Inference for mixed models of ANOVA type with high-dimensional data

Inference for variance components in linear mixed models of ANOVA type, including estimation and testing, has been investigated when the number of fixed effects is fixed. However, for high-dimensional data, this number is large and would be regarded as a divergent value as the sample size goes to infinity. In this paper, existing tests are extended to handle this problem with a sparse model structure. To avoid the impact from insignificant fixed effects, the proposed tests are post-selection-based with an orthogonality-based selection of SCAD type applied to selecting significant fixed effects into working model. The selection and estimation of fixed effects are under the assumption on the existence of second order moments for errors. Two types of tests for random effects are considered and some new insights are obtained. The proposed tests are distribution-free, though they request the existence of the fourth moments of random effects and errors. The proposed methods are illustrated by simulation studies and a real data analysis.

[1]  C. Crainiceanu,et al.  Restricted Likelihood Ratio Testing for Zero Variance Components in Linear Mixed Models , 2008 .

[2]  Jianqing Fan,et al.  Nonconcave penalized likelihood with a diverging number of parameters , 2004, math/0406466.

[3]  A Reparametrization Approach for Dynamic Space-Time Models , 2008, Journal of statistical theory and practice.

[4]  D. Stram,et al.  Variance components testing in the longitudinal mixed effects model. , 1994, Biometrics.

[5]  Sanjoy K. Sinha,et al.  Bootstrap tests for variance components in generalized linear mixed models , 2009 .

[6]  H. Bondell,et al.  Joint Variable Selection for Fixed and Random Effects in Linear Mixed‐Effects Models , 2010, Biometrics.

[7]  Reiko Aoki,et al.  Assessment of variance components in nonlinear mixed-effects elliptical models , 2012 .

[8]  On Variance Components in Semiparametric Mixed Models for Longitudinal Data , 2010 .

[9]  U-tests for variance components in one-way random effects models , 2008, 0805.2316.

[10]  Helmut Küchenhoff,et al.  Size and power of tests for a zero random effect variance or polynomial regression in additive and linear mixed models , 2008, Comput. Stat. Data Anal..

[11]  P. Sen,et al.  U-tests for variance components in linear mixed models , 2013 .

[12]  Jianqing Fan,et al.  Variable Selection via Nonconcave Penalized Likelihood and its Oracle Properties , 2001 .

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

[14]  Geert Verbeke,et al.  Testing multiple variance components in linear mixed-effects models. , 2013, Biostatistics.

[15]  Zaixing Li,et al.  Variance Components Testing in ANOVA‐Type Mixed Models , 2014 .

[16]  D. Ruppert,et al.  Likelihood ratio tests in linear mixed models with one variance component , 2003 .

[17]  W. Fung,et al.  Variance component testing in semiparametric mixed models , 2004 .

[18]  Pranab Kumar Sen,et al.  Decomposability of high-dimensional diversity measures: Quasi-U-statistics, martingales and nonstandard asymptotics , 2009, J. Multivar. Anal..