Sample sufficiency for estimation of mean of productive traits of rye

Abstract

The objective of this work was to determine the sample size (number of plants) to estimate of the mean rye traits, evaluated in the grain harvest period. Eight uniformity trials (blank experiments) were conducted. Five with the cultivar ‘BRS Progresso’ and three with the cultivar ‘Temprano’. At the stage of grain maturation 780 plants (total of eight trials) were randomly collected and 17 traits were evaluated in each plant. Descriptive statistics were calculated and the sample size to estimate the mean at amplitudes of the confidence interval of 95% (ACI_95%) 5 (greater precision), 10, 15, 20, 25, and 30% (lower precision) of the mean - was determined by resampling with replacement. There is variability of sample size between traits, cultivars, and sowing dates. The sample size for the cultivar ‘BRS Progresso’ is lower than for the cultivar ‘Temprano’. For the mean estimate, for all traits and sowing dates, with ACI_95% maximum of 15, 20, 25, and 30% of the mean, 352, 197, 127, and 87 plants of ‘BRS Progresso’ cultivar and 609, 341, 227, and 154 plants of the ‘Temprano’ cultivar are required, respectively.