Suppose we want to compare the taste preferences of consumers for three different brands of bottled water (say, Brands A, B, and C) using a random sample of 15 bottled water consumers. Set up a completely randomized design for this purpose—that is, assign the experimental units to the treatments for this design.