recently, the college board announced that the latest sat scores from the class of 2015 show that 6 out of 10 graduating seniors are woefully unprepared for college. the average sat score for the class of 2015 was 1490 out of 2400, down 7 points from last year and the lowest since the test was redesigned in 2005. meanwhile, the college board’s frenemies at the act pessimistically announced that act scores also remained stagnant this year.
whoa, this sounds bad. and it’s been all over the news. should we all be equally freaking out over the state of american education? well, the answer is “no.” at least not in the way the college board and the act would like you to be.
you see, the college board and the act have a vested interest in making the case that their respective tests are an accurate measure of academic achievement. in the past few years, the act has overtaken the sat in popularity. and in response, the sat is going to be drastically changing in march 2016 in the interest of reclaiming some of this market share. the new sat is supposed to be aligned with current common core standards in schools (and is startlingly act-like in many ways). in other words, the sat is working extra hard to fight for its relevance.
every summer in recent memory, alarmist news reports about the state of sat and act scores flood the media. but it isn’t helicopter parents or indignant students driving these reports. it’s the college board and the act. it’s pretty smart on their part. such reports make the automatic assumption that the act and sat are accurate measures of college potential. and this is a rather questionable assumption, as many teenagers who consider themselves to be excellent students but poor test-takers would not hesitate to point out. this is not to say that there aren’t some correlations between test scores and academic potential, but the sat and act are far from fair judges.
test scores are declining because more students are taking the test
ok, so the tests aren’t entirely fair, and there’s an ulterior motive behind widespread reports about declining scores–but why are scores decreasing?
the simplest answer is that there is a far more diverse group of students taking the sat and act, including more low-income and underrepresented groups. the number of students using a fee-waiver, indicating they are low income, was at its highest in recent years, and the number of underrepresented minority students increased to 32.5%, compared to 29% in the class of 2011. and this is good news, indicating more students than ever before are being encouraged to pursue higher education. as andrew ho, harvard professor in the graduate school of education, argues, headlines about declining test scores could just as easily be about increasing test participation.*
why sat and act test scores will never change that much
it’s also worth mentioning that standardized test scores are subject to a process called equating. it’s not exactly a curve, but it’s a process that ensures that a student’s score should be roughly the same no matter when he or she takes the test. and sat and act questions are carefully tested on students to make sure that there won’t be any drastic changes in difficulty between test forms. this means we aren’t likely to ever see test scores increase or decrease by large amounts year to year–this would mean the testmakers aren’t doing their jobs. if students as a whole do become far worse or far better at the test, then scaled scores would be adjusted to reflect this (in other words it would become either easier or harder to get a certain score).
because the sat has a lot of “points” to work with, a shift of a few points in either direction (when we are talking about a total of 2400 points) is not a huge deal. it also makes sense that act scores (which range from 1-36) would remain “stagnant” in recent years, as opposed to declining or increasing, because it would take a lot more to move the average score from, say, a 21 to a 22.
so, ultimately, doomsday reports about sat and act scores are both rather hyperbolic and rather ingenious marketing strategies on the part of the sat and act.
what we should be talking about regarding declining sat and act scores
opponents of standardized testing are often quick to point out that it is income that is often the strongest predictor of test scores. a glance a bit further down in the college board’s annual report reveals this correlation:
| family income | avg. sat score |
|---|---|
| $0 – $20,000 | 1314 |
| $20,000 – $40,000 | 1399 |
| $40,000 – $60,000 | 1458 |
| $60,000 – $80,000 | 1500 |
| $80,000 – $100,000 | 1544 |
| $100,000 – $120,000 | 1581 |
| $120,000 – $140,000 | 1591 |
| $140,000 – $160,000 | 1616 |
| $160,000 – $200,000 | 1636 |
| more than $200,000 | 1720 |
chart data compiled from the college board’s 2015: college-bound seniors total group profile report.
and income is not the only factor with a strong correlation to test scores; there are also distinct differences between ethnicity groups:
| test-takers who self-describe as: | avg. sat score |
|---|---|
| american indian or alaska native | 1423 |
| asian, asian american, or pacific islander | 1654 |
| black or african american | 1277 |
| puerto rican | 1347 |
| mexican or mexican american | 1343 |
| other hispanic, latino, or latin american | 1345 |
| white | 1576 |
chart data compiled from the college board’s 2015: college-bound seniors total group profile report.
in the last decade, students who identify as black or african american, mexican or mexican american, and other hispanic or latino have all seen greater score drops than the average, according to fairtest.org.
so while we really shouldn’t be alarmed at small shifts in sat scores or “stagnant” act scores, we should be alarmed at the discrepancy in scores between different socioeconomic and ethnic groups–and what this means for both our schools and the fairness of these tests.
*“why the dip in sat scores may not be such a bad thing”
leave a reply