Experimentation on a plot of land is shown in Figure 5. Suppose that we are considering
two factors (A and B) and each has three levels, requiring nine treatments in total. If a
completely randomised design is used, all nine treatments are randomly assigned to nine
subplots of land. If the split plot arrangement is used, factor A is used as a whole plot
factor, and its three levels are randomly assigned on three plots of land with levels of
factor B randomly assigned within each subplot. In split block, three levels of both factorA (A1, A2, and A3) and B (B1, B2, and B3) are randomly assigned across the land.