A Monte Carlo study of REML and robust rank-based analyses for the random intercept mixed model

ABSTRACT Restricted maximum likelihood (REML) methods are traditionally used for analyzing mixed models. Based on a multivariate normal likelihood, these analyses are sensitive to outliers. Recently developed robust rank-based procedures offer a complete analysis of mixed model: estimation of fixed effects, standard errors, and estimation of variance components. The results of a large Monte Carlo study are presented, comparing these two analyses for many situations over multivariate normal and contaminated normal distributions. The rank-based analyses are much more powerful and efficient than the REML analyses over all non-normal situations, while losing little power for normal errors.

[1]  Risto Lethonen.  Multilevel Statistical Models (3rd ed.) , 2005 .

[2]  Joseph W. McKean,et al.  Nonparametric Statistical Methods Using R , 2014 .

[3]  J. Jurecková,et al.  Nonparametric Estimate of Regression Coefficients , 1971 .

[4]  J. Hess,et al.  Analysis of variance , 2018, Transfusion.

[5]  Steven P. Reise,et al.  Multilevel Modeling and its Application in Counseling Psychology Research , 1999 .

[6]  D. Bates,et al.  Mixed-Effects Models in S and S-PLUS , 2001 .

[7]  Louis A. Jaeckel.  Estimating Regression Coefficients by Minimizing the Dispersion of the Residuals , 1972 .

[8]  Joseph W. McKean,et al.  R Estimates and Associated Inferences for Mixed Models With Covariates in a Multicenter Clinical Trial , 2012 .

[9]  H. D. Patterson,et al.  Recovery of inter-block information when block sizes are unequal , 1971 .

[10]  G. A. Marcoulides.  Multilevel Analysis Techniques and Applications , 2002 .

[11]  Hira L. Koul,et al.  An Estimator of the Scale Parameter for the Rank Analysis of Linear Models under General Score Functions , 1987 .

[12]  Simon J. Sheather,et al.  High-Breakdown Rank Regression , 1999 .

[13]  Joseph W. McKean,et al.  Rfit: Rank-based Estimation for Linear Models , 2012, R J..

[14]  Joseph W. McKean,et al.  A Robust Two-Stage Multiple Comparison Procedure with Application to a Random Drug Screen , 1989 .

[15]  Joseph W. McKean,et al.  Rank-Based Estimation and Associated Inferences for Linear Models With Cluster Correlated Errors , 2009 .