It would also be possible to use a block design in this study. Sort the before CFU counts into three blocks (representing low, medium, and high CFU counts). Each block should have 10 units. Analyze the data using the three blocks to test for differences in the hand- washing treatments. This is an unbalanced (i. e., unequal sample size in each group) block design with one blocking factor and one cleansing factor—no covariate should be used.