Why So Many Kids Struggle to Learn

When Eric Kalenze was getting his master’s degree in education in the 1990s, he was immersed in pedagogical theories that have prevailed at ed schools for a century. Learning proceeds best, he was told, when focused on skills like critical thinking and tailored to the interests of individuals. Rather than assuming the role of a “sage on the stage,” depositing facts into children’s passive brains, a teacher should be a “guide on the side,” enabling students to learn primarily through inquiry and hands-on activities. Kalenze was dubious. None of this jibed with what he recalled of his own school experience or his gut-level sense of what works.

But when he became a high school English teacher, he figured he would give it a try. When he covered The Great Gatsby, he didn’t explicitly teach his students about the symbolism of the green light at the end of Daisy’s dock. Instead, he had them spend two or three days looking up meanings of the word green, finding magazine ads using that color, cutting them out, and making collages. He would ask questions like, “Why do you think they used green for this product that’s a lotion?” and hope the response would be something like, “Because it makes you feel fresh and youthful.”

Kalenze’s students loved him, and other teachers observed his classes because they’d heard the buzz. But after a few years, writing assignments and class discussion showed that his apparently engaged students weren’t grasping Gatsby’s significance. He also discovered books by E. D. Hirsch Jr. and Diane Ravitch that reinforced his initial feeling that what was called “progressive” or “constructivist” pedagogy didn’t place enough value on building students’ knowledge through explicit instruction. Kalenze realized that his students didn’t know enough about the 1920s to appreciate why Gatsby was considered the Great American Novel and not just another tale of unrequited love.

So he stopped having his students make collages and started supplying some of that history. He told them that the era was a time of liberation for many, especially women, but also a time of growing economic inequality—and “then this book comes out that basically says all that amid a love story.” That background provided “an anchor,” he says, for subsequent class discussion, and his students’ understanding of the book “changed wildly.”

“No one ever told me to do that,” he says. “They just told me to have the kids read for enjoyment, and make it as engaging as possible.”

Kalenze’s experience lines up with what scientists have discovered in recent decades: that acquiring factual information isn’t a useless, soul-crushing exercise; it’s the prerequisite for higher-order thinking. Asking students who don’t know much about a topic to learn through inquiry or “discovery” is inefficient at best. Projects and hands-on activities often waste precious time. Engagement is crucial, but it’s quite possible for students to be highly engaged without learning anything important.

The foundations of modern cognitive science go back 70 years, with many of its advances stemming from research in artificial intelligence and information technology. Application of the principles to teaching methods is a more recent development; a landmark 2006 article by three cognitive psychologists—Paul A. Kirschner, John Sweller, and Richard E. Clark—used a title that attempted to cover the range of approaches favored by educators: “Why Minimal Guidance During Instruction Does Not Work: An Analysis of the Failure of Constructivist, Discovery, Problem-Based, Experiential, and Inquiry-Based Teaching.” “The past half-century of empirical research on this issue,” the authors wrote, “has provided overwhelming and unambiguous evidence that minimal guidance during instruction is significantly less effective and efficient than guidance specifically designed to support the cognitive processing necessary for learning.”

John Sweller had come up with a useful way of looking at the learning process in the late 1980s. Called “cognitive load theory,” it rests on a well-established concept that cognitive scientists call working memory: the aspect of consciousness that takes in new information about the world and tries to make sense of it. The essential fact about working memory is that it can hold only a few items—maybe five, or seven, or nine—for just 20 or 30 seconds. If your working memory is trying to juggle too many things at once—if the “cognitive load” is too great—you may not be able to fully comprehend, analyze, or retain information. The best way to get around constraints on working memory is to store lots of information in long-term memory, which has a capacity that is theoretically unlimited. If you can withdraw items from long-term memory when you need them, you don’t have to use space in working memory to make sense of them.

One practical implication of cognitive load theory is that you can’t teach students critical thinking in the abstract; you can think critically about a topic only if you have ample information about it stored in long-term memory. It’s also important to provide students with “retrieval practice”: having them recall information from long-term memory to make it more readily available, perhaps by giving frequent quizzes. Students should have some freedom of choice, but if the curriculum is guided by their individual interests, they likely won’t acquire the information needed to gain new knowledge later on.

