Abstract
This paper is concerned with the double k-class estimation methods for econometric model in which we especially consider the k-class estimation method and the h-class estimation method. The properties of both estimation (i.e., mean square error, bias and variance) for k-variables are studied by a Monte Carlo simulation. In consequence of the experiment we obtain the optimum ranges of k-variables for both estimation methods. Generally the value of k which gives minimum mean square error for the k-class estimator is shown by 0≦k≦1. On the other hand for the h-class estimator, 0≦k≦kU (where, 1≦k_U).
