It used to be that an acceptance letter from a good college was simply a pleasant prelude to the game of life. No more. In 21st century America, getting into the best universities has become a ferociously competitive, high-stakes game. This year, the University of California received 340,000 applications for 40,000 places. There are many more qualified students than selective schools can accommodate, and the hunt is on for the best students at public and private institutions alike.
But who are the best students? American colleges and universities have long answered this question by looking at applicants’ high-school grades in academic subjects and their scores on standardized college-entrance tests.
These tests come in two varieties: achievement and general reasoning. Achievement tests measure what students have learned in high-school courses, such as history, math, and foreign languages. General-reasoning tests seek to assess students’ academic potential by measuring their skills in solving reading and math problems largely, by design, independent of high-school curricula. Since 1926, the dominant general-reasoning test in the U.S. has been the SAT, sponsored by the College Board.
The SAT has a long pedigree in American higher education. Yet the problem with general-reasoning tests like the SAT is their premise: that something as complex as intellectual promise can be captured in a single test and reflected in a single score. It is tempting for admissions officers — and parents, legislators, policymakers and the media — to read more into SAT scores than the numbers can bear. Although measurement experts know that tests are only intended as approximations, the fact that scores frequently come with fancy charts and tables can create an exaggerated sense of precision.
For quite some time, an over-reliance on these scores has skewed the outcome of the admissions game. The more competitive admissions become, the more small differences in SAT scores affect a student’s chances. As a result, deserving students, including low-income and minority applicants, are crowded out of the game. These concerns led the University of California to consider eliminating the SAT entirely as a requirement for admission in 2001.
The College Board responded with a revised SAT, introduced in March 2005. The new SAT is a dramatic improvement over the old. The mathematics section is more demanding, but also more fair; while the old SAT featured questions that were known for their trickery but required only a beginning knowledge of algebra, the new math section is more straightforward and covers higher-level math.
Instead of deconstructing esoteric verbal analogies, students must now perform a task they will actually face in college: writing an essay under a deadline. These changes have already galvanized high schools and students to put more effort and attention into writing and college-preparatory math. The new SAT, in other words, has gone a long way toward becoming an achievement test.
But has it gone far enough? The College Board’s own recent assessment concludes that the new SAT is not substantially better than the version it replaced in its ability to predict student performance in the first year of college. Although the essay adds significant value to the new SAT, it appears the critical-reading section does not. The new SAT is almost an hour longer than the old SAT. And its content is still not as closely tied to college-preparatory curricula as a true achievement test should be.
The new SAT is looking more like a promising first draft than a final product. Any plans for revision should consider a series of University of California studies that have unsettled some entrenched assumptions about testing students’ readiness for college.
The studies, conducted over the past decade, suggest that achievement tests are better than general-reasoning tests in predicting how well students are likely to perform in college, that they are fairer to low-income and minority students, and that they reinforce teaching and learning in a way the SAT — even the new SAT — does not. Achievement tests help students understand where they are strong academically and where they need to improve — and that they can improve if they invest the time and work.
The most intriguing aspect of this research, however, is not what it says about tests but what it says about that old-fashioned admissions criterion: high-school grades. The studies concluded that a student’s performance over four years of high school remains the fairest and most meaningful measure of his or her accomplishments and the most reliable indicator of future success in college. We need standardized tests to correct for grade inflation and give students useful feedback. But we must be very careful about the tests we choose, and the California findings give us persuasive reasons to move toward achievement tests.
Like the new SAT, standardized testing is itself a work in progress. We present two possible routes for the future.
The first option is to revise the new SAT to keep the writing and mathematics sections but significantly reduce the critical-reading component. Along with this newer SAT, require students to take two achievement tests of their own choosing: candidates are the SAT Subject Tests and Advanced Placement (AP) exams, both offered by the College Board.
This strategy yields a shorter SAT while preserving its current strength in assessing two indispensable skills for academic success–writing and mathematics. It also tells students that they must be prepared to demonstrate not only an ability to write clearly and think quantitatively, but also mastery of two subject areas.
The second is not to require a single, comprehensive test at all, whether the new SAT or its long-standing rival the ACT. Instead, have students take a combination of achievement tests in various academic subjects, again using the SAT Subject Tests or AP exams, with a choice of at least some of them. This strategy recognizes a fundamental problem with any effort to develop a national achievement test: the absence of a standardized high-school curriculum in the U.S.
American College Testing, sponsor of the ACT, has sought valiantly to overcome this difficulty through national curriculum surveys, but the ACT does not measure student achievement to the same depth as do discipline-specific tests like the SAT Subject Tests or AP exams. It may be that no single examination, however well-designed, will be satisfactory in a country that lacks a national curriculum and has a long tradition of local control.
In the unrelentingly competitive world that college admissions has become, we owe students the chance to be judged on criteria as fair and rigorous as we can make them. The current ferment of research on standardized testing, including several major studies now underway, suggests that we may be on the verge of opening a productive new chapter in the long national conversation on what academic merit is and how it should be measured. One thing is clear: There is still a lot more to say. ♦
This article was originally published by Forbes magazine, August 13, 2008.