Skip to Main content Skip to Navigation
Journal articles

Robustness of the linear mixed model to misspecied error distribution: Robustness of the linear mixed model

Abstract : A simulation study is performed to investigate the robustness of the maximum likelihood estimator of fixed effects from a linear mixed model when the error distribution is misspecied. Inference for the fixed effects under the assumption of independent normally distributed errors with constant variance is shown to be robust when the errors are either non-gaussian or heteroscedastic, except when the error variance depends on a covariate included in the model with interaction with time. Inference is impaired when the errors are correlated. In the latter case, the model including a random slope in addition to the random intercept is more robust than the random intercept model. The use of Cholesky residuals and conditional residuals to evaluate the fit of a linear mixed model is also discussed.
Document type :
Journal articles
Complete list of metadatas

https://www.hal.inserm.fr/inserm-00084214
Contributor : Helene Jacqmin-Gadda <>
Submitted on : Thursday, December 14, 2006 - 9:17:13 AM
Last modification on : Wednesday, August 19, 2020 - 11:17:03 AM
Long-term archiving on: : Monday, April 5, 2010 - 11:51:52 PM

Identifiers

Collections

Citation

Hélène Jacqmin-Gadda, Solenne Sibillot, Cécile Proust, Jean-Michel Molina, Rodolphe Thiébaut. Robustness of the linear mixed model to misspecied error distribution: Robustness of the linear mixed model. Comput Stat Data Anal, 2007, 51 (10), pp.5142-5154. ⟨10.1016/j.csda.2006.05.021⟩. ⟨inserm-00084214⟩

Share

Metrics

Record views

816

Files downloads

4005