On the Effect of Blocking in Randomized Block Design: An application to Competition of Soybeans Varieties against Weeds

The research considered the effects of blocking in ramdomized block design with application in the competition of the varieties of soybeans against weeds in Northern Nigeria. It focused on the Randomized Complete Block Designs and compared it with these Designs when block effects are not considered. Data of the competition between Soybeans varieties against weeds were considered and rearranged in four different tables seen as table 1, table 2, table 3, and table 4 each with sample sizes of 24, 48, 72 and 96 respectively as seen in the example in this paper. Each of these tables were analyzed with or without blockings. The study considered ANOVA, Mean Square Error, F-test, Coefficient of determination, Akaike Information criterion (AIC), Schwarz Bayesian criterion (SBC) and Relative Efficiency of a pair of design. The effect of blocking in a design was seen to be of great importance, since in most cases, the analysis with blocks showed to be superior to those without blocking. The effects of sample size were also observed, as it was seen that changes in sample sizes have no effect on the calculated F-statistic and the larger the sample size the larger the values of AIC and SBC.