Cuadernos de Investigación UNED
versão impressa ISSN 1659-4266
Resumo
VARGAS ROJAS, Jorge Claudio e NAVARRO FLORES, Juan Ramón. Size and shape of experimental units for rice (Oryza sativa) yield trials, in Guanacaste, Costa Rica. Cuadernos de Investigación UNED [online]. 2019, vol.11, n.3, pp.355-360. ISSN 1659-4266. http://dx.doi.org/10.22458/urj.v11i3.2653.
“Size and shape of experimental units for rice (Oryza sativa) yield trials, in Guanacaste, Costa Rica”.Introduction: The size and shape of an experimental unit affect the accuracy of field trials. Objective: Finding the appropriate size and shape of an experimental unit for rice yield trials in Bagaces, Guanacaste, Costa Rica. Methods: We applied a multiple regression analysis to data from an uniformity trial. The residual variability, measured as a coefficient of variation (and related to different sizes and forms of a replicable experimental unit), was modeled according to length and width. The next step was to determine partial derivatives per regressor and adjust them to -1. With the equation system it is possible to obtain a combination of width and length which decreases the coefficient of variation. Results: For this work, the resulting combination was 2,93m long by 7,95m wide. Conclusion: In practical terms, the size of the experimental unit that decreases the coefficient of variation is 3m long and 8m wide.
Palavras-chave : Plot size; plot shape; rice; multiple regression; experimental error.