LINDLEY-LIKE MEAN CORRECTION IN THE IMPROVED ESTIMATION OF REGRESSION-MODELS WITH NONSCALAR COVARIANCE-MATRIX

In the linear regression model with disturbances covariance matrix depending upon a few unknown parameters, the risks of the feasible generalized least squares and improved estimators with and without Lindley-like mean correction are derived and dominance conditions of one estimator over the other are obtained when sample size is large. For a specific choice of the loss function it has been observed that the improved estimator without Lindley-like mean correction dominates the FGLS estimator whenever the number of explanatory variables is greater than three, whereas the estimator with Lindley-like mean correction dominates the FGLS estimator whenever the number of explanatory variables is greater than six. © 1990.