Empirical Bayes Prediction in Random Effects Models with Covariates Subject to Measurement Error

Abstract

Abtsrcat

The paper considers a nested error linear regression model with multiple covariates subject to measurement error. An efficient empirical Bayes (EB) predictor is obtained using the information on both the observed response variables and the observed covariates. We derive a correct second- order unbiased approximation to the mean squared error (MSE) of the EB predictor.