Why I'm Not Writing About NAEP Reading and Math Scores

The media hoopla over low results distracts attention from the reasons for them.

---

Well, okay, I will write a bit about the depressing national test scores. But the media hoopla only distracts attention from the underlying causes of test score declines—which are about far more than the pandemic.

There’s been plenty of coverage of the reading and math scores from the National Assessment of Educational Progress, or NAEP, that were officially released on October 24. You can read about them in detail via the New York Times, NPR, Axios, USA Today, the Washington Post, and many other outlets.

But I’ll summarize: It’s bad. Scores fell sharply, although the decline was steeper in math than reading, and gaps between high- and low-scorers increased. Whether schools remained physically open during the pandemic didn’t make a difference.

As federal education official Peggy Carr said during the release event, math scores are generally more “sensitive to schooling,” so it makes sense that a period of remote learning—or in-person learning in schools thrown into disarray by the pandemic—would have a greater impact there than in reading. By the same token, though, the math decline may be easier to reverse once schools are back to normal. And if you compare current math scores to those in 1990, when the NAEP in math was first administered, there’s been progress.

But reading scores are right back to where they were in 1992, when the first reading test was given. Plus, unlike math scores, reading scores have been stagnant or declining for over 20 years.

Reading instruction needs to fundamentally change

At the NAEP release event, there was only one hint that something fundamental might need to change, and it was easy to miss. During a discussion of tutoring—the go-to solution for Covid-related learning loss, along with longer school hours—NAEP governing board member Tonya Matthews observed that it may not be enough for a tutor to sit with a child and have them read, because the child might have just memorized the book. Tutors need training to teach reading skills, she suggested.

Dr. Matthews was on the right track, but she could have added that reading teachers need that training. Watching the NAEP event, you’d never know the education world has been roiled for years by revelations that the standard approach to reading conflicts with a mountain of scientific evidence on how children actually learn to read. Prospective teachers are trained to use methods that result in kids only appearing to read because—yes—they’ve memorized the book, or they’re just guessing at words based on the pictures or the context.

Gee, could that have something to do with the fact that two-thirds of students routinely score below proficient on NAEP reading tests? The tests supposedly assess reading comprehension, but a recent study found that the fourth-graders who score lowest actually can’t decipher, or “decode,” the words.

But decoding is only part of the problem. Most of the time spent on reading in elementary school—which averages at least two hours a day—is devoted to comprehension “skills and strategies.” Kids practice the skill-of-the-week on books on random topics that are easy for them to read, because the theory is that if you get good at, say, “finding the main idea,” you’ll be able to apply that skill to any text. That assumption has led elementary and sometimes middle schools to marginalize or eliminate social studies and science, especially where scores are low.

But there’s solid evidence that the key factor in reading comprehension isn’t abstract skill, it’s how much relevant knowledge the reader has—either of the topic, or general academic knowledge and vocabulary. In the current system, kids don’t get a chance to acquire that kind of knowledge because they’re jumping from one topic to the next, without the repetition and deep dives that would enable them to retain information and vocabulary. Tutoring for kids who struggle with comprehension—and most kids who struggle with reading after second or third grade are presumed to have comprehension problems, even if they can’t decode fluently—consists of yet more ineffectual practice with comprehension “skills.”

Another huge factor in reading comprehension is familiarity with complex sentence structure. For many kids, the only time they see complex “grade-level” text is when they take a reading test—and when they get to high school. Children from more highly educated families generally do better on reading tests, largely because they get more exposure to academic knowledge and complex syntax at home.

There’s been much less attention focused on problems with comprehension instruction than on schools’ failure to teach phonics, so you might think it’s understandable that NAEP officials would be unaware of them. Except that the NAEP governing board convened an expert panel back in 2018 that provided basically the same explanation I’ve just offered. Was anyone there listening?

Reading tests shouldn’t be a guide to instruction

I’m not saying we should stop giving the NAEP, or even that we should stop paying attention to the scores (although I have argued that we’d be better off if we eliminated so-called reading comprehension tests, including the NAEP, and gave tests focused on content areas instead). But we do need to recognize that reading tests only tell us that a lot of kids can’t read well. They can’t tell us what to do about that.

Nevertheless, reading tests have routinely been used as a guide to instruction: kids spend hours reading passages on random topics followed by comprehension questions. But that doesn’t equip kids with the knowledge to understand passages on other random topics, including the ones on the tests. And the media circus surrounding the release of NAEP results reinforces the idea that all we need is more of that kind of instruction. It’s as though we’ve been prescribing medicine that is actually toxic—and when the patient fails to improve, or even gets worse, we just call for larger doses.

Someday I’d like to see a NAEP release event—or perhaps I should say another NAEP release event, after the apparently overlooked one in 2018—where someone explains that we’ve been teaching reading in a way that conflicts with science. I’d also like to hear that person say that if states and districts want to boost reading scores—at least in the long run—they should switch to an elementary curriculum that teaches foundational reading skills systematically, is rich in content, and is structured to build academic knowledge in a coherent, logical fashion.

There are now half a dozen such curricula. An increasing number of schools and districts are adopting them—and, ideally, providing training to teachers in how to use them—but they’re still far from the norm.

I confess that I didn’t watch the entire four hours(!) of the NAEP release event. After two hours, I switched to an online presentation by the authors of a recent book called How Teaching Happens: Seminal Works in Teaching and Teacher Effectiveness and What They Mean in Practice, which digests decades of research in cognitive science—the science of how we learn—and conveys it to teachers. It will come as surprising news to nearly all of them, if they read it.

That’s because the ideas teachers absorb during their training, not just about reading but about learning in general, generally conflict with what science has found actually works. That has a lot to do with poor scores in reading and math—not to mention American history, civics, science, writing, and other subjects the NAEP occasionally gets around to assessing. And kids from lower-income families suffer the most.

Wouldn’t it be nice if the media covered that story with the same avidity they devote to parsing the latest NAEP results?

This post originally appeared on Forbes.com.