Question: Research suggests that hypertension adversely affects the speedy retrieval of newly learned information. One test involves showing subjects two, four, or six digits on a video screen and later asking whether a particular digit was part of the set. It is anticipated that hypertensive individuals will take longer to respond than nonhypertensive individuals. Consider forming two groups of equal size to test this. Suppose that the spread (SD) in response times is about three seconds and that researchers consider a mean difference of five seconds to be practically important. What sample sizes should be used to make a 95% confidence interval small enough to distinguish a five-second difference? (Data from "How Hypertension Affects Memory," Science News 143 (March 20, 1993): 186.)