Performance of Simple Linear Regression Analysis Under a Randomized Complete Block Design

Abstract

In experiments conducted under a randomized complete block design, the fitting of the simple linear regression model can be performed under different combinations of the number of treatments and the number of replications. In order to determine the best combination, considering the same number of experimental units, it was concluded through a data simulation study that the quality of the fit increases when regression is performed in experiments with fewer treatments and more replications. Therefore, for model fitting, if linearity is expected, it is recommended to use two treatments. Otherwise, three treatments are recommended. All of this applies to experiments with coefficients of variation between 10% and 30%.Keywords: Treatments, Replications, Experimental precision.