Reduced bias estimation of variance for repeated measurements model

Abstract

This paper discusses two important points; the first part describes the one-way repeated measurements model, which had two random effects in addition to the random error and two fixed factors. The model was also written as a set of matrices, which showed how the random effects and fixed factors interacted. The second part shows the maximum likelihood method for determining the variance components for a one-way repeated measures model. In this method, it was seen that the estimates of the variance combinations have some bias. The goal of this study is to use the Mean bias reduction method of the one-way repeated measurements model to reduce the bias in the estimators of the variance components.