There has been much discussion about the sharp decline of SAT scores. Some (including its sponsor, the College Board) attribute the decline to an increase in the number of test-takers. Others say that the decline can be attributed by the increased diversity of the test-takers, meaning that when more low-scoring students take the tests, the scores go down.
Carol Burris took the time to review the data and come up with a data-driven discussion of what really happened.
The bottom line, she writes, is that the score decline was large and significant:
SAT scores for the Class of 2015 were the lowest since the test was revised and re-normed in 2005. The score drop in one year was 7 points — a drop that Inside Higher Ed characterized as significant.
She says there was very minimal increase in the number of students taking the tests.
Between 2014 and 2015:
11 states saw an increase in the proportion of seniors who took the SAT.
3 states remained exactly the same.
36 states saw decreases in the percentages of members of the Class of 2015 taking the test when compared with 2014.
What about the assumption that the increase in fee waiver students is responsible for the decline? It implies that a greater proportion of test takers come from low-income households. The increase in fee waivers does not necessarily mean, however, that the percentage of low-income test takers has increased. It could be attributed to more students being encouraged to apply for the waiver. In fact, the percentage of low-income students who are test-takers has been remarkably stable.
The diversity of the test-takers is relatively stable:
Here are the percentages of test takers with family incomes below $20,000 during the past five years: 2011 — 13 percent, 2012 — 14 percent, 2013 — 14 percent, 2014 — 13 percent, 2015 — 14 percent. That same stability runs across all bands of income.
And what of the College Board’s claim that this is “the most diverse group ever”? Time Magazine implied that the increase in fee waiver students and increases in diversity are the cause of the decline.
How much has diversity increased? Since 2011 the percentage of Black or African American students taking the test has been a steady 13 percent every year. There have been small proportional increases in 2 of the 3 categories that describe students who are Hispanic or Latino, but there are also small proportional increases in Asian students and international students whose test scores exceed the average by large amounts. Asian students’ average scores this year were a whooping 164 points above the total average. They are hardly dragging scores down.
And in this year of the big drop, the proportions of Black, Latino/Hispanic and Asian test takers are exactly the same as they were in 2014.
What does the decline mean? What does the College Board advise the schools to do?
Burris writes:
Nearly every article on the topic included the same quote from the chief of assessment of the College Board, Cyndie Schmeiser:
“Simply doing the same things we have been doing is not going to improve these numbers. This is a call to action to do something different to propel more students to readiness.”
Well, riddle me this one: Does Ms. Schmeiser talk to her boss? College Board chief David Coleman certainly created “something different” back in 2010. And given that the Class of 2015 had five years of exposure to his Common Core State Standards (of which he was the co-author of the English Language Standards), as well as spending their entire school career in the era of NCLB accountability, it doesn’t look like “something different” is working very well.
Of course, his new solution is to make next year’s newly designed SATs align with the Common Core. Expect ACT registration, which is already on the rise, to increase.
Reformers like Coleman are now the status quo, and the evidence of the effectiveness of their strategies have yet to appear. And if the past four years of SATs are a measure, then their reforms are having a negative effect on scores.
Reflecting on the dropping SAT scores, corporate reform super-fan, Mike Petrilli, of the Thomas B. Fordham Institute asked this question in The Washington Post, “Why is education reform hitting a wall in high school?”
This former 15-year high school principal can answer your question, Mr. Petrilli. Education reform isn’t hitting a wall. It is the wall.

Who would have thought that amateur educators, reformers and admins would actually lower test scores. What say you?
LikeLike
“This former 15-year high school principal can answer your question, Mr. Petrilli. Education reform isn’t hitting a wall. It is the wall.”
Nice!
LikeLike
All important information from Carol. Still, isn’t the main point here, as it has been always, that this test is a high-stakes, normed test, which means no matter what anyone does, no matter how much kids thrive and learn, or fail to thrive and learn, the spread of scores will remain as it is? That’s what a normed test is all about–comparing what kids know to what other kids know. Period. And to make this comparison the College Board uses income distribution as a proxy for “spread,” so that SAT scores follow income distribution. This distribution is fixed and pre-determined, and has nothing to do with real learning or human development (in fact it has become a serious impediment to both). Carol’s genius should not be wasted on this junk.
