The objective of this work was to verify the increase of corn cultivar discrimination by using the final plants population to adjust grain yield. The adjustment, or not, of the corn cultivars grain yield considering the population variate for 31 corn trials was made with the methods: without correction, three rule, Zuber method, covariance of the average population, covariance of the ideal population, method proposed by Cruz, method proposed by Vencovsky & Cruz and the method of stratified correction. The Fasoulas differentiation index was used to verify the cultivar discrimination improvement of each method in relation to the data without correction. The profit depends on the method used to correct the grain yield, producing greater profits with the methods of stratification corrections, covariance for the ideal population and covariance for the average population. Bigger profits are obtained in assays with heterogeneous populations, however the cultivar discrimination is smaler than the obtained in assays with homogeneous populations. Assays with populations more spaced away from the desired levels allow greater profits, but lesser discrimination in relation to the assays with adequate populations.
Zea mays; experimentation; variety trials; statistical methods
