Tamanho de amostra para a estimação do coeficiente de correlação linear de Pearson entre caracteres de mamoneira

Abstract

In the study of linear relationships, it is important to define correctly the sample size, to estimate the Pearson correlation coefficient among pairs of characters, with acceptable reliability. The aim of this research was to determine the sample size (number of plants) to estimate the Pearson correlation coefficient among 21 characters of castor bean. It was evaluated 41 and 55 plants of the Sara and Lyra hybrids, respectively, regarding to the characters of seed, seedling, adult plant and yield in Santa Maria, Rio Grande do Sul State, Brazil, in the agriculture year of 2007/2008. It was calculated the sample size through bootstrap resampling, to estimate the Pearson correlation coefficient, for the amplitudes of the bootstrap confidence interval 95% equal to 0.20, 0.35 and 0.50, for each of 210 pairs of characters. Regardless of the castor bean hybrids, 96 plants are sufficient to estimate the Pearson correlation coefficient, to the amplitude of the bootstrap confidence interval 95%, maximum of the 0.52.