Cognitive load theory also helps explain why projects and inquiry-based learning are generally ineffective. Given the limitations of working memory, students tend to remember whatever their attention is focused on—which could just be cutting out ads with the color green. And asking learners to search for a solution to a problem involving an unfamiliar topic places an unnecessary burden on working memory that leaves little space for productive mental effort.

Education researcher Dylan Wiliam has called cognitive load theory “the single most important thing for teachers to know.” And yet it’s unlikely that prospective American teachers will encounter it or almost any theory grounded in cognitive science. Take, for example, reading instruction, which has come to dominate the elementary school day. The term reading is used to encompass two processes: reading or “decoding” words, and comprehension. And the standard instructional approach to each process conflicts with a mountain of scientific evidence.

Many children, perhaps most, need systematic instruction in phonics and other foundational reading skills to become fluent decoders. And yet, teacher-prep programs either encourage candidates to develop their own philosophy of teaching reading or endorse the now-standard “balanced literacy” approach, which embraces phonics in theory but retains a skepticism that has deep roots in the education world. Former elementary teacher Lori Sappington remembers that in her undergraduate courses, she learned only the definitions of terms like phonics. “I don’t recall ever learning here’s how you might teach this.”

She’s not alone. In a 2019 survey conducted by the publication Education Week, kindergarten through second-grade teachers and those trained in special education were asked where they had learned most of what they knew about reading instruction. Only five percent cited their preservice training. A Facebook group created in August 2019 called “Science of Reading—What I Should Have Learned in College” attracted more than 100,000 members in less than two years.

The scientific evidence on reading comprehension isn’t as voluminous as the data on phonics, but it’s well established that relevant knowledge and vocabulary are central to success. That evidence accords with cognitive load theory: if readers don’t have sufficient background knowledge of a topic, their working memory quickly becomes overwhelmed. There haven’t been many studies of whether building kids’ academic knowledge will boost reading comprehension, as measured by standardized tests, but the database is growing, and it’s promising.

Nevertheless, in ed schools—and K–12 classrooms—reading comprehension is typically reduced to a menu of “skills and strategies,” like “finding the main idea” or “making text-to-self connections,” to be practiced by students on a random variety of books they can read easily on their own. To make time for such practice, many elementary schools have marginalized or eliminated knowledge-building subjects like social studies and science. The pressure to raise reading and math scores is one factor, but even more fundamental is the notion that it just isn’t important to have facts about the world stored in long-term memory. In one ed school class I observed on elementary literacy, the instructor invoked a common trope in dismissing the idea of quizzing students on dates and names because those “are things that kids can now access at the touch of a phone.” But looking things up can place a heavy burden on working memory.

One result of the disjunction between teacher training and science is that teachers’ jobs are much harder than they should be. Another is that students who come from highly educated and usually affluent families are more likely to thrive, since they generally have greater access to the academic knowledge or support missing at school. Others—who are disproportionately low income and minority—are left to flounder. They lag behind in reading, despite years spent practicing comprehension “skills,” which are arguably not true skills since they can’t be taught directly or applied generally. They also often arrive at upper grades without the background knowledge suddenly assumed by the curriculum. It’s not uncommon for students at high-poverty high schools to lack basic information like the difference between a city and a state or a country and a continent, or what the American Revolution was. The problem is not that they can’t learn; it’s that no one has deemed it important to give them access to the knowledge for which they’re now being held accountable.

The Common Core State Standards in literacy, released in 2010, were intended to address this problem. They introduced more complicated text and more nonfiction, including in the elementary grades where reading had long been limited to simple stories. Although supplemental materials urge schools to adopt elementary curricula that systematically build children’s knowledge, most educators and policymakers have focused only on the standards, which read like a list of skills—albeit some that are new, like “know and use various text features (e.g., headings, tables of contents, glossaries, electronic menus, icons) to locate key facts or information in a text.” Most “Common Core–aligned” reading curricula have also maintained a focus on skills. But combining skills-focused comprehension practice with nonfiction, which often assumes a good deal of background knowledge, can make instruction even less effective. Research on whether the Common Core improved overall education outcomes or narrowed score gaps has been inconclusive.

On national reading tests, about two-thirds of students score below “proficient,” as do about four-fifths of those at the lower end of the socioeconomic spectrum; more than 40 percent of 12th-graders whose parents didn’t finish high school score below “basic.” Scores have declined in recent years, and the gap between high scorers and low scorers has widened. The tests don’t distinguish between decoding problems and lack of background knowledge, but many students suffer from both and are hopelessly behind. As a novice teacher 15 years ago, Ashley Johnson was asked to evaluate the reading levels of 150 ninth-graders at a high-poverty school in Washington, D.C. She found that the average student read at a third-grade level. Eventually she cofounded an organization called Literacy Lab, which provides tutoring in foundational reading skills for students from age three to third grade.

