Do you think that selecting a sample of three thousand individuals could yield an accurate picture of the eating habits of Americans? Why or why not?