Review of the REBLUP method for estimating variance components under the nested error model

Abstract

E-mail: sebrek@uni-corvinus.hu The present work aims to analyse the REBLUP (robust empirical best linear unbiased prediction) method as proposed by Sinha–Rao [2009] for computing robust estimators of variance components under the nested error unit-level model. It explains the theoretical and computational aspects associated with the REBLUP method to reveal the strengths and weaknesses of the proposed approach. A Monte Carlo study is then conducted to analyse the method’s performance under different scenarios.