It’s not just students from less-educated families who lack academic knowledge. On the most recent administration of a test on U.S. history given to a representative sample of American eighth-graders, only 15 percent scored proficient or above; about a quarter cleared that bar in civics. That’s a problem for all of us. If American citizens don’t have the skills or background knowledge to understand a newspaper, they may not be able to vote responsibly—and many may not vote at all.

To be sure, schools of education aren’t solely responsible for this situation. And a grounding in the principles of cognitive science wouldn’t be enough to ensure effective teaching. But shouldn’t teachers be equipped with this knowledge rather than get information contradicting it? How and why has teacher training diverged so markedly from science? And can anything be done to bring the two together?

Many observers trace the problems with teacher-prep programs to their 19th-century origins as “normal schools”: institutions that admitted mostly female students with an eighth-grade education and trained them to deliver the basics of reading, writing, geography, and arithmetic. In the 20th century, those schools became state colleges or were absorbed by state universities, but admissions standards have remained relatively low, as have the criteria for professors. The rest of academia has looked down on schools of education, partly because of those lower standards and partly because of a common assumption that the teaching profession—unlike law or medicine—doesn’t require any special preparation or knowledge. Education professors have been understandably resentful, and there has been little communication between them and, say, professors of psychology on the same campus. The degree of divergence is reflected in the old quip that 120th Street in Manhattan—which separates Columbia Teachers College from the rest of the university—is “the widest street in the world.”

Internally, several forces have shaped the ed school curriculum. One was a reaction against the didactic approach of 19th-century teachers, who were perceived as emphasizing harsh discipline and rote memorization—a reaction that continues to shape teacher training. The early-20th-century philosopher John Dewey and his acolytes developed the still standard “child-centered” approach in opposition to those methods. Another factor has been the pressure to establish academic credibility. Education courses are criticized for being too theoretical, but to gain tenure, instructors need to prioritize research, which often centers on theory. For all their deficiencies, normal schools were at least single-mindedly focused on practice, sometimes requiring candidates to try out techniques in model schools where instructors critiqued their performance.

Ed school students generally take some courses in content outside the education program. Those who plan to go into secondary teaching often major in the subject they’re expecting to teach, and many prospective teachers enroll in master’s programs after getting an undergraduate degree in a subject like English or history. Elementary teacher candidates, however, typically get undergraduate degrees in education, which may leave little time for other courses. They’re required to take some arts and sciences courses before applying to an education program but receive little guidance on which courses will cover the content they’ll be expected to teach—and the absence of a standard K–12 curriculum makes that hard to predict. More fundamentally, candidates are themselves products of a defective K–12 system that can leave significant gaps in their own knowledge. Fewer than half of aspiring elementary teachers pass teacher licensure tests on their first try.

All candidates for education degrees also take “foundations” courses in ed school, in subjects like developmental psychology, along with courses in teaching methods. They may get exposure to aspects of cognitive psychology in a foundations course, but what they learn in a methods course may diverge from or contradict it—or the course might focus on how to read aloud expressively or decorate classroom bulletin boards. According to a 2005 paper by two education professors, methods courses aren’t intended to transmit information about effective methods of instruction but rather to function as “complex sites in which instructors work simultaneously with prospective teachers on beliefs, teaching practices and creation of identities—their students’ and their own.”

Even if the goal were to impart best classroom practices, that would be difficult when the content K–12 teachers are expected to cover is vague.  States have academic standards that set out in broad terms what students should know and be able to do at specific grade levels, and in theory districts adopt detailed curricula aligned to the standards. But ed schools generally train candidates to “teach the standards,” since it’s unclear what the curriculum will be—and because many teacher-educators are suspicious of any curriculum, believing that teachers should be free to teach whatever their students want or need. Reading standards rarely specify any content, focusing instead on comprehension skills. But even when content is defined, as with social studies standards, it’s often fragmented or so broad as to be almost meaningless.

