Last week, the sponsor of the SAT college admissions exam claimed that a seven point average score decline, the largest drop in 31 years, was caused by a reduction in the number of high school seniors who took the test more than once. Speaking at a National Press Club news conference College Board president Gaston Caperton said that three percent fewer students took the test multiple times, noting that scores typically increase by about 30 points on a second administration.

In a letter delivered to Caperton this morning, the National Center for Fair & Open Testing (FairTest) charged “the College Board’s math does not add up” and offered an analysis demonstrating, “the reduction in students re-taking the SAT cannot explain even one point of the seven point score plunge.”

The FairTest letter explained that if 44,000 additional students had taken the SAT multiple times and if their scores had gone up an average of 30 points, as the College Board projected, overall scores would have improved by nine-tenths of a point. “That means less than one point of the decline, not seven, was due to the reduced number of re-tests.”

“It appears that the College Board has once again given the public inaccurate, self-serving information, just as it did when the SAT scoring error was first disclosed earlier this year,” wrote FairTest Public Education Director Robert Schaeffer. “Clearly, factors other than the re-testing pattern must have played more significant roles. Was it the additional length of the exam leading to student fatigue, increased difficulty stemming from new items types, or some other flaw in the hastily redesigned SAT, which the College Board rushed to market after the University of California threatened to stop requiring the test?”

The official College Board news release on this year’s SAT scores, which is still posted on the firm’s web site, claims, “Much of the score difference this year can be attributed to this decline in the number of students retaking the test and gaining the advantage of a score increase.” The FairTest letter said the College Board’s statement is “not factually accurate.”

Schaeffer continued, “Test-takers and their parents who paid more than $100 million in registration fees for the new’ SAT deserve an accurate explanation for the score decline, especially since the College Board repeatedly promised that results from the revised exam would be consistent with its predecessor. So do the admissions and scholarship officials who relied on the scores as well as the journalists who reported the College Board’s false claim.”

FairTest promised to make public any response it receives from Caperton. “Our letter asks the College Board to correct us if we are wrong.’ If they cannot do so, their credibility  and that of the SAT  will further erode,” Schaeffer said.

A copy of the FairTest letter to the College Board follows

September 5, 2006 Gaston Caperton, President The College Board 45 Columbus Avenue New York, New York 10023 via overnight delivery

Dear Governor Caperton:

At last week’s National Press Club presentation of College Bound Seniors 2006 SAT results, considerable attention was focused on a seven point score decline, the largest drop in 31 years. The Board’s news release stated, “Much of the score difference this year can be attributed to this decline in the number of students retaking the test and gaining the advantage of a score increase,” noting that three percent fewer students took the exam multiple times and that scores increase by about 30 points on a second administration.

Unfortunately, the College Board’s math does not add up. A simple analysis reveals that the reduction in students re-taking the SAT cannot explain even one point of the seven point score difference. Thus, the College Board’s public statement is not factually accurate.

The calculation is straight forward. The total number of high school seniors who took the exam at least once was 1,465,744, as reported by the College Board. If three percent fewer re-tested, that would come to approximately 44,000 students. If the scores of 44,000 students each rose 30 points, that would produce a total gain of 1,320,000 points (44,000 x 30). Dividing that increase by the number of test-takers, yields an average gain of nine-tenths of a point across the test-taking pool (1,320,000/1,465,744).

That means that less than one point of the decline, not seven, was due to the reduced number of re-tests.

Please correct us if we are wrong, but it appears that the College Board has once again given the public inaccurate, self-serving information, just as it did when the SAT scoring error was first disclosed earlier this year. Clearly, factors other than the re-testing pattern must have played more significant roles. Was it the additional length of the exam leading to student fatigue, increased difficulty stemming from new items types, or some other flaw in the hastily redesigned SAT which the College Board rushed to market after the University of California threatened to stop requiring the test?

Test-takers and their parents who paid more than $100 million in registration fees for the “new” SAT deserve an accurate explanation for the SAT score decline, especially since the College Board repeatedly promised that results from the revised exam would be consistent with its predecessor. So do the college admissions and scholarship officials who relied on the scores as well as the journalists who reported your false claim.

On behalf of all these groups, the National Center for Fair & Open Testing (FairTest) calls on the College Board to tell the American public the real reasons for this year’s SAT score decline.

Sincerely,

Robert A. Schaeffer Public Education Director

>

About the Author:

eSchool News