To the Editor:
In his Commentary “UnSATisfactory: Why Education’s Most Famous Test Fails the Test “ (June 14, 2006), Walt Gardner is willing to accept assumptions, the advertised claims of commercial coaching firms, and statements from an unpublished study over fact and dozens of published scientific studies. In fully embracing these unsubstantiated statements and dismissing accumulated evidence, he promotes folklore over science, accepting misstatement and perception as fact.
There have been 10 separate, peer-reviewed studies published of the effects of coaching on the SAT across different groups of students. The majority of these studies were conducted by academic researchers with no affiliation to the College Board. They all have concluded that, after about 40 hours of classroom time, additional homework, and fees as high as $1,000 per student, coaching increased scores on the SAT verbal section (now “critical reading”) only by from 8 to 15 points, and on the SAT math section only by from 15 to 23 points. The most recent study was conducted by Derek C. Briggs, a professor at the University of Colorado at Boulder. Mr. Gardner ignores all of this research.
Mr. Gardner also cites a study by one selective, test-optional college demonstrating that, after 20 years, there was essentially no difference in graduation rates and only 0.05 in grade point average (on a 4.0 scale) between those who did and did not submit admission-test scores. He implies that this single study dismisses the accumulated evidence for validity of admissions tests from hundreds of studies at hundreds of colleges.
Meta-analyses of these studies confirm that admissions tests predict freshmen grades, cumulative grades, persistence, and graduation rates nearly as well—and better for some disciplines—as four years of high school grades. But the grades vs. test scores argument is a false dichotomy, since colleges can use both grades and test scores. When they do, validity is maximized for all groups of students.
Mr. Gardner does not offer a solution, but rather takes the position that admissions tests are not useful. This year, 42 percent of college-bound seniors taking the SAT reported having an A average in high school and a mean GPA of 3.33. At some elite private colleges, about 70 percent of students who sent SAT scores reported having a high school GPA of 4.0 or higher. When the vast majority of a college’s applicants all have A’s, grades are no longer helpful in distinguishing among applicants. Such institutions are forced to use other factors to make admissions decisions, such as activities, the rigor of courses completed, and the quality of the high schools attended.
Unfortunately, students have little influence over the quality of their high schools, and in some instances lack access to rigorous courses like Advanced Placement and opportunity to engage in school activities. Eliminating the only standardized measure available is not helpful to anyone. This point is well understood. In fact, the vast majority of colleges continue to use the SAT.
Educational discussions, including those about college admissions, are too often based on misinformation and speculation. Mr. Gardner reinforces this harmful, anti-scientific practice of characterizing anecdote as fact by ignoring research that could illuminate both the benefits and the limitations of standardized testing, grades, student essays, references, and other tools. We need to consider facts when debating the important educational decisions that are made concerning our children.
Vice President of Research
The College Board
New York, N.Y.
A version of this article appeared in the July 12, 2006 edition of Education Week as Critic of SAT Favors Anecdotes Over Facts