Test items that everyone gets "right" or everyone gets "wrong" provide . no basis for a comparison of test takers' abilities proof the test is not biased against minorities evidence the test yields a wide range of scores support for the validity of the test questions