Sankhya: The Indian Journal of Statistics
1998, Volume 60, Series B, Pt. 3, 433--436
A NEW COVARIANCE ESTIMATOR IN RANDOM COEFFICIENT REGRESSION MODEL
NESAN CHELLIAH Griffith University, Brisbane
SUMMARY. A new method is given to obtain an estimator for the covariance matrix of the response variable in a linear regression model. This method gives positive variance components and offers a mechanism for improving the performance of the weighted least squares estimator of the regression parameters.
AMS (1991) subject classification.62F10.
Key words and phrases. Residual type statistics, random coeficient.
Full paper (PDF)
This article in Mathematical Reviews