What Do We Mean by "Reading Comprehension Instruction"?

And is there too little of it in schools or too much?

---

---

A recent study found, to the researchers’ dismay, that on average, teachers spend less than a quarter of their instructional time on reading comprehension instruction. Yet I and others have been arguing that schools spend too much time on comprehension instruction.

Who’s right? The answer may depend on which grade levels and subjects you look at—and how you define comprehension instruction.

But first, some background. Back in the late 1970s, a researcher named Dolores Durkin observed third- through sixth-grade classrooms to see how teachers were teaching reading comprehension. She found that, according to her definition, they basically weren’t.

Yes, teachers would ask questions about what a text meant and correct children’s answers if they were wrong. But that, said Durkin, wasn’t teaching comprehension. Rather, it was assessing it.

Durkin’s paper arrived at around the same time cognitive psychologists started looking into how reading comprehension works, whether it can be taught—and if so, how. Since then, hundreds of studies of reading comprehension instruction have been published that show positive effects from certain practices, and the federal government has released multiple guides for teachers on how to align their instruction with the data.

Surely, comprehension instruction must have improved since Durkin’s time—or so a team of researchers assumed. They recently undertook a systematic review of the literature, analyzing 66 studies done between 1980 and 2023. The data covered observations of over 1,700 teachers in K-12 classrooms.

Lead researcher Philip Capin and his colleagues did find more instruction than Durkin had. One reason is that they used a broader definition, including practices like defining words. Durkin believed that only instruction aimed beyond the word level counted as comprehension instruction.

Still, they found that teachers across grade levels approached comprehension mostly through “low-level questions … in which the answers were largely derived directly from details in the text.” Rarely were students asked questions that required them to justify their answers with textual evidence or make inferences within and across texts.

A basic problem, the researchers said, was that students weren’t spending much time reading. Especially in science and social studies, Capin told Jill Barshay of the Hechinger Report, teachers were building students’ knowledge, but not through reading texts.

Another study, led by Christy S. Murray in 2021—also discussed by Barshay—confirmed that finding. Surveying 124 middle school teachers, the researchers found that 70 percent of science teachers and about half of social studies teachers spend less than six minutes a day reading text in class.

Most teachers surveyed wanted to spend more time on reading but said the texts they were supposed to use were too difficult for their students. Echoing Capin’s observation—and Durkin’s—the survey found that science and social studies teachers emphasize “content delivery rather than learning content by reading text.” They tend to rely on methods like PowerPoint presentations, films, and handouts with notes.

Too much or too little focus on text comprehension?

As I mentioned, I and others have been arguing that teachers spend too much time trying to teach comprehension skills and strategies and too little time building the kind of knowledge and vocabulary students need to understand complex text. Are we wrong?

Actually, if you look closely, Capin’s findings support many of the points we’ve been making. He found, for example, that elementary teachers spend more time than others teaching comprehension—or trying to. And, indeed, the knowledge movement has identified elementary school as the place where teachers spend way too much time on comprehension skills. Meanwhile, social studies and science—subjects that have the most potential to build the knowledge and vocabulary that fuel reading comprehension—are marginalized.

Capin also told Barshay that teachers tend to ask “generic comprehension questions, such as ‘What is the main point?’ without considering whether the questions are appropriate for the reading passage at hand.” That too aligns with what advocates of knowledge-building have observed. Often there’s a skill of the week—maybe “finding the main idea” of a text, or “making inferences.” Teachers try to use a text to teach the skill rather than focusing on a text or topic and bringing in whatever skills or strategies are likely to help students understand that particular content.

It also makes sense that Capin saw more comprehension instruction in studies done after 2000. That’s when standardized reading tests started to become the barometer of all progress under the federal No Child Left Behind law. Those tests purport to measure reading comprehension skills, using passages on random topics. Under pressure to raise scores, schools patterned their instruction after the tests.

That approach hasn’t actually raised scores. Comprehension skills and strategies work only if readers have enough background knowledge to understand a passage at least at a superficial level. If they don’t, they never get a chance to demonstrate their skill at making inferences, or whatever the test appears to be measuring. Just think about what it’s like to read a text about, say, molecular biology if you don’t know anything about the subject. Can you “find the main idea”?

It certainly makes sense to us knowledge advocates that, as Murray found, teachers beyond elementary school encounter students who struggle to read grade-level text. There’s plenty of other evidence for that.

One reason is that texts at higher grade levels assume students have background knowledge relating to history or science that many never had the opportunity to acquire—because their elementary schools spent so much time on comprehension skills. Another is the widespread system of leveled reading, which has students practice skills on texts at their supposed individual reading level. That approach prevents them from developing familiarity with the complex vocabulary and sentence structure they’ll encounter in texts later on.

Knowledge Is Important but Not Enough

A couple of additional points have come up in commentary about the Capin study that need to be addressed. In his interview with Barshay, Capin remarked that “some advocates of knowledge-building criticized his analysis, arguing that knowledge-building alone is beneficial for reading comprehension and it doesn’t matter if the teacher uses slides or actual texts.”

