Is It Time to Update NAEP?

Common Core classrooms shift away from what the nation’s report card measures

Gloria’s diving scores from a recent competition are represented in the stem-and-leaf plot shown above. In this plot, 3|4 would be read as 3.4. (A sample eighth-grade math question on the NAEP, http:​/​/nces​.ed​.gov​/nationsreportcard​/itmrls​/portal​.asp​?questionlist​=2003-8M10:8)

On the most recent nationwide U.S. math test, known as the National Assessment of Educational Progress or NAEP, an eighth grader might have been asked a question about stem-and-leaf plots (see the sample question in the graphic on this page or here).

Have no idea what those are? Probably the eighth grader didn’t either. It’s an old-fashioned way to depict data, one that isn’t used much anymore. The Common Core standards, used in more than 40 states, intentionally pared down the list of data representations that students should learn to ones that are commonly used in the real world, such as bar charts and time-series graphs, so that students could spend more time mastering them.

But the NAEP, unchanged since 1990, tested eighth graders on stem-and-leaf plots anyway. “It becomes something of an IQ test to figure out what the heck you’re looking at,” said Fran Stancavage at the American Institutes of Research, a nonprofit research organization that has helped develop the NAEP exam for the U.S government. 

So when the 2015 NAEP results came out last month, showing the first declines in math scores in 25 years (a two-point drop in fourth-grade math and a three-point drop in eighth-grade math between 2013 and 2015), Stancavage didn’t think the problem was only that teachers needed more practice and training to teach the new Common Core material effectively.

“The kinds of things we’re seeing, it wouldn’t matter if someone had become an expert implementer,” said Stancavage.  “Some of these things the kids have never been exposed to.”

In other words, last month’s NAEP scores might have gone down even if every teacher in the nation had taught the new Common Core content perfectly.

Stancavage has made a career out of developing and maintaining the NAEP exam and her work is one of the reasons that the NAEP is prized as an unbiased yardstick of educational progress. She recently co-authored an October 2015 study, comparing NAEP content with the Common Core standards, and found several areas of what she calls “mismatch”.

It’s all in the timing

For example, certain geometry concepts, particularly terminology and vocabulary, have been purposely shifted from fourth to fifth grade in Common Core, so that students can spend more time with fractions in fourth grade. Even if students have the knowledge to solve a geometry problem, an unfamiliar word could confuse a student and throw him or her off course. Not surprisingly, results from the geometry section of the fourth-grade test were down five points, more than twice the size of the overall math decline.

Data analysis was another category that saw an outsized drop (four points) in the fourth-grade NAEP results. Stancavage says that many of these concepts aren’t being taught until middle school now, under Common Core.

By eighth grade, students seemed to have learned some geometry, according to the NAEP results, but the data analysis subsection was still five points lower than two years ago. Again Stancavage found that students hadn’t been exposed to a lot of the content yet. Most of these data analysis concepts gets covered later in high school. But some of it, such as the stem-and-leaf plots, doesn’t.

On the flip side, kids are learning a lot of new material under Common Core that NAEP isn’t testing. Stancavage calculates NAEP isn’t testing 42 percent of the math content that a Common Core-educated eighth grader has been taught.

Stancavage acknowledges that NAEP isn’t supposed to be a “carbon copy of Common Core,” but a reflection of American education. Still, she argues that it’s time to revamp NAEP. “If you start getting too far away from what’s going on in classrooms, which appears to be what is happening now, you don’t really have a trend — you’re picking up an artifact of the test not matching the curriculum.” 

“They haven’t changed it since 1990,” she added. “They probably are due. It’s 25 years.”

Stancavage isn’t alone. In her October 2015 study, 15 prominent education experts urged the board that oversees NAEP to consider a revamp. Privately, some people who are involved in administering NAEP tests worry that urban school districts, which voluntarily participate in extra testing designed to give the public more data on their students, might pull out of the NAEP testing program if they feel that the tests are unfair.

The problem is, if you change NAEP, you won’t be able to compare new NAEP scores with old NAEP scores. It would be like comparing apples and oranges, and so we’d never know if our nation’s great Common Core experiment is working, and producing academic gains.

Diane Briars, president of the National Council of Teachers of Mathematics, says we should keep NAEP as is for exactly that reason. She hopes that Common Core will help students become so much better at fractions and open-ended word problems, for example, that it will more than offset the areas on the NAEP exam that students aren’t exposed to. Others argue it’s too soon to update NAEP before we get a chance to tweak Common Core and get the new standards right first.

Stancavage admits it’s a conundrum. One way out would be to administer two tests simultaneously, one that’s updated to reflect what American school children are learning and one that you can compare with past years. But it would be hard to get Washington to pay for that, or to get anyone to agree to more testing.

Even if we could agree to update NAEP, Stancavage says the development process is slow. The soonest we could roll out a new exam would be 2021. Until then, it’s probably going to look like American students aren’t making the progress they should be in math.

[Cross-posted at The Hechinger Report]

Jill Barshay

Jill Barshay is the founding editor and writer of Education By The Numbers, The Hechinger Report's blog about education data.