A New Way To Bring Teaching in Line With Cognitive Science

There are days when I feel we’re making real progress in aligning teaching with cognitive science—particularly, moving away from ineffective skills-focused reading comprehension instruction and towards an approach that centers knowledge-building.

One day earlier this week, for example, I did two remote presentations on literacy instruction and cognitive science that reached hundreds of teachers. After each one I was inundated with expressions of gratitude in the chat, while applause and heart emojis wafted across my screen. Clearly, many teachers are hungry for this information and receptive to the message.

Plus, more and more school districts are adopting effective knowledge-building elementary literacy curricula like Amplify CKLA, Wit & Wisdom, and EL Education. According to the Center for Education Market Dynamics, the combined total reach of those three curricula is now over 600 districts and seven million students. And that figure doesn’t include the other nine knowledge-building curricula currently available. I don’t have comparable figures for ten years ago, but I would guess that virtually no schools were using knowledge-building literacy curricula then, because such curricula didn’t yet exist.

No curriculum is perfect, of course, including those that build knowledge. They may not be implemented effectively—and some curricula that are labeled “knowledge-building” don’t actually do a good job of building knowledge. Still, many teachers around the country have told me that after they switched to a knowledge-building curriculum, their students were far more engaged and started achieving at levels they never would have expected. They frequently tell me that adopting such curricula has changed their lives, and, I hope, those of their students as well.

But some days I’m reminded, usually via email, that many educators out there haven’t gotten the message.

An elementary teacher I’ll call Ms. Smith emailed me because she was distraught at a test-prep lesson she had just observed: Students were given a six-paragraph excerpt from The Diary of Anne Frank and asked to “infer” how Anne felt about her friends—which, Anne confided to her diary in the excerpt, she no longer had.

The kids knew nothing of Anne Frank, including the fact that she was hiding in an attic, or of World War II. The teacher had been told not to provide any background information because if she did, students wouldn’t “learn to use the text evidence.” As a result, some kids “inferred” that Anne must have been autistic.

“This type of nonsense happens daily,” Ms. Smith told me, “but no one will take me seriously when I complain that it feels like a waste of a lesson.”

Another educator, who I’ll call Ms. Anderson, told me she’s well-connected in education circles in her state but has “literally never heard anyone at the state or district levels talk about the importance of building knowledge and improving reading comprehension.” Nor do they talk about the science of learning. People are very aware of the “science of reading,” but to them it means only systematic instruction in foundational reading skills like phonics—and, perhaps, comprehension skill instruction of the type that Ms. Smith witnessed.

Ms. Smith and Ms. Anderson live in different states, and I’ve spoken in both of them about the importance of building knowledge, multiple times, including at statewide conferences. Apparently that’s not enough.1

I’m not ready to declare, as one longtime advocate for knowledge has, that knowledge-building curriculum can “never win.” It is winning in many places. But I do think we need to figure out if there’s more we can do to promote widespread change.

My writing and speaking, like professional development in general, have been focused on giving educators information in the hope that it will change their beliefs. The assumption is that a change in beliefs will lead to a change in practice. That’s the way it works for some teachers, but clearly not for all.

In fact, the work of Thomas Guskey and other researchers suggests that we’ve gotten it backwards. The evidence shows that change actually starts with changes in teacher practice—plus, crucially, evidence showing teachers that those changes are resulting in improved learning. It’s the experience of successful implementation of a new approach, Guskey says, that eventually changes teachers’ beliefs. “They believe it works,” he writes, “because they have seen it work.”

And there’s the rub: If a knowledge-building curriculum is implemented well, which is in itself a big “if,” teachers should see that it “works.” But how will they know it’s working? You might assume they would just look to whether students are learning the concepts and vocabulary that have been taught. And some certainly do.

But because of their training and the institutional pressures to which they’re subject, many teachers look instead to scores on standardized reading comprehension tests. Those include not just the high-stakes tests at the end of the school year but also the “benchmark” tests used throughout the year to monitor progress. On those tests, the reading passages have nothing to do with the content students have been taught.

Those scores may not show improvement for several years, because that’s how long it can take for students to acquire the critical mass of knowledge and vocabulary that enables them to understand texts on topics they don’t already know about. So even if students are acquiring knowledge that will eventually improve their general reading comprehension, teachers may be getting feedback from tests telling them their students aren’t making progress—and that what they really need is more practice in, say, the “skill” of making inferences. That feedback can lead teachers back to their old ineffective approach.

How can we ensure teachers get more accurate feedback about whether knowledge-building is working? One way—maybe the only broadly effective way—is to change the nature of state tests.

Testing is powerful. What gets tested, it’s frequently said, gets taught. Recent commentators say one big reason for declining student achievement is that we’ve loosened up on holding schools accountable for test scores. But while that may be true for math, it’s almost certainly not the case for reading comprehension.

Share

Standardized reading tests don’t just mislead teachers about student progress. They also distort curriculum and instruction in ways that hold students back. Most schools devote two or more hours a day to reading or “English language arts,” much of which is spent on instruction that mimics the tests: Read a random excerpt, then answer comprehension questions. As the Anne Frank incident suggests, that approach doesn’t improve comprehension.