That’s not exactly what this advocate of knowledge-building would argue. Slides or videos can build knowledge of a new topic, and that knowledge does help with comprehension. But instruction shouldn’t end there.

Written language is more complex than oral language in terms of its vocabulary and sentence structure, and students need to become familiar with that complexity if they’re going to become proficient readers. So teachers can begin by relaying information in simple terms, but they also need to read aloud from complex texts and guide students to understand them—and then, of course, have students read the texts themselves. The ultimate goal is to enable students to read complex text independently.

Reading researcher Tim Shanahan has also written a couple of posts about the Capin study (see here and here). His position is that if instruction improves students’ ability to understand the text at hand but doesn’t equip them to understand other texts, it’s not comprehension instruction.

Christopher Such has already taken on Shanahan’s argument in a Twitter thread, and I’m not sure I can improve on his response. Briefly, Such argues that Shanahan’s distinction is a false dichotomy that could lead teachers to downplay building students’ “broader grasp of written language and the world it describes.”

“It is essentially impossible to teach pupils to better understand a specific text without also developing their potential to understand other texts,” Such writes.

Shanahan also argues that there’s insufficient evidence to show that building students’ knowledge improves their general reading comprehension. He concludes that “at this point, building knowledge, per se, should not count as reading comprehension instruction.”

I’ve already addressed that argument elsewhere (see, for example, here and here) and won’t go into detail now. Suffice it to say that there is evidence that building knowledge boosts comprehension, although much of it is indirect. That’s because it can take years to see the effects of building knowledge on standardized tests, and experiments that last years are hugely expensive and generally difficult to conduct. So we don’t have many of them, and we’re unlikely to get many more.

That, however, doesn’t justify perpetuating a demonstrably ineffective approach to teaching comprehension. Yes, there is evidence that some kinds of strategy instruction boost comprehension, but that evidence doesn’t support the kind of comprehension instruction typically seen in classrooms. Even in the studies, kids get all the benefits of skills-focused instruction very quickly—and it’s not clear how long those benefits last.

Changing Our Concept of Reading Comprehension

Where does all this leave us? Students need to be able to understand what they read, and clearly, many don’t, especially as the grade levels go up. It’s also clear that teachers are generally ill-equipped by their training and materials to deal with the problem. A RAND study last year found that nearly three-quarters of teachers in grades three to eight say they need more resources to identify and support struggling readers.

I’d say the solution starts with changing our concept of reading comprehension. Rather than seeing it as a set of discrete skills that can be taught independently of content—and only during the elementary “reading block”—we need to start seeing it as part of teaching. Just as it’s impossible to separate helping students understand one text from helping them understand other texts, it’s impossible to separate teaching comprehension from teaching content. Teachers need to guide students to comprehend whatever texts they’re expected to read, across grade levels and across the curriculum.

That means—yes—building knowledge. And it means checking to see if kids have literally understood what you think you’ve taught or what they’ve read. How can you address comprehension if you don’t know what students have or have not comprehended?

But it also means building fluency—the ability to read with appropriate expression at an appropriate pace—which has been called “the bridge to comprehension.” Many students at higher grade levels are hampered by a lack of fluency, expending so much cognitive effort on deciphering and making sense of individual words that they don’t have the bandwidth to understand the text as a whole.

And it means teaching students about things like prefixes and suffixes and word origins, knowledge that becomes increasingly important to comprehension as texts become more complex. It means bringing in strategies and skills as appropriate, including those that require students to engage in analysis, generalization, and inferential thinking. It should also include teaching students to write about what they’re learning, across subjects and grades.

As I’ve discussed elsewhere, writing—if taught explicitly and manageably—is a powerful way of deepening and reinforcing the knowledge that enables reading comprehension. It can also deepen comprehension by having students ask and answer “why” and “how” questions, something cognitive scientists call elaborative interrogation. And it can familiarize students with the complex syntax of written language by teaching them how to use that kind of syntax in their own writing. Yet none of the discussions of reading comprehension I’ve referred to in this post even mentions writing.

Teachers Need Better Curriculum

Incorporating all these things in instruction may sound like a tall order for teachers, especially those who haven’t gotten good training in teaching reading or writing—which is the case for most teachers, especially those at higher grade levels who teach subjects other than English language arts. Better teacher prep would help, but it will take a long time for that to come about, and it would still leave out the millions of teachers already on the job. In the meantime, we risk condemning yet more students to functional illiteracy.

What we need is better curriculum—across the board, not just in English language arts—with on-the-job professional development grounded in whatever specifics the curriculum covers. That curriculum should incorporate principles backed by cognitive science that teachers rarely learn about during their training.

The answer, then, to the question “Too much or too little reading comprehension instruction?” is: “Both.” There’s too much instruction that treats comprehension as a transferable set of narrow skills to be taught solely in elementary English language arts. And there’s too little of the kind that has teachers guiding students to understand complex texts across grades and subjects, bringing in whatever is needed to achieve that goal. If we want students to become proficient readers, we need to recognize that every literacy teacher also needs to be a content teacher, and vice versa.