On The Problem of the Estimation of Variance Components Based On Non-linear Maximization Approach

Ahmad Khaldi

Mutah University, Faculty of Science, Jordan

khaldiahmad1221@gmail.com

Abstract

In this paper, we study the problem of estimating variance components in the two-way classification with interaction in the random effect linear model by non-linear maximization. We assume the model according to the assumptions and give the theory of derivation of the estimators of these components, then apply these estimators on real data and obtain the estimates. We estimate these components by two other methods: the solution of the expected equation of mean square in the analysis of the variance table, and the minimum variance quadratic unbiased estimator.

Keywords: Variance; Linear model; Non-linear maximization; Estimation theory