Least Squares and Grouping Method Estimators in the Errors in Variables Model

Authors:

Abstract
The probability density function of the least squares estimator of the slope coefficient in the errors in variables model is presented. It is shown how the bias and mean-square error of the least squares estimator b depend on the parameters of the model. In particular, for a given sample size, b converges to the true parameter as one of the distribution parameters increases indefinitely. The analysis is supplemented with numerical computations of the relative bias and mean-square error. The distribution function of the grouping method estimator has the same form as that of 6. The biases and mean-square errors of b and are compared. For the case of zero within-group variance, the use of always reduces the magnitude of the relative bias and generally reduces the mean-square error. For large values of the within-group variance, use of may result in an increase in mean-square error.