Suppose that a placebo gives headache relief to 15% of patients who take it. A drug company tests a new drug using a sample of 1,000 people. They find that the new drug gives relief to a higher proportion of people, and the difference is statistically significant.