In one elementary social studies methods class I observed, the assignment was to pick a standard, do a class presentation on a children’s book connected to that standard, and create activities that could be used to teach the standard, along with an “artifact.” One candidate chose a fifth-grade standard that covered significant events of the 20th century, including “the world wars.” In a fairly typical presentation, she described a book about three elephants that were starved to death at a Tokyo zoo during World War II as a precaution against the zoo being bombed and the animals escaping. Neither the presentation nor the book mentioned any specifics about World War II, and students were to do a project on how to prevent what happened to the elephants from ever happening again. The candidate’s artifact was a stuffed green elephant.

In addition to coursework, there’s student teaching, often referred to as clinical experience or fieldwork. In theory, this is a crucial aspect of teacher prep in which students can apply what they’ve learned and apprentice under a master teacher. In practice, candidates in university-based programs may be placed with teachers who are struggling and need help—or who model instruction that departs from what ed school faculty have taught is best practice—and instructors exercise little effective supervision or control. Alternative programs, like Teach for America, often supervise fieldwork more closely and may have different course requirements—although few if any have integrated cognitive science into their curricula.

Ed school faculty and administrators may agree in principle that their programs could be improved, but reform efforts have had little discernible effect. According to Edward Crowe, who heads TPI-US, an organization that inspects teacher prep programs, “the culture is like Jell-O: it’s infinitely adaptable to the larger environment.” (That’s true, he adds, of higher education generally.)

If outsiders try to initiate reform, long-standing resentments can make communication difficult. But even efforts by insiders can encounter resistance. In the 1980s, leaders of some preeminent American ed schools mounted an initiative to address perceived deficiencies in their programs. Their first report noted that “a flurry of concern swirled about their discussions, as if the assembled leaders were plotting revolution.” Although that effort and others touched on improving the ed school curriculum and making it more rigorous, a prime goal has been forging a stronger link between the theories that candidates learn in their courses and what they do in their clinical practice. Few have questioned the validity of the theories themselves.

That has, however, been a focus of the most high-profile ed school reform initiative of recent years, the National Council on Teacher Quality (NCTQ). Founded in 2000, NCTQ periodically releases reports that analyze various aspects of teacher prep. In a report published in 2016, it found that textbooks used in educational psychology and methods courses barely mentioned findings supported by evidence from cognitive science. But the organization is best known for grading programs on the narrower and more contentious issue of early reading.

NCTQ’s first report on that subject, in 2013, gave only 35 percent of programs an A or B, with many receiving failing grades. Ed school faculty and administrators were furious, charging that NCTQ was biased—it did favor phonics, but the “bias” was grounded in scientific evidence—and that its methodology was flawed. Rather than conducting site visits, the organization reviews documents like syllabi and textbooks. Kate Walsh, who heads NCTQ, says it isn’t feasible to visit the approximately 2,000 providers of teacher prep in the United States, let alone the 27,000 or so individual programs they encompass. Nor, given the resistance even to providing documents—some of which NCTQ has obtained by suing under open records laws—is it likely that schools would allow classroom observations. Still, critics have called NCTQ’s data inaccurate and argued that rating a program based on syllabi is like judging a restaurant by its menu.

Walsh denies the accusations of inaccuracy and maintains that syllabi are detailed enough to provide a good basis for review. But she acknowledges early mistakes in communication and says that programs are now allowed to review their scores and provide additional documentation before reports are published. NCTQ has also found that programs are improving: in its 2020 report, the number deemed to provide sufficient coverage of early literacy was up to 51 percent, crossing the halfway mark for the first time. Still, that leaves 49 percent that aren’t meeting that bar—and Walsh says NCTQ’s 2020 criteria were too generous and will be tightened in the future. Moreover, a lot of hostility toward the organization still exists on ed school campuses. Walsh herself estimates that about a third of the programs think NCTQ “might have something useful to say,” a third are unsure, and the rest “think that we are the evil empire.”

In 2015, another organization began aiming for a more collaborative approach. Called Deans for Impact (DfI), it’s a consortium of ed school leaders seeking to improve their programs. The organization’s interest in cognitive science was apparent from the beginning: shortly after it was formed, it released a document called The Science of Learning. One of its authors was University of Virginia cognitive psychologist Daniel T. Willingham, who translates academic findings into terms understandable by educators and the general public.

The impetus for the focus on cognitive science came from DfI’s founder and executive director, Benjamin Riley, who says he was looking for something that could apply to a broad range of programs. “No matter where you are, no matter what context you’re preparing teachers to teach in, they’re all teaching kids with minds,” he says. “And we have some insight into how those minds work.”

