ACCURACY STIMATION OF A MULTIVARIABLE REPEATED MEASURES DESIGN

Abstract

In order to control their possible effect and to obtain greater reliability on the estimation of treatments, the repeated measures design could be extended to more complex structures when introduces a classification factgor of the subjects. Generizability theory assumes that other sources of variance in addition of the individual differences, permits to integrate each one of tghese sources of variance in a general structure that admit specific application of the sample statistical theory.