Gender Bias Victory Wins Millions for Females But National Merit Test Remains Biased

University Testing

Hundreds more young women received prestigious National Merit Scholarship awards this year due to changes in that competition's qualifying test which were made to settle a FairTest gender discrimination complaint. But remaining bias in the exam still costs females millions of dollars in scholarships they would have earned on a level playing field based on genuine academic "merit."

More than 45% of the nearly 7600 college tuition awards made to the high school class of 1999 will go to females, according to a FairTest analysis of the names of semifinalists. In past years, young women won fewer than 40 percent of National Merit Scholarships, even though females earn higher grades than their male counterparts in both high school and college when matched for identical courses.

Scores from a three-hour, largely multiple-choice exam, the Preliminary SAT/National Merit Scholarship Qualifying Test (PSAT/NMSQT), are the sole factor used to determine semifinalist status for about $28 million in annual awards.

Historically, the male-female ratio of scholarship winners has closely tracked the gender distribution of the semifinalist pool.


In 1994 FairTest filed a complaint with the U.S. Department of Education's Office for Civil Rights (OCR) charging the Educational Testing Service and College Board with illegally discriminating against females for designing and administering the PSAT/NMSQT (see Examiner, Winter 1993-94) in a manner that violates the Title IX gender equity requirements of the Civil Rights Act. To settle that case, the test-makers agreed to add a multiple-choice "Writing Skills" section to the exam beginning with the high school class of 1999 (see Examiner, Fall 1996) and use results from it in calculating scholarship eligibility. Due to these changes, the gender gap in this year's qualifying test scores narrowed by 40% from last year's level (see Examiner, Winter 1997-98).


Continuing Battle

The fact that such minor changes in one exam can have so profound an impact clearly shows that test scores do not measure merit. But National Merit's reliance on a one-shot exam, albeit a modestly reformed one, continues to cheat young women. More than 56% of students in the Class of 1999 who took the PSAT/NMSQT are females, and they earned better average grades than males while taking equally rigorous high school course loads. Except for the biased tests, a majority of scholarships would have gone to young women.

The ongoing, unjustified gender disparity may spur further federal action. In its settlement agreement with the test-makers, OCR maintained continuing jurisdiction over the case to assure that the revised exam "sufficiently addresses the allegations" in the original complaint. FairTest is not convinced that a modest reduction in the PSAT/NMSQT gender gap leading to some improvement in the National Merit Scholarship award ratio is sufficient to meet the Title IX legal standard, though it is a step in the right direction.


The narrowing of the PSAT/SAT gender gap due to adding the "Writing Skills" section raises several other important policy questions. Why, for example, have similar changes not been made on the SAT, the GRE and related exams which show comparable bias? Even more fundamentally, why are instruments on which results can be so quickly "adjusted" ever used as the sole or primary factor to determine college admissions or award scholarships?


Coupled with the recent court decision suspending the National Collegiate Athletic Association's test score-based initial eligibility and scholarship rules because of racial discrimination (see related article), the National Merit data should serve as a warning to any program which relies exclusively on standardized exam scores to make high-stakes decisions. Such practices risk legal action for violating the anti-discrimination provisions of federal civil rights laws, both Title VI and Title IX.


The Citizens Scholarship Foundation of America (CSFA), which runs local "Dollars for Scholars" programs, awards about as much money each year as National Merit, but does not rely heavily on test scores to select its winners. As a result, the majority of CSFA award recipients are female (see Examiner, Summer 1994).


Despite this example of how to run a competition fairly, the National Merit Scholarship Corporation (NMSC) continues to resist any process other than reliance on test scores. Rather than overhauling its selection mechanism, NMSC has spent its staff time warning other institutions that it holds the trademark to the slogan "Merit Scholarship" and threatening colleges which use that term to identify their own programs with legal action.

* FairTest's Fact Sheet on the National Merit scholarship is available by clicking here.