What has any of this got to do with reading the Iliad, or learning about why American history cannot be done properly without studying slavery (and why Black History Month is not separate from American History Year, but an integral part of it), or what a great two-page expository essay looks like, or why the Moon’s rotation about its axis occurs in same time period as the Moon’s revolution around planet Earth? What has any of this stupidity got to do with providing kids with the time and guidance to learn how to play a clarinet or the drums?
The amount of time and focus we waste on testing mania (and related nonsense) has become, in itself, an important impediment to the education of young people. It is a waste of time, and a serious one at that. It makes all of us stupider. (Reformers’ kids are not subject to it!!!!!)
Waste the time of the kids of the 99% while feeding the minds and souls of the kids of the 1%. Now that’s a national educational policy that works!
LikeLiked by 1 person
Steve, you are so right. The efforts to adapt to the Common Core ARE an impediment to real education. I call the novel and untested curricula it spawns “mutant” curriculua. Tortured analyses of random texts! It bears no resemblance to what previous generations of wise men and women considered good education. To adapt Tacitus: they created a wasteland and called it educational utopia.
LikeLike
WaPo has already posted this article so see my comments on there. Carol is wrong about the percentages of minorities as Hispanics increased by a full % point of the test takers and their scores dropped by 2 pts. Explains most everything.
LikeLike
Assuming you are correct, this fact explains nothing about what really matters. Absolutely nothing.
LikeLike
I’m with you, Virginia. Adults put too much weight on test scores.
It’s fun to watch the ed reformers rush out with the suddenly-important “context” when they get a set of numbers they don’t like, though- especially as I’m reading a whole bunch of astro-turfed ed reform op-ed pieces that claim the Common Core test scores are The Truth. Why are those tests The Truth? Because they serve the political objectives of ed reformers to claim 60% of existing public schools are failing?
LikeLike
So the entire shift can be explained by a tiny change in the percentages? I find this an interesting comment from someone who supports the use of growth algorithms that are shaky at best.
I guess that’s the fun thing about statistics. They can be used to prove anything anyone wants them to prove.
I’m also going to add my usual caveat here. SGPs are piece of the puzzle but not the whole puzzle. Qualitative measures matter a ton in teaching as well. Stats are usefull but easily manipulated.
LikeLike
That the 2 points lower scores of an additional 1% of the test takers coupled with the lowered 2 points of perhaps 13% of test takers comes nowhere near explaining a 7 point drop on average for 100% of the test takers. 1% + 13% = 14%. Multiply that by 2 points and one gets 28 as a drop factor. Whereas 7 points x 100% = 700 as the total drop factor. I’m sorry, Virginia, 28 is nowhere near 700. Now maybe if Hispanics scored 50 points lower on average, you could make the case that it was “their fault.”
LikeLike
I think readers should read the comments left by SGP on the Washington Post blog. Comment after comment after comment. A continual rant with every blog I write. He knows Diane would not allow those personal attacks here. It is very sad. And I will never dignify them with a response.
LikeLike
It is sad and sickening, and shouldn’t be allowed on any serious blog.
LikeLike
I agree. Now I can figure out why his initial stands for “Supernova Gaijin Planet.” Even his former boss in the Navy won’t hire a guy with that mentality.
LikeLike
Any objective measure of learning that can’t be fixed by the corrupt school districts not only shows no improvement but rather a decline in student performance.
LikeLike
Leonard,
“Any objective measure of learning. . . ”
Please define what an “objective measure of learning” is.
TIA,
Duane
LikeLike
Is there any kind of disinterested analysis of these claimed “remediation rates”? I ask because many of the community college students here believe that colleges are shunting them into remediation mostly as a revenue-enhancer. They say it enough to make me think it’s not whining, although of course it could be whining. Frankly it is difficult for me to believe that the current crop of 18 year olds are hugely more stupid than when I was 18 and I went to a community college and I’m middle aged. I think extraordinary claims require more than an assertion by an interested party.
I know Carol Burris once questioned Duncan’s reciting these stats. Is there ever any examination of the remediation stats that ed reform seems to rely on for 90% of their press releases? If I’m wrong and these numbers are verified as valid in some fashion other than self-reporting by colleges that’s fine- I’m asking.
