Dominic wants to test if one of two types of fertilizer (a standard fertilizer and a new product) is better for producing melons
of equal weight after a three-month fertilization program. Dominic has two identical land plots (1 and 2) and is considering how to assign 30 melon plants, each of which has a numbered tag, to the fertilizer. What is the best way to do this as a matched-pairs design? A. Put all the odd-numbered melons in plot 1 and the even-numbered melons in plot 2. Then give the standard fertilizer to plot 1 and the new product to plot 2.
B. Assign the melons completely at random to the two plots using a coin flip, heads meaning plot 1 and tails meaning plot 2. Then give the standard fertilizer to plot 1 and the new product to plot 2.
C. Obtain pairs of melon plants whose seedling heights are roughly equal at the start of the experiment and randomly assign one of the pair to plot 1 and the other to plot 2. Then give the standard fertilizer to plot 1 and the new product to plot 2.
D. Obtain pairs of melon plants whose seedling heights are roughly equal at the start of the experiment and put the taller of each pair into plot 2. Then give the standard fertilizer to plot 1 and the new product to plot 2.
E. Divide the 30 melon plants into two groups, with the 15 tallest seedlings in one group. Randomly choose which plot to assign the tallest plants and assign the smallest plants to the other plot. Then give the standard fertilizer to plot 1 and the new product to plot 2.
A. is wrong because having a different height for each seedling and not trying to keep it equal is simply worse than making sure the heights are identical.
B. Completely wrong because it would likely result in 1 plot having more melon seedlings than the other.
C. Correct because it makes sure the seedlings are the same height at the beginning of the experiment, allowing less randomness in the experiment.
D. Putting the taller of a each pair for 1 plot makes it the experiment rigged in favor of the plot with taller seedlings
