Bayesian Covariance Selection in Generalized Linear Mixed Models

Summary The generalized linear mixed model (GLMM), which extends the generalized linear model (GLM) to incorporate random effects characterizing heterogeneity among subjects, is widely used in analyzing correlated and longitudinal data. Although there is often interest in identifying the subset of predictors that have random effects, random effects selection can be challenging, particularly when outcome distributions are nonnormal. This article proposes a fully Bayesian approach to the problem of simultaneous selection of fixed and random effects in GLMMs. Integrating out the random effects induces a covariance structure on the multivariate outcome data, and an important problem that we also consider is that of covariance selection. Our approach relies on variable selection‐type mixture priors for the components in a special Cholesky decomposition of the random effects covariance. A stochastic search MCMC algorithm is developed, which relies on Gibbs sampling, with Taylor series expansions used to approximate intractable integrals. Simulated data examples are presented for different exponential family distributions, and the approach is applied to discrete survival data from a time‐to‐pregnancy study.

[1]  G. Barrie Wetherill,et al.  Random Effects Models , 1981 .

[2]  J. Ware,et al.  Random-effects models for longitudinal data. , 1982, Biometrics.

[3]  P. McCullagh,et al.  Generalized Linear Models , 1992 .

[4]  Scott L. Zeger,et al.  Generalized linear models with random e ects: a Gibbs sampling approach , 1991 .

[5]  A. Raftery,et al.  How Many Iterations in the Gibbs Sampler , 1991 .

[6]  John Geweke,et al.  Evaluating the accuracy of sampling-based approaches to the calculation of posterior moments , 1991 .

[7]  R. Schall.  Estimation in generalized linear models with random effects , 1991 .

[8]  D. Rubin,et al.  Inference from Iterative Simulation Using Multiple Sequences , 1992 .

[9]  C. Weinberg,et al.  Reduced fertility among women employed as dental assistants exposed to high levels of nitrous oxide. , 1992, The New England journal of medicine.

[10]  Reduced fertility among women employed as dental assistants exposed to high levels of nitrous oxide. , 1992 .

[11]  David R. Cox,et al.  Nonlinear component of variance models , 1992 .

[12]  N. Breslow,et al.  Approximate inference in generalized linear mixed models , 1993 .

[13]  J. Q. Smith,et al.  1. Bayesian Statistics 4 , 1993 .

[14]  E. George,et al.  Journal of the American Statistical Association is currently published by American Statistical Association. , 2007 .

[15]  D D Baird,et al.  Pitfalls inherent in retrospective time-to-event studies: the example of time to pregnancy. , 1993, Statistics in medicine.

[16]  C. Mcgilchrist.  Estimation in Generalized Mixed Models , 1994 .

[17]  Stephen W. Raudenbush,et al.  Random effects models. , 1994 .

[18]  J. Geweke,et al.  Variable selection and model comparison in regression , 1994 .

[19]  Adrian E. Raftery,et al.  Accounting for Model Uncertainty in Survival Analysis Improves Predictive Performance , 1995 .

[20]  W. Gilks,et al.  Adaptive Rejection Metropolis Sampling Within Gibbs Sampling , 1995 .

[21]  A. Raftery.  Approximate Bayes factors and accounting for model uncertainty in generalised linear models , 1996 .

[22]  Xihong Lin.  Variance component testing in generalised linear models with random effects , 1997 .

[23]  Daniel Commenges,et al.  Generalized Score Test of Homogeneity Based on Correlated Random Effects Models , 1997 .

[24]  S. Chib,et al.  Bayesian Tests and Model Diagnostics in Conditionally Independent Hierarchical Models , 1997 .

[25]  Walter R. Gilks,et al.  Corrigendum: Adaptive Rejection Metropolis Sampling , 1997 .

[26]  C. McCulloch.  Maximum Likelihood Algorithms for Generalized Linear Mixed Models , 1997 .

[27]  Adrian F. M. Smith,et al.  Bayesian Statistics 5. , 1998 .

[28]  R. Kass,et al.  Nonconjugate Bayesian Estimation of Covariance Matrices and its Use in Hierarchical Models , 1999 .

[29]  P. Damlen,et al.  Gibbs sampling for Bayesian non‐conjugate and hierarchical models by using auxiliary variables , 1999 .

[30]  Jim X. Chen,et al.  Approximate Line Scan‐Conversion and Antialiasing , 1999, Comput. Graph. Forum.

[31]  R. Kass,et al.  Bayes Factors and Approximations for Variance Component Models , 1999 .

[32]  Ming-Hui Chen,et al.  Monte Carlo Estimation of Bayesian Credible and HPD Intervals , 1999 .

[33]  Dongchu Sun,et al.  PROPRIETY OF POSTERIORS WITH IMPROPER PRIORS IN HIERARCHICAL LINEAR MIXED MODELS , 2001 .

[34]  Daniel B. Hall,et al.  Order‐restricted score tests for homogeneity in generalised linear and nonlinear mixed models , 2001 .

[35]  S. Sinharay.  Bayes factors for variance component testing in generalized linear mixed models , 2001 .

[36]  Eric R. Ziegel,et al.  Generalized Linear Models , 2002, Technometrics.

[37]  A-5 Default Bayes Factors for Variance Component Models , 2002 .

[38]  Purushottam W. Laud,et al.  Predictive Variable Selection in Generalized Linear Models , 2002 .

[39]  Guang Guo,et al.  The Mixed or Multilevel Model for Behavior Genetic Analysis , 2002, Behavior genetics.

[40]  H. Chipman,et al.  Bayesian Treed Generalized Linear Models , 2003 .

[41]  P. Dellaportas,et al.  Bayesian variable and link determination for generalised linear models , 2003 .

[42]  Joseph G. Ibrahim,et al.  Prior elicitation for model selection and estimation in generalized linear mixed models , 2003 .

[43]  Michael J Daniels,et al.  Modelling the random effects covariance matrix in longitudinal data , 2003, Statistics in medicine.

[44]  Brian Neelon,et al.  Bayesian Inference on Order‐Constrained Parameters in Generalized Linear Models , 2003, Biometrics.

[45]  Mark Von Tress,et al.  Generalized, Linear, and Mixed Models , 2003, Technometrics.

[46]  D. Dunson,et al.  Random Effects Selection in Linear Mixed Models , 2003, Biometrics.

[47]  R. Kohn,et al.  Efficient estimation of covariance selection models , 2003 .

[48]  Merrill W. Liechty,et al.  Bayesian correlation estimation , 2004 .

[49]  A. Gelman.  Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper) , 2004 .

[50]  Edward I. George,et al.  Bayesian Treed Models , 2002, Machine Learning.

[51]  David J Nott,et al.  Sampling Schemes for Bayesian Variable Selection in Generalized Linear Models , 2004 .

[52]  P. Gustafson,et al.  Conservative prior distributions for variance parameters in hierarchical models , 2006 .

[53]  S. Fienberg.  When did Bayesian inference become "Bayesian"? , 2006 .

[54]  Satkartar K. Kinney,et al.  Fixed and Random Effects Selection in Linear and Logistic Models , 2007, Biometrics.