College Press Service
If you ask Van Williams, his college entrance test scores do not reflect his academic talent or potential.
Williams, a senior African-American at DuSable High School on Chicago's South Side, has a 3.6 grade point average and is a member of the school's academic decathlon team. Yet he scored a 17 on the American College Testing and a lackluster 980 on the Preliminary Scholastic Aptitude Test, the sole determining factor for students applying for National Merit Scholarships.
"My score, in essence, doesn't reflect what I am as a student and is nowhere a reflection of what I know," Williams said. "The math problems may be universal, but the stories and issues in the reading portion are hard for minorities to relate to."
Williams is one of thousands of students who feel they are being shortchanged by a college admission system that places too much emphasis on standardized test scores -- not on the individual accomplishments of the person. Yet many college officials say standardized tests add an element of consistency to an otherwise subjective selection process.
"It's a measuring stick that everyone uses," said Lisa Hibbs, athletic academic coordinator at the University of North Carolina-Charlotte.
Although the differences in Scholastic Aptitude Test and ACT exam scores between groups has narrowed somewhat in the last two decades, males have historically scored higher than females, and white students higher than African-Americans.
Mean SAT scores for males in the high school graduation class of 1995 were 429 on the verbal portion of the test and 503 on the math; females scored 426 and 463, comparatively, on the exam's 200- to 800-point scale. The gap between white and African-American scores is more marked: Whites scored 448 on the verbal and 498 on the math, while blacks averaged scores of 356 and 388, respectively.
On the 1995 ACT male grads outperformed females by just three-tenths of a point, 21 to 20.7 on the test's 36-point scale. The average composite scores for whites and blacks, however, were 21.5 and 17.1.
"Girls score lower, despite the fact that they get higher grades than boys in both high school and college," said Bob Schemer, director of public education for the National Center for Fair & Open Testing (FairTest), a Cambridge, Mass.-based organization that advocates making the tests optional in college admissions.
"Research suggests that fast-paced, multiple choice formats tap into skills and knowledge that boys have more of. Girls are less inclined to give a quick first answer, a strategy that's rewarded on both tests. In college or life, it's necessary to contemplate shades of meaning or puzzle out a problem from context, but that will hurt you here."
Minority students that aren't part of "mainstream society" are also at a disadvantage, said Schemer, because language differences force them to take more time answering questions.
However, Gretchen Rigol, executive director of admissions and guidance services for the College Board, which administers the SAT through the Educational Testing Service, said Fair Test's assertions aren't grounded in reality.
"The myth that's developed about multiple choice formats working against females and minorities is both a sexist and a racist assumption," she said. "Girls don't fall apart under pressure any more than boys -- these stereotypes are self-fulfilling prophesies. There is no statistical basis to support differences in guessing patterns.
"The big problem is that someone has billed the SAT way out of proportion. It's simply intended to help aid the transition to college, to provide a yardstick for where the student fits in. In any case, test scores are rarely used as the single determined factor in admissions."
To underscore that point, the College Board does not endorse the establishment of cut-off scores by colleges to narrow fields of applicants, and recommends considering scores alongside grades and other aspects of a student's academic record.
In an effort to avoid the slightest hint of cultural bias on exams, both the College Board and ACT programs employ panels of racially and ethnically diverse experts to screen each question, said Kelley Hayden, the ACT's director of corporate communications. Questions are then pre-tested; if disparities in the response among different groups of test-takers crop up, the question is dropped.
"We've been sensitive to the question of bias for years, whether it's found to exist in overall content or a single phrase," Hayden said. "We make every effort to ensure the tests are multicultural and don't under- or overestimate any one group's performance.
"It's our feeling that the bias exists in the system, in society. For example, test scores do go up with family income, but is that a bias against poor people?" asked Hayden. "Well, no. It means people with greater income have more advantages, including access to better schools. It comes down to a matter of preparation. Students who've taken the proper amount of core classes -- English, math, social science and natural science -- will naturally score higher."
Hayden said the ACT's gender gap is closing, in part, because more girls are taking advanced classes in science and math -- areas in which males have traditionally posted the highest scores. "If girls can't do as well as boys," he added, "why have they outscored them on the English portion of the test for 35 years?"
Still, there is evidence to suggest that standardized tests can adversely affect the performance and, consequently, collegiate status of women and minorities.
A recent study at the University of California at Berkeley, which bases the first 50 percent of its undergraduate admissions solely on a composite average of SAT scores and high school grade point average, found that the school's admissions formula underpredicted women's subsequent GPAs and reduced the number of females entering each class by 5 percent, or by about 200 to 300 students.
Meanwhile, Claude Steele, a professor of social psychology at Stanford University, has attracted the attention of media and scholars alike with his theory of "stereotype vulnerability," the expectation that one's membership in a stigmatized group will impede individual performance in test-taking situations.
Following a seven-year research project, Steele concluded that situational factors (asking students to check off their race on a form, for example, or telling students that a math test may show gender differences) in the test-taking experience can depress the academic performance of women and African-Americans in college environments. His research refutes the arguments that content bias is the culprit for differences in achievement. In addition, Steele's research also casts doubt on theories that racial differences in intelligence test performances are genetically based, as was posited by Richard J. Herrnstein and Charles Murray in their controversial 1994 book The Bell Curve.
FairTest claims that more than 200 U.S. colleges and universities have provided high school graduates with an alternative by making SAT and ACT scores optional in admissions evaluations. Such schools range from exclusive, private colleges to larger state-governed institutions in Oregon and California.
Peter Burns, director of admissions at tiny, progressive-minded Goddard College in Plainfield, Vt., said the school's test-optional policy has not resulted in a less qualified pool of undergraduate applicants.
"Test Scores are meaningless in determining how successful a candidate will be," Burns said. "We believe in evaluating the whole person, and rely heavily on personal essays and interviews."
Goddard's policy certainly benefited Gyllian Pressey, a sophomore and classical violinist who recently transferred from a state school in Maine. Although an honors student in high school, Pressey's low SAT scores caused her first-year college to place her on academic probation during her first semester. When she decided to transfer to Goddard, she was relieved she didn't need to submit any standardized test scores.
"I completed my freshman year with a 3.8 GPA," she said. "But I did terrible on my SAT, so I didn't bother to send it in."
In the future, student achievement may be assessed differently. Nationwide reform movements involving performance-based assessment of elementary and high school students are researching different ways to measure students' abilities.