Researchers gave a spatial reasoning test to men and women at Lafayette College in Easton, PA, to determine if men or women were better at this skill. But FIRST they asked the test-takers some statistical questions - one group were asked their genders, a second group were asked why they chose a liberal arts college, and the control group were asked about their childhoods. The control group had "typical" scores for the test - men scored 15-20% higher. Those asked about their gender beforehand had men scoring 25-30% higher. And those asked why they chose a liberal arts school, a question which allowed the testers to think of their strengths, nearly closed the gap between men and women. Go figure!
Perhaps statistical questions should be always asked at the end of an exam, then.