FUNCTIONALITY OF FOUR MODELS FOR GENOTYPIC EVALUATIONS IN SERIES OF EXPERIMENTS

Abstract

The quality of the analysis of the data produced in combined experiments conducted to evaluate genotypes depends primarily on the degree of correspondence between the real experimental situation and the model used to represent it. In this study, the expected mean squares of each of four models, each including three cases, used to represent situations in which "g" genotypes are evaluated in "" locations dhring a years are presented: thus plant breeders can identify for each case the procedure to test hypotheses. Additionally, the nature and relationships among factors are discussed in the context of the typical experimental situations relative to genotypic evaluation of annual crops.