed effects model. We illustrate the behavior of the proposed criterion through a simulation experiment and a case study and we recommend its use as an alternative to various existing BIC versions that are implemented in available software.

Citation

Download Citation

Maud Delattre. Marc Lavielle. Marie-Anne Poursat. "A note on BIC in mixed-effects models." Electron. J. Statist. 8 (1) 456 - 475, 2014. https://doi.org/10.1214/14-EJS890

Information

Published: 2014
First available in Project Euclid: 2 May 2014

zbMATH: 1348.62186
MathSciNet: MR3200764
Digital Object Identifier: 10.1214/14-EJS890

Subjects:
Primary: 62J02 , 62J12

Keywords: Bayesian Information Criterion , BIC , Mixed effects model , Variable selection

Rights: Copyright © 2014 The Institute of Mathematical Statistics and the Bernoulli Society

Vol.8 • No. 1 • 2014
Back to Top