Do We Need Another National Reading Panel?

Or might it do more harm than good?

---

Back in 2000, a panel of reading experts convened by the federal government issued a report that reviewed the research on how children learn to read. The findings of the National Reading Panel helped shift reading instruction away from ineffective techniques that involve having students guess at words and toward systematic instruction in phonics. No doubt many children have benefited as a result.

Now there’s a move afoot in Congress, spearheaded by Rep. Josh Harder (D-Calif), to convene another NRP. After all, advocates say, there’s been quite a bit of research related to reading in the 25 years or so since the last panel issued its report.

Harder has managed to get language that would fund a new NRP into the House budget bill for 2027—which, it should be noted, has not yet been enacted into law.¹

The bill calls for the new NRP to address some issues that the earlier panel didn’t take up. Those include “the relationship between reading books rather than passages in building reading stamina; the relationship between writing instruction and reading comprehension; optimal amounts of time for different aspects of reading and writing instruction in the average classroom; [and] the essential components of effective reading curricula.”

The bill also mentions an examination of the relationship between teacher preparation, professional development, and effective reading and writing instruction, along with “out-of-school factors” like oral language.

These issues are worthy of attention, and a new report could address some misconceptions that have proliferated in recent years. Literacy instruction has come to center far too much on excerpts or brief passages; some students graduate from high school without ever having read an entire book. And it’s welcome news that writing is mentioned in the bill. The first NRP said nothing about writing, despite evidence that writing about texts can boost reading comprehension.

The “optimal amounts of time” issue could help prevent schools from spending way more time on phonics than the 20 or 30 minutes experts say most kids need. Some schools reportedly spend as much as two hours a day on phonics, leaving little or no time for things like read-alouds—or actual reading by students.

Inevitably, however, there are some important components of reading that didn’t make it into the bill—for example, familiarity with complex sentence structure, which recent research indicates is more important to comprehension than vocabulary, especially at higher grade levels. And the bill’s passing mention of research regarding “early reading development and instruction” suggests that the drafters see reading instruction as limited to the early grades—when in fact, students at higher grade levels often desperately need support in decoding more complex vocabulary and sentence structure.

No Mention of the Role of Knowledge

Of course, the panel wouldn’t be limited to looking at the factors listed in the bill. But a crucial factor I worry a new panel would overlook is the same one that was overlooked by the original NRP: the role of background knowledge in reading comprehension.

The original NRP report vastly oversimplified the process of comprehension by focusing on evidence related to teaching reading comprehension strategies like “asking and answering questions” or summarizing.² One reason, apparently, was that the panel prioritized experimental and quasi-experimental studies that met the gold standard of a randomized controlled trial—studies in which there is a “treatment” group and a comparable “control” group. Such studies allow researchers to infer that the “treatment,” or intervention, actually caused whatever changes they observe.

The panel’s decision to focus on such quantitative studies is understandable—when it comes to phonics. The “whole language” advocates of the 1980s and 1990s who opposed systematic phonics instruction often relied on qualitative, observational studies of reading instruction.

The problem with such studies is that it can look like kids are learning to read even if they’re not. They might, for example, have just memorized the simple words they’re being asked to read rather than having learned the phonics patterns that would enable them to transfer their decoding skills to words they don’t already know. To ensure the reliability of its conclusions, the NRP wanted to exclude those observational studies.

But the panel applied the same criteria to studies of reading comprehension, and in that area such a standard can be misleading. There are hundreds of experimental studies showing that teaching certain comprehension strategies can improve reading scores. But there are other factors—like background knowledge—that are actually more important to comprehension but far more difficult to study directly.

Strategy Instruction Is Easier to Study

One reason there are so many studies of strategy instruction is that researchers can see the results on standardized tests very quickly—even after just a few hours of instruction. Most research studies last no more than six weeks, so if you’re looking to publish positive results, strategy instruction is a highly attractive subject to focus on.

We also have studies demonstrating the role of knowledge in comprehension, but most are limited to showing that it’s helpful to have knowledge of the topic you’re reading about. If you’re a baseball expert, for example, it’s clear that your comprehension of an article on baseball will be better than that of someone who knows little or nothing about baseball—even, perhaps, if they’re a better all-around reader.

But some reading experts discount those studies because they don’t measure the effects of background knowledge on general reading comprehension. For that, they say, you need to use standardized tests that rely on passages unrelated to any particular topic. Those tests purport to measure comprehension skill or ability in the abstract.

But it’s virtually impossible to do that. Background knowledge and vocabulary almost inevitably affect comprehension, especially as grade levels go up and reading passages become more complex. As cognitive psychologist Daniel Willingham has said, standardized reading comprehension tests are actually “knowledge tests in disguise.”

