Once researchers have developed hypotheses, they can be tested. As we have already seen, a hypothesis suggests relationships between variables. Suppose we hypothesize that men and women drink different amounts of coffee during the day during finals. The independent variable in this case is gender, while the dependent variable is number of cups of coffee. We collect data and find that to average number of cups of coffee consumed by female students per day during finals is 6.1, and the average number of cups of coffee consumed by males is 4.7. Is this finding meaningful? The answer appears to be straight- forward (after all, 6.1 is larger than 4.7), but sampling error could have distorted the results enough so that we conclude there are no real differences between men’s and women’s coffee consumption.