Monthly Archives: October 2020

Bush/Obama Ed Reform: Core Damage?

History will probably never adequately address the reasons for Common Core’s educational failure to improve results (as opposed to political failure, which I’ve outlined over the past year).

Contemporary analyses can’t even agree that they’ve failed.

  1. We need more implementation time to show the results. (ex: Stay the Course)
  2. The standards were too easy. (ex: Common Core Has Failed. Now What?)
  3. Standards don’t make a difference. (ex: Common Core Has Not Worked)
  4. What are you talking about? Common Core didn’t fail! ex: In California, Common Core Has Not Failed)

Before NAEP failed to show any improvement, everyone pushing Common Core called NAEP the “gold standard” of educational testing. After NAEP scores failed to show any improvement, some Core proponents including, sadly, Michael Kirst, blamed NAEP.

But to me, there’s always been another interesting question: granted that Common Core didn’t improve academic performance, did it do something worse? Did it actively slow or retard student progress?

NAEP Reminder

To begin with, for those who don’t actively pore over NAEP scores, here’s the reformer nightmare:

Click to enlarge. ELA on left, fourth grade on top. These scores go from 1990 to 2019. Notice that the only steep hike is in 4th grade math, next steepest in 8th grade math, and both of the major growth happens not only before Common Core, but before NCLB–that is, long before reformers got their way at the federal level. My take on this: the growing criticism from “A Nation at Risk” on clearly convinced states to take academic achievement more seriously, teaching more content earlier, thus leading to a boost in scores.

That boost in achievement, as anyone who knows NAEP will tell you, never translated to high school:

There are three explanations of why high school NAEP scores remain flat. First, we are holding onto more students, so their ability level is lower, thus pulling down scores.  Next, the population demographics have changed, and the growth rate has been among the races with lower test scores, which prevents the average from increasing. That’s certainly possible, but looking at 17 year old scores by race:

They all seem pretty flat, so surely that would have resulted in a decrease? Eh.

My own pick for why scores stay flat in high school is that the question is reversed. Instead of asking why scores stay flat in high school, ask yourselves why so many more blacks and Hispanics do well in younger grades. I first wrote about this in my seminal article The false god of elementary math scores–well, seminal in the Ed Realist oeuvre, at least:

We should take to heart the Wise Words of Barbie. Math achievement will fall off as the courses get more challenging. Students who excelled at their times tables and easily grasped fractions might still struggle with complex numbers or combinatorics.

Or, as Steve Sailer said once, Usain Bolt wasn’t much faster than any of his age peers–at six months old.

State Tests

I’ve always preferred state test scores over NAEP. Granted, they aren’t standardized over geography or time. But the entire state population is tested by grade on the same assessment.  You’d think that would be a baseline requirement, but in fact NAEP just selects kids at random, allowing the states some selection sculpting, and then tests those kids on a subset of the entire question set.

So what did the state tests show about Common Core?

God Bless Education Week.

Common Core 2015 Test Results

Blue states are SBAC, red states are PARCC, purple states either didn’t adopt Common Core or didn’t adopt the tests, using their own.

Step through every state’s results from 2013-14 to 2014-15 and you’ll notice that all the purple states saw little or no change in test scores. Meanwhile, the red and blue state scores plummeted, with the singular exception of Missouri, whose English scores on the SBAC were an improvement from the year before. Someone should ask Missouri why. All  the rest, all the way down: state-designed tests, no change. PARCC or SBAC, steep drop in proficiency.

Edweek compared overall student populations over a year period. I picked three states: California, Illinois, and Colorado (an SBAC state and two PARCCs) and broke them out by grade and growth/decline over a five year period.

Orange is before; blue is after. So, a massive hit in test scores. And this was the norm, fairly close to universal, for all states that adopted the PARCC or SBAC. Which is why so many states dumped the tests.

They were designed to be more difficult. Education reformers desperately want to expand charters beyond their primary base of low income parents looking for a way out of chaotic schools. They wanted to break into the suburbs and wreak the same kind of budgeting havoc in wealthy school districts as they do in poor ones. The path towards achieving this, they thought, was to “convince suburban parents that their schools sucked, too!” as Michael Petrilli said in a podcast several years ago.

