Tamanho de amostra e número de repetições para avaliação de caracteres agronômicos em milho-pipoca

• Autor(a) principal: Catapatti, Thiago Rodrigues
• Data de Publicação: 2008
• Outros Autores: Gonçalves, Manoel Carlos, Silva Neto, Maurílio Rezende, Sobroza, Rubens
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1413-70542008000300023; http://repositorio.ufla.br/jspui/handle/1/6703
• topic: Amostragem (Estatística); Milho de pipoca; Sampling (Statistics); Popcorn

An adequate experimental design and planning are essential in order to obtain high quality data. Experimental plan ensures correct plot size and shape as well as adequate number of replications and samples. The optimun sample sizes have increasing importance, due to the research cost implication and attempts of keeping to a minimum the research error, remaining low the errors committed in the inferences. The adequate number of replications is of basic importance for minimizing experimental error and increase experimental precision. The objective of our research was to evaluate the ideal sample size for characters of pre- harvest characteristics in popcorn and to study the implications of the number of replications in the residual variance in quality and productivity characteristics. The following genotypes were used: variety BRS ÂNGELA, triple hybrid ZÉLIA, variety UFV-M2 and advanced population of triple hybrid ZÉLIA. The experimental design was the one of completely randomized blocks with two, three, four, five and six replications in a split plot treatment design. The genotypes were laid out in the plots and the sample sizes were laid out in the subplots (5, 10, 15, 20 and 25) plant plot-1. It was observed that, to evaluate pre-harvest characters one may be assessed from 5 to 25 plants sample size, without impairing the experimental precision. With the increase of the number of replications there was increase in the amount of significant characters with 1% of probability, for the effect of genotypes. There was a greater amount of significant differences among means of genotypes, for the characters of pre-harvest and yield, for the numbers of replications greater than four. Thus five plants in the plot may be sampled, and four replications may be adopted.