The class of 2010’s SAT scores were flat, except for Asian-Americans, already the highest-scoring group. From the Wall Street Journal:

Overall, the average score for the graduating class of 2010 in reading remained at 501; climbed in math to 516 from 515; and dropped in writing to 492 from 493, according to scores released Monday.

However, Asian-American students widened their lead by gaining three points in reading, six points in writing and four points in math. “More than two-thirds took at least four years of science in high school, versus 59% of all test-takers, and 48% of the Asian-Americans took calculus, versus only 28% of the rest of the pool,” College Board officials told the Journal.

Not surprisingly, students who took college-prep courses outperformed those who didn’t. The number of test takers rose by 1.2 percent.

Scores mean little without the questions. Years ago, the College Board recentered the SAT. Same for the NAEP Math, which, I have read, raised scores by the simple device of reducing the number hard questions (addition and subtraction of fractions with different denominators) from the 4th grade Math assessment.

Still, relative rank (by race, parents’ level of education, family income) tell us something. Also, scores of the top-performing students, by race, tell us something.

malcolm, if the questions were getting easier or standards were being lowered across the (College) board, why are scores flat then? wouldn’t you expect all race groups to raise their scores?

so are asians getting smarter, richer, more hardworking, or some combination thereof?

My guess is that Asians are still paying for test prep while whites aren’t.

I’m neither shocked or surprised at the SAT scores for Americans of Asian descent. I work as a volunteer in 7th & 8th Grade American History classes in a Chicago suburban school district. For the last six years I’ve observed that American of Asian descent or even recent immigrants far more often than not, are more serious about education, work harder, pay attention, contribute more in group learning activities and are the least disruptive of any groups that we pay attention to. Must be something in the water.

Scores are flat because measures inflate while performance drops. That’s my guess, anyway.

Black students in Massachusetts made significant gains.

http://tinyurl.com/2g4cqwz (links to a Boston Globe article)

“Black students made strong gains in all three subjects covered by the test in 2010 when compared with 2009: an eight-point gain in critical reading, a nine-point gain in mathematics, and a seven-point gain in writing.

The math and writing gains outpaced national increases and means African-American students on average now outscore their peers nationally in both sections of the college entrance exam.”

I venture to say that dropping the MCAS is a really, really stupid idea.

Perhaps this is a silly question, but wouldn’t you expect the average of the entire test-taking population on a norm-referenced test, such as the SAT, to lie around the 50th percentile?

On the sample used to norm it, yes.

I would (naively?) assume that the sample chosen to norm the exam would be chosen to represent the whole as closely as possible?

Am I missing something? It is big news when the average of the 2009 SAT population is… around the 50th percentile? (http://professionals.collegeboard.com/profdownload/SAT-Percentile-Ranks-2009.pdf)

If the characteristics of “the whole” are changing, one would expect the average of some statistics to change also.