Reformers thought most people approved of their goals. They thought the public had their backs. They thought the public shared in the disapproval of those dumbed-down NCLB era tests. They were wrong.

Why Make The Tests Harder?

The tests could have been made more difficult, the cut scores higher, without making any underlying change in the materials learned. But Common Core standards, at least in math, were much harder.

Why? Because the Common Core developers had their own ideas about the falloff between elementary and high school scores.  They understood, as I’ve pointed out, that elementary school focuses on arithmetic fact and algorithm mastery.  Math curriculum gets dramatically more difficult and more abstract in high school.  Thus, elementary school test scores are always higher than high school scores. It’s easier to achieve mastery of arithmetic and algorithms.

But they didn’t even consider the Wise Words of Barbie.  The Common Core developers, as well as all education reformers and progressives, see ability as irrelevant to policy planning. They see the falloff as a failure of instruction and expectations.

Solving the huge influx of abstract math in high school required,er, flattening the difficulty curve. Teach young children the theory behind arithmetic, rather than just the algorithms and math facts. Introduce conceptual math much earlier into the educational time line. Students would master difficult arithmetic concepts earlier and be ready for the higher difficulty of advanced math.

As far as I can tell, I’m one of the only people who correctly observed this plan seven years ago, in Core Meltdown Coming, (I’m kind of proud of this, given that the Common Core math criticism was that the standards were too easy). You can read some of the details of how they pushed the difficulty down in that article. Or you can just read the thousands of articles delineating the angry woes of elementary school kids throughout the country.

One could easily explain the perceived failure of the past fifty years of educational policy as nothing more than the failure to see ability as highly relevant to educational achievement.

Teaching vs. Learning

So Common Core’s failure to improve academic achievement could be seen as the imposition of ruthless reality: Introducing difficult math concepts earlier didn’t lead to earlier mastery.  

But Stripe wasn’t harmed by the attempt to teach him whistling.

What if teaching more abstraction not only didn’t achieve understanding but also prevented the understanding previously taken for granted? What if kids who’d previously been able to grasp the basics of math facts and algorithms were now struggling with them? If you don’t tell a kid that 3+8=11, but rather continually ask him to prove it, maybe the kid’s own intelligence influences understanding of math facts.  Bright kids will realize that there’s a pattern to the “proofs”, that they are actually just using patterns to reflect reality. Less able kids might  never get around to realizing that math facts are facts, as opposed to opinions they can prove kind of like in writing class, just by finding a cool quote.

Possibly–just possibly–Common Core math standards interfered with algorithm and math fact mastery.

That would explain the falloff in NAEP math performance, although once again NAEP isn’t focused enough to pick up on this in any comprehensive manner. It would also explain why the lowest achievers were the hardest hit.

I have no evidence. Zip, nada. (Although I was just reminded that Spotted Toad wondered the same thing, and that’s a good sign.) It’d be an interesting research project, requiring a deep dive into particular question types. Someone should check it out. My theory has face validity, at least. Intuitively, teachers all understand that teaching students aggressively beyond their capabilities is damaging. It’s why so many of us reject the demand for “higher standards” and often actively support “dumbing down” as a way for children to learn more effectively.

Imagine being an education reformer shilling and then defending Common Core. NAEP scores, which you routinely describe as the “gold standard of education measurement” flatten or drop in fourth and eight grade math and reading in apparent response to a  hugely expensive, howlingly unpopular standards change.  Then it appears that the lowest performers are declining more than high performers, when your argument for Common Core was that higher standards were what weaker kids needed to know what is expected of them.

At that point, you’re left with “It was the implementation!” or “Stay the course!” or “The NAEP is testing the wrong stuff!”

And so, the national standards dream died a horrible death once more.

And so, finally, I have finished all but the final chapter of my Rise and Fall of Bush/Obama Education Reforms. The pandemic derailed my focus, alas, and I probably could have done better on Common Core. Each one of these pieces has something interesting to say, I think, but ideally I would have collapsed them all into two. Sorry about that.