QUALITY CONTROL OF CULTIVAR COMPETITION EXPERIMENTS THROUGH THE ANALYSIS OF THE STATISTICS EMPLOYED

• Main Author: LÚCIO, ALESSANDRO DAL'COL
• Publication Date: 1999
• Other Authors: STORCK, LINDOLFO, BANZATTO, DAVID ARIOVALDO
• Format: Article
• Language: por
• Source: Pesquisa Agropecuária Gaúcha (Online)
• Download full: http://revistapag.agricultura.rs.gov.br/ojs/index.php/revistapag/article/view/424
• topic: competition of varieties, maize, rice, soybean, wheat, beans, barley, oat, triticale, sorghum.; competição de variedades, milho, arroz, soja, trigo, feijão, cevada, aveia, triticale, sorgo.

The objective of this work was to study the distribution of the statistics coefficient of variation (CV) and minimum significam difference (MSD) as a way to obtain the limits for quality control, in competition experiments of maize, rice, soybean, wheat, beans, barley, oat, triticale and sorghum. Data of 2017 cultivar competition assays, conducted in the state of Rio Grande do Sul, Brazil, from 1987 through 1995 were included. The CV and MSD values were submitted to the Lilliefors test. The class limits of these statistics were established based on the averages and variante of the variables tested. Using the analogous management criteria for conduction of the experiments, each assay was dismembered for each crop evaluated. We concluded that the diversity of management procedures employed in cultivar competition experiments adversely affects their quality control. This control must be performed according to the crop and management employed. No information is lost when CV and MSD classified as too high are disregarded.