SAT Records Biggest Score Dip in 31 Years

The College Board released its report on the new SAT, which reports the biggest score drop in 31 years according to the Washington Post.

Some people will no doubt attribute the score drop to the longer length of the test, as did one student quoted in the Washington Post:

Anita Kinney, a Catholic University freshman who was one of nearly 1.5 million high school seniors who took the new SAT, said it was ridiculous to discount exhaustion. "The test is four hours long. Enough said," she said. "The members of the College Board obviously have not sat down and taken the new SAT."

On the contrary, I am sure the College Board tested the concept and made sure that fatigue would not contribute to a massive decline in scores. After all, there are a number of breaks in the test day. I am not prepared to accept the word of one student, or even a million students, who complain about the length of the exam. The College Board responded to such charges:

Research has shown that fatigue is not a factor. A College Board analysis of the performance of more than 700,000 test-takers on the critical reading and mathematics sections during the spring and fall 2005 SAT administrations showed no difference in student performance. There was no difference in either number of items correct or number of items omitted, between sections that appeared early in the test and sections that appeared later in the test.

That is what real statistics show.

So what caused the drop in scores? The College Board notes:

The most significant factor in the overall decline is mainly attributable to a change in student test-taking patterns, according to the College Board. The most notable change in test-taking behavior involved a decrease in retesting. Typically, students who take the test a second time see a 30-point increase on their combined score. 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.

I tend to think that the new test causes people to either take the old test before implementation or wait a cycle to see what will happen with the test. Thus the decline in the number of test takers is likely to an aberration.

There is some good news:

Students for whom English is a second language increased 5 points in critical reading to 467 and 2 points in mathematics to 523. They scored 469 on the writing section. Although this is 28 points below the mean for all students, it is a smaller gap than these students experienced for critical reading. More ESL students took the SAT this year.

snip

African American and Mexican American critical reading scores, 434 and 454 respectively, improved over last year by one point each. American Indian mathematics scores improved by 1 point to 494 and Mexican American mathematics scores improved over last year by 2 points to 465.

But my concern about reading skills is coming home to roost:

The largest declines in critical reading were seen among White and Other Hispanic students, both down by 5 points to 527 and 458, respectively.

snip

Males experienced a greater score decline than females in critical reading across all ethnic groups. In addition, males outscored females in critical reading in all ethnic groups except for African American in which females scored 7 points higher than males.

While the reading section was altered to include longer passages and more questions related to those passages, I still don't think the change can be fully attributed to changes in format. The College Board no doubt did some significant testing of the new exam to ensure consistency in these areas, so a majority of the decline has to be a loss of skill over time.

I am very troubled by these results and I do hope they are an aberration, but I am not convinced.

Be sure to check out some of the facts at a glance on teh College Board's Website, like this one:

Forty-three percent of 2006 college-bound seniors reported grade averages of A+, A, or A-. Ten years ago, the figure was 36 percent, and in 1987, the first year these data were collected in the same manner, the figure was 27 percent. This year's average grade point average was 3.33, compared with an average GPA of 3.21 in 1996 and 3.07 in 1987.

Hmm--makes you wonder about grade inflation (both by the schools and the students) and the value of a grade point average as a determinant of success on the SAT. If seven percent of students have a better GPA, why are scores going down?