. . . "Davies is right. 12,000 people is a large number of people to survey." .