Métodos estatísticos para análise robusta de dados oriundos de experimento agrícola em áreas irrigad

Abstract

PALAVRAS-CHAVE: Homogêneas, análise robusta, abacate comercial STATISTICAL METHODS FOR ROBUST ANALYSIS OF DATA ORIGINATED FROM AGRICULTURAL EXPERIMENT IN IRRIGATED AREASABSTRACT: Usually, the conclusions originated from field or green house experiments are obtained from Variance Analysis (ANOVA). However, the use of this technique demands two fundamental presuppositions: that the errors must be normally distributed and that the variance of the treatments must be homogeneous. However, most of the time, these exigencies are neglected due to lack of theoretical knowledge, fact that may lead the researcher to equivocated conclusions. Thus, it was made an analysis of an experimental data in entirely random blocks with four soil water treatments (25, 50, 75 e 100% of ETo) and four repetitions summing up 96 1 Doutor, professor UFCG/CES/UAE.