Generalized S-Estimators for Linear Mixed-Effects Models

• Activity Number: 18
• Type: Topic Contributed
• Date/Time: Sunday, August 4, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Inna Chervoneva*+ and Mark Vishnyakov
• Companies: Thomas Jefferson University and Thomas Jefferson University
• Keywords: Robust estimation ; Clustered data ; qRT-PCR efficiency ; Redescending M-estimator ; Breakdown point

Abstract

High breakdown redescending estimators are currently the robust methods of choice for multivariate linear regression, but such estimators so far have been developed only for the limited class of completely balanced linear mixed effects (LME) models. For a general unbalanced hierarchical LME model, we propose a class of redescending M-estimators that include S-estimators in a fully balanced model as a particular case. Therefore, they are referred to as generalized S-estimators. The asymptotic properties of the proposed estimators are established. A small simulation study is conducted to compare performance of the generalized S-estimates and M-estimates for unbalanced LME models. Finally, the proposed generalized S-estimators are used for robust analysis of age-related changes in hemoglobin levels of sickle cell disease patients. The proposed generalized S-estimators are used to estimate the PCR efficiency in serial dilution experiments in the context of qRT-PCR studies.