Testing on a Level Field
The developers of the Preliminary Scholastic Aptitude Test, in ordering a revision to eliminate bias against girls, have cast new light on the larger problem of school admissions based narrowly on standardized testing.
Boys generally outscore girls on the standardized PSAT exam. It is taken in the junior year of high school and is regarded as preparatory for the Scholastic Achievement Test, which is given heavy weight by many colleges in the admissions process. The PSAT also determines semifinalists for the prestigious National Merit Scholarship Program. Boys win the majority of those college scholarships, although more girls than boys take the test. The score gap is a result of higher math scores achieved by boys. It is generally acknowledged that social, cultural and historical pressures have discouraged girls from excelling at math.
The long-standing disparity resulted in charges of gender discrimination by the watchdog National Center for Fair and Open Testing, based in Cambridge, Mass. It also prompted the American Civil Liberties Union to apply pressure on the Educational Testing Service, which administers the test for the College Board. Without admitting gender bias, the testing service agreed to more measurements of writing skills. That is expected to narrow the gap because girls tend to do better than boys on tests of verbal ability.
The new PSAT exam, according to a statement from the College Board, “will help measure the varied talents of an increasingly diverse student population.” Educators have long argued about the ability of standardized tests to predict success in college for women and minority students. The PSAT revision goes in the right direction.
At the same time, the University of California is moving backward. A UC study shows that at least its premier campuses will see a marked decrease in black and Latino enrollment when racial and ethnic preferences are banned as of 1998. Test scores and grades alone are not enough. Merit comes in many forms and should be measured by many methods.