But maybe it’s possible to use testing as a lever for bringing instruction more in line with cognitive science. I have a couple of ideas.

First, instead of putting the overwhelming emphasis on reading, along with math, states could start testing social studies regularly. Ironically, that might end up doing more to improve students’ reading ability than the current regime of requiring reading tests.

A study by the Thomas B. Fordham Institute found that more time on social studies in elementary school is associated with improved reading scores by fifth grade, whereas more time on reading is not. Most likely, that’s because the kids who get more social studies are acquiring more of the knowledge and vocabulary that enables them to understand the passages on the reading tests.

You might think this data would lead schools that want to improve their reading scores to spend more time on social studies. But the study has been out there for six years now, and it doesn’t seem to have had that effect. Social studies, particularly in elementary school, is basically an afterthought.

Teachers are far less likely to be provided with a curriculum for social studies than for other subjects, meaning they have to cobble together materials themselves. They’re also less likely to get professional development or evaluation in that area.

A major reason, according to a RAND report, is that only “a handful of states” require tests in social studies that are factored into their accountability formulas. The obvious solution is to add tests in social studies, ideally in third through eighth grades and once in high school, which is what states currently require for reading and math.

Of course, teachers—and parents and students—are unlikely to be enthusiastic about more testing. In my ideal world, I would solve that problem by simply eliminating the reading comprehension tests and replacing them with social studies tests.

Students should of course read literature, and especially whole novels. But the emphasis on “English language arts” has distorted our educational priorities. Is it really more important for fourth-graders to learn about character traits and metaphors than about history and geography? Wouldn’t our democracy be better off if more kids knew which country we fought against in the American Revolution, or who won the Civil War, or that there’s such a place as Washington, D.C.?

As the country approaches its 250^th birthday, there’s been a lot of hand-wringing about how little students know about history and civics, but little effort to do anything meaningful about the problem.

Realistically, though, we’re probably stuck with reading tests for the foreseeable future. So why not ground the passages on state reading tests in the social studies topics listed in state academic standards?

Those standards may not be great—and we certainly need better social studies curricula, especially at the elementary level—but at least they specify some content, whereas reading comprehension standards list only comprehension skills.2 And the Fordham study found positive results from additional social studies instruction based on the social studies standards we currently have, so apparently they’re good enough to improve reading ability.

My other testing innovation would be to place more emphasis on writing quality. Again, what gets tested gets taught, or at least gets attention. When Louisiana gave more weight to the writing section of its state literacy tests, districts began to pay more attention to writing instruction—which, like social studies, is a generally neglected area. Fortunately, most districts in the state were already using a content-rich curriculum, which gave students the raw material they needed to engage in writing. Like most curricula, though, it didn’t actually teach them to write.

Some districts discovered a method of explicit writing instruction called The Writing Revolution, which they could adapt to their curriculum, not just in ELA but across the board. [Note: I’m the co-author of a book describing the method and serve on the advisory board of The Writing Revolution organization.] When I visited one of those districts—Monroe City, a high-poverty area—I was told that the result was not only that students became far better able to write but also better able to learn.

“At the start,” high school teacher Justin Overacker told me, “the focus was on writing instruction. But something kind of unexpected, kind of amazing began to happen. We realized that teaching students to write clearly was actually teaching them to think clearly.”

Why? Probably because when teachers embed explicit writing instruction in rich content, they’re providing students with many of the benefits of cognitive-science informed instruction—notably, retrieval practice and elaboration, increasing their access to information stored in long-term memory and deepening their comprehension of whatever they’re writing about.

What I saw and heard in Monroe is a prime example of the process Thomas Guskey describes: First teachers’ practice changed, with training provided directly by The Writing Revolution organization to ensure successful implementation. Then they saw evidence it was working, both in their students’ writing and on state tests.

I haven’t done a study of whether teachers’ beliefs eventually changed, but I suspect many came to intuitively understand the power of science-informed instruction, whether or not they would put it in those terms. In any event, the process didn’t start with a change in beliefs. Teachers in Monroe told me they’d never heard of concepts like retrieval practice and cognitive load theory before changing their approach to writing instruction. They just wanted to help their students write.

Ideally, a state testing regime wouldn’t try to evaluate student writing in the abstract—which doesn’t work any better than trying to evaluate reading comprehension in the abstract. Writing ability is inextricably bound up with knowledge of the topic you’re writing about. So the writing prompts on state tests should be grounded in material students have learned about—like, for example, the content specified in state social studies standards.3

I’ll continue to write and speak about building knowledge and concepts like working memory as long as people want me to—or until I just can’t do it anymore. Even if I’m not reaching every educator, I’m reaching many of them, and that’s important.

But I’ll also hope that some state education officials out there are bold enough to experiment with state tests as I’ve suggested. I can’t guarantee it will work, but changing state testing regimes is likely to be our best chance for changing teachers’ practice in ways that enable all students to reach their full potential—and maybe, ultimately, changing teachers’ beliefs as well.