LikeLike
Chiara, I am not aware of any reliable data on what percentage of students actually need remediation. However, I assume it’s largely a result of more students attending college.
If 1/3 of students have traditionally attended college, it’s fair to say most of the other 2/3 would need remediation prior to admission. As every politician pushes more college attendance, more of those 2/3 are entering and many need remediation. Thus, the traditional college bound students are not dumber, again this is a mix issue.
Furthermore, college doesn’t make sense for the new attendees. When college diplomas were scarce, they were valued. With each additional diploma, its value drops including for the traditional graduate. But making masters worse, these borderline college students can’t hope to ever receive a huge return since they were not strong students to begin with. Subsidies entice them into massive loans along with the promises of politicians and admission officers. They lose on both ends.
We should be developing other sectors via apprenticeship programs like Germany has. There is a reason Germany makes most of the high precision tooling equipment in the world. But if you say not everybody should go to college, you are instantly shot for being an elitist hater.
LikeLike
“We should be developing other sectors via apprenticeship programs like Germany has. There is a reason Germany makes most of the high precision tooling equipment in the world. But if you say not everybody should go to college, you are instantly shot for being an elitist hater.”
You know, Virginia I hear this constantly about how we should be using apprenticeship programs “like Germany” but Germany makes a determined effort to retain skilled trades and a higher-paid workforce. It’s a pro-working people country, with a strong labor influence.
Germany didn’t destroy the skilled trades pipeline in search of ever-cheaper labor like the the US did. If we wanted more skilled trades maybe we shouldn’t have tried to exterminate labor unions and devalued work. Actions have consequences. The US decided obscene executive compensation and shareholder gains were more important than valuing and investing in w workforce.
Oh, well. Greed and short term thinking has consequences. One of them is a lack of trained workers.
LikeLike
THANKS, CHIARA, Germany has a policy of protecting jobs and unions. It is the most powerful nation in Europe because it protects its middle class. In the latest OECD report, the US has a far higher proportion of college graduates than Germany, disproving the canard that our economy depends on sending more people to college than any other nation. We can learn from Germany. We should rebuild our middle class and the unions that help the middle class grow.
LikeLike
Germany never has had a Taft Hartley law that debars all with a supervisory title from joining a union. In Germany only the very highest executives are not in the union, thus there is solidarity rather than division among members of the work force. Repealing Taft Hartley used to be a standard part of the Democratic party platform — until it was quietly dropped. A big mistake.
LikeLike
Diane, did you just say that an economy does NOT depend on sending an ever increasing percentage of its citizens to college? I think we are in violent agreement on that point. Why don’t you make that a more substantive part of your campaign? It would eliminate a lot of student debt in the process.
LikeLike
Virginia, please read my book “Reign of Error.” I point out the relative standing of the U.S. in producing college graduates and show that Germany, the engine of Europe, has far fewer college graduates than we do. Germany has protected unions and its industrial base. It has not willy-nilly outsourced jobs to low-wage countries.
LikeLike
Decision makers, listen up!
LikeLike
“College Board. Propelling students to readiness since 1900.”
LikeLike
Sounds like this Carol Burris lady is playing dirty pool. After all , using a data driven discussion to beat the reformers at their own game. Oh the humanity!
LikeLike
By their own standards they have failed, they should be forced out of power and conspicuously ignored. Their own petard will spear them.
LikeLike
Interesting take and I agree in spirit. But for so long, many (including myself) have criticized any use of the SAT as a reliable measure of educational quality in the U.S. I’m not sure that using it now for that purpose isn’t disingenuous. Thoughts?
LikeLike
Concur!
LikeLike
Ditto
LikeLike
After looking at the data – is it Common Sense to conclude Common Core (and NCLB) is the cause of this downward trend?
Of course, these results will be used as an excuse to double down on the current educational trends instead of switching course and putting things “back to the way it was” prior to all this interference from non-educators (ones who insist Piaget is passé).
LikeLike
Both Colemaning and Petrillium engineering programs are sell-out.
LikeLike
Reblogged this on Network Schools – Wayne Gersen and commented:
This SHOULD compel those “reformers” who advocate evidence based decision-making to re-think the test and punish model they’ve championed for over a decade. But based on the pending federal legislation it appears that testing will continue…
LikeLike