A mean squared error criterion is used to compare five estimators of the coefficients in a linear regression model: least squares, principal components, ridge ...

The objectives of this paper are to examine the mean square error criterion for rejecting or adopting restrictions on the parameter space in a regression model, and ...