We do have a handful of studies showing that building children’s knowledge of lots of topics, in a logical fashion, improves their general reading comprehension. But those studies also indicate that it can take three years or more to see the results. Studies that last that long are difficult to conduct and extremely expensive.

So it’s not surprising that there are more studies of strategy instruction than of knowledge-building. That doesn’t mean strategy instruction is more effective than knowledge-building; it just means it’s easier and less expensive to study. In fact, without a minimum level of relevant background knowledge, no comprehension skill or strategy is likely to work.

The Risks of a Second NRP

The original NRP’s failure to mention knowledge as a factor in comprehension helped entrench a pre-existing belief among many educators that teaching comprehension skills and strategies was all you needed to do to enable students to understand text. We’ve been paying the price ever since—or rather, students who are less likely to acquire academic knowledge outside of school have been paying that price.

It’s not clear what a second NRP would conclude about the role of background knowledge. One possibility is that it would cast its net farther than the original panel in terms of data—including not only the multi-year studies showing that building knowledge can boost standardized test scores but also the ample evidence on the importance of knowledge from the field of cognitive science.

This wouldn’t necessitate relying on relatively subjective observational studies. There are many experimental studies showing that having prior relevant knowledge stored in long-term memory expands the capacity of working memory to understand and absorb new information.

The other possibility, though, would be that the panel would again limit its focus primarily to experimental studies conducted by reading researchers. If a new report simply weighs the total number of studies of strategy instruction against the total number of studies of knowledge-building, without considering other data and factors, it could declare that the evidence shows that strategy instruction is more effective than knowledge-building. That might be an even worse outcome than the first panel’s failure to mention knowledge-building at all.

There are reasons to believe that might happen. Not only does the NRP bill fail to mention the role of knowledge in comprehension, its sponsor, Josh Harder, seems unaware of the issue. Harder was recently a guest on a podcast produced by the Thomas B. Fordham Institute, an education think tank. One of the hosts, Fordham president Michael Petrilli, set Harder up to address the knowledge issue, saying this:

[The] National Reading Panel did not say much about the importance of knowledge building. This was something that even back then, E.D. Hirsch had been writing about and arguing about for quite some time, and marshalling a lot of evidence that reading comprehension, once you do learn the foundational skills, is largely about learning history and science and other kinds of content and vocabulary.

But that was not a part of the National Reading Panel report. And you could argue maybe that’s why it took us so long to finally get to the point, really just very recently, that we start to see more people talking about the importance of that.

So I guess the question is, what if a new National Reading Panel gets it wrong? Isn’t there some risk of mischief here?

Rather than discussing knowledge, Harder talked about the importance of gathering evidence on implementation. “The biggest problem that we have on reading right now,” he said, “is that almost every state has passed a science of reading bill, but very few states are actually seeing sustained gains in student outcomes.”

It’s undoubtedly true that one reason most states haven’t seen sustained gains is that implementation is often flawed, But another likely reason is that virtually none of those states require that schools build the knowledge that kids need to understand the increasingly complex text they encounter as the grade levels go up.

It’s also worrying that Harder’s bill specifies that the panel is to include “leading scientists in reading research,” along with education school faculty, reading teachers, and school administrators. Without representatives from the broader community of cognitive scientists—and, ideally speech and language pathologists, who have been more aware of the role of complex syntax in comprehension—we’re likely to get another panel that views reading comprehension in isolation from learning in general.

Of course, it’s possible that the make-up of the panel would go beyond what the bill calls for—or that the “reading researchers” appointed to the panel will take a broader view of reading research than their predecessors. Even if that’s the case, there’s still the risk that the report will be misinterpreted. If, for example, it provides a list of factors backed by research, educators may see that as a checklist rather than items to be integrated into instruction—which is what often happened with the first NRP.

It’s common these days to attribute all the problems in our education system to problems with math and reading instruction—and, when it comes to reading, to problems with instruction in reading individual words. But those problems are just the tip of a very large iceberg. If we want to improve education outcomes, we need to connect reading and writing instruction to the “content areas” like history and science—and bring instruction in all those areas in line with evidence from cognitive science.

I’m not sure another federally convened panel will have much effect—or at least, not a positive one. But if we’re going to have a panel, I would prefer one focused on learning, and not just on “reading.”

1. The NRP provision is not to be confused with the READ Act, sponsored by Sen. Bill Cassidy (R-La), which would restrict federal literacy funding to “evidence-based approaches” and also contains a controversial definition of dyslexia.

2. The NRP did mention the role of vocabulary in comprehension but didn’t connect that to knowledge, when in fact the two are largely inseparable. It didn’t help that the popular “five pillars of early literacy,” derived from the NRP report, lists “vocabulary” and “comprehension” separately.