Bayesian modeling of the dependence in longitudinal data via partial autocorrelations and marginal variances

Many parameters and positive-definiteness are two major obstacles in estimating and modelling a correlation matrix for longitudinal data. In addition, when longitudinal data is incomplete, incorrectly modelling the correlation matrix often results in bias in estimating mean regression parameters. In this paper, we introduce a flexible and parsimonious class of regression models for a covariance matrix parameterized using marginal variances and partial autocorrelations. The partial autocorrelations can freely vary in the interval (-1, 1) while maintaining positive definiteness of the correlation matrix so the regression parameters in these models will have no constraints. We propose a class of priors for the regression coefficients and examine the importance of correctly modeling the correlation structure on estimation of longitudinal (mean) trajectories and the performance of the DIC in choosing the correct correlation model via simulations. The regression approach is illustrated on data from a longitudinal clinical trial.

[1]  Alan Agresti,et al.  Bayesian inference for categorical data analysis , 2005, Stat. Methods Appl..

[2]  Mohsen Pourahmadi,et al.  Modeling covariance matrices via partial autocorrelations , 2009, J. Multivar. Anal..

[3]  M. Pourahmadi.  Joint mean-covariance models with applications to longitudinal data: Unconstrained parameterisation , 1999 .

[4]  Anne Lohrli.  Chapman and Hall , 1985 .

[5]  Dean Phillips Foster,et al.  Calibration and Empirical Bayes Variable Selection , 1997 .

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

[7]  A. Verbyla,et al.  Modelling Variance Heterogeneity: Residual Maximum Likelihood and Diagnostics , 1993 .

[8]  S. Chib,et al.  Bayesian analysis of binary and polychotomous response data , 1993 .

[9]  G. Gastaldi.  Bayesian analysis of covariance matrices and dynamic models for longitudinal data , 2002 .

[10]  M. Daniels,et al.  A Note on MAR, Identifying Restrictions, Model Comparison, and Sensitivity Analysis in Pattern Mixture Models with and without Covariates for Incomplete Data , 2011, Biometrics.

[11]  M. Pourahmadi.  Maximum likelihood estimation of generalised linear models for multivariate normal covariance matrix , 2000 .

[12]  Bill Ravens,et al.  An Introduction to Copulas , 2000, Technometrics.

[13]  Xiao-Li Meng,et al.  Modeling covariance matrices in terms of standard deviations and correlations, with application to shrinkage , 2000 .

[14]  Crump Sl,et al.  The present status of variance component analysis. , 1951 .

[15]  R. Kohn,et al.  Parsimonious Covariance Matrix Estimation for Longitudinal Data , 2002 .

[16]  M J Daniels,et al.  Dynamic conditionally linear mixed models for longitudinal data. , 2002, Biometrics.

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

[18]  Peter D. Hoff,et al.  A Covariance Regression Model , 2011, 1102.5721.

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

[20]  M. Segal,et al.  A parametric family of correlation structures for the analysis of longitudinal data. , 1992, Biometrics.

[21]  H. Joe.  Generating random correlation matrices based on partial correlations , 2006 .

[22]  C. Czado.  Multivariate regression analysis of panel data with binary outcomes applied to unemployment data , 2000 .

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

[24]  A. Rotnitzky,et al.  Missing Data in Longitudinal Studies: Strategies for Bayesian Modeling and Sensitivity Analysis by DANIELS, M. J. and HOGAN, J. W , 2009 .

[25]  Bayesian modeling of several covariance matrices and some results on propriety of the posterior for linear regression with correlated and/or heterogeneous errors , 2006 .

[26]  Dean P. Foster,et al.  Calibration and empirical Bayes variable selection , 2000 .

[27]  M. Clyde,et al.  Flexible empirical Bayes estimation for wavelets , 2000 .

[28]  J. Berger,et al.  Estimation of a Covariance Matrix Using the Reference Prior , 1994 .

[29]  Nicole A. Lazar,et al.  Statistical Analysis With Missing Data , 2003, Technometrics.

[30]  Tom Leonard,et al.  The Matrix-Logarithmic Covariance Model , 1996 .

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

[32]  C. Durand,et al.  A controlled dose‐ranging study of remoxipride and haloperidol in schizophrenia ‐ a Canadian multicentre trial , 1990, Acta psychiatrica Scandinavica. Supplementum.

[33]  B. Efron,et al.  Multivariate Empirical Bayes and Estimation of Covariance Matrices , 1976 .

[34]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[35]  J. Magnus,et al.  Matrix Differential Calculus with Applications in Statistics and Econometrics , 1991 .

[36]  Matt Simpson,et al.  Bayesian inference for a covariance matrix , 2014, 1408.4050.

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

[38]  R. Kass,et al.  Shrinkage Estimators for Covariance Matrices , 2001, Biometrics.