Rigging the Test

It's a familiar phenomenon among recreational carpenters: Miss the nail, hit the thumb, blame the hammer. It's this same blame-the-tool mentality that prompted the U.S. Department of Education's Office of Civil Rights to bully the Educational Testing Service (ETS) and the College Board into changing the Preliminary Scholastic Assessment Test (PSAT) so that girls' scores would improve.

The PSAT is used to determine the finalists for the prestigious National Merit Scholarships, which distributed $28 million in scholarships to 7,400 students last year. The problem, civil rights enforcers say, is that boys outperform girls on the test, producing a gender imbalance among National Merit Scholars. In 1997, for example, only 44 percent of scholarship recipients were girls. Equity activists and the OCR argued that adding a writing section to accompany the math and verbal sections would give girls a chance to even up the scores.

While the ETS and College Board went along with the wishes of the feds, an ETS study released last May (www.ets.org) suggests that nothing short of eliminating the test and drawing finalists by random lot is likely to produce a 50/50 gender split. The "ETS Gender Study" documented what researchers have long known: While the average (mean) performance of girls and boys is quite similar, there is greater variance in the scores of boys–in other words, boys are more likely than girls to be very strong or very weak academic performers.

For instance, there are five boys for every four girls in the top 10 percent and bottom 10 percent of the test distribution, and the gender split becomes more exaggerated as you move further up and down the scale. Since National Merit Scholarship recipients are among the top 1 percent of all test takers, it's not surprising that more boys than girls win the scholarships.