More test takers, lower SAT scores

SAT scores declined slightly this year, as a wider range of students try the college-admissions test.

Last spring’s seniors scored on average 502 out of a possible 800 points on the critical reading section of the country’s most popular college entrance exam, down from 503 for the class of 2006. Math scores fell three points from 518 to 515.

Scores also fell three points on the writing section, which is still in an experimental stage, from 497 to 494.

A rising percentage of test takers speak English as a second language, come from non-college-educated families and request a fee waiver. Maine, which now requires all high school students to take the SAT, saw a sharp drop in scores as a result.

Male students, who make up 46 percent of test takers, outscore female students in math (533 to 499); boys are more likely to take advanced math classes. Female students do better in writing (500 to 489). Reading scores are similar: 504 for boys, 502 for girls.

About Joanne


  1. “as a wider range of students try the college-admissions test.”

    This is inaccurate. You don’t know why scores declined. The College Board suggests that this is the cause, but they have no evidence. It’s very unlikely, in fact.

    The most likely reason for the decrease–which is minimal–is that they reduced the number of questions on the test when they changed it. They really need to address that problem.

  2. Cardinal Fang says:

    But Cal, how can that be? Scores declined this year from last year, yet both this year’s students and last year’s students took the new test.

  3. It’s unlikely? Are you serious?

    As more students take the test, do you really think the distribution of new students (who otherwise wouldn’t have taken the test) are evenly distributed over academic levels? Isn’t it more logical and intutitive to think the increase in students are largely from mediocre academic backgrounds?

    I agree it would be impossible to prove this statistically but there are a lot of reasonable assumptions that are impossible to prove statistically.