After some experimentation, DfI launched its first Learning by Scientific Design network in the fall of 2019, with six schools that committed to embracing one or more of six “focus principles”—for example, “deepening meaning and learning” and “practicing with purpose.” As an initial step, the group administered an assessment to about 1,000 candidates and 22 teacher-educators. Some of the results, a report said, were “sobering.” Candidates often mistook engagement for learning, and only six percent correctly identified the following statement as false: “Any kind of repeated exposure to information makes it more likely the information will be moved into long-term memory.” Teacher-educators who took the test did better, especially on knowing how to apply the principles in the classroom. But on questions about the basic principles themselves, the overall average was only 60 percent. Two further rounds of tests, in 2020 and 2021, showed that teaching about cognitive science improved results to varying degrees, with more hours of instruction generally leading to better scores.

The Learning by Scientific Design effort—which includes helping programs set specific goals and plans and providing training, coaching, and site visits—is supported by philanthropy, including some big names like the Bezos Family Foundation and the Chan Zuckerberg Initiative, each of which has donated more than a million dollars. Among education funders, the initiative headed by Mark Zuckerberg and his wife, Priscilla Chan, has shown the most interest in cognitive science. It has also funded an organization that trains teachers in cognitive neuroscience, a sister discipline to cognitive psychology that focuses on the brain rather than the mind but arrives at essentially the same conclusions. But philanthropic interest in cognitive science has been minimal compared with other education initiatives, many of which are in tension with scientific findings. Both the Chan Zuckerberg Initiative and the Bill & Melinda Gates Foundation have directed sizable funding to “personalized learning”—learning tailored to each student’s needs and interests, often delivered by technology. That approach has been viewed with skepticism by many in the cognitive science camp, including DfI founder Riley.

Although the schools incur no financial obligation by joining the Learning by Scientific Design network, they contribute an enormous amount in “sweat equity,” Riley says. I spoke to administrators and faculty members at three of the institutions, and their comments were uniformly positive. Rather than telling faculty what to do, they said, DfI staff members begin by asking which of the principles a school is interested in addressing and what they can do to help. “We have taken the approach,” Riley says, “that this work will happen only if we find ways to align it to the interests and desires and needs of the faculty and other folks that we are asking to do it.”

Perhaps the most enthusiastic faculty participant I encountered was Hilary Dack, a professor at the ed school at the University of North Carolina–Charlotte who teaches prospective middle school teachers (and is a former middle school teacher herself). Contrary to what DfI generally advises, Dack decided to teach all of the learning science principles in one year—the only one in the network, she thinks, to do that. She acknowledges that was ambitious but, she says, “I couldn’t not teach all of it. I liked it, and I thought it was important. I had been teaching the ‘what,’ but my students needed the ‘why.’ ” The scientific evidence, and in-class activities she devised that brought the principles home, also helped students understand what Dack calls “the equity lens”—the recognition that certain standard pedagogical approaches will inevitably leave some students behind.

UNC-C undergrad Arrinna Poessnecker has taken several courses with Dack and feels passionate about the cognitive science she has learned. In contrast to Kalenze’s experience 25 years before, Poessnecker found that Dack’s classes gave her “the terminology and research to back up my ‘feelings’ on how a classroom should handle assignments and feedback,” as she wrote in a class submission. In one assignment, she created a video comparing the effectiveness of activities for teaching the parts and functions of a cell. After beginning with a diagram showing the interaction between attention, working memory, and long-term memory, the video displays an image commonly found on the Internet: a round cake meticulously decorated with candies labeled nucleolus, mitochondrion, and so on. “This would be an example of a task that draws a student’s attention away from the objectives entirely,” Poessnecker says in the video, “even though it started with good intentions.” Far more effective, she concludes, would be an activity that requires students to recall the names of the parts of a cell and their functions and draw them—a form of retrieval practice that helps embed information in long-term memory.

“A problem in education,” says TPI-US chief executive Edward Crowe, “is that there’s no agreement on what science is—or even what evidence is.” That’s especially apparent in reading instruction, where faculty members may ask, when confronted with the evidence supporting phonics, “Are you talking about your science or my science?” Crowe says he has encountered graduate students who were creationists preparing to be science teachers at one top-rated school of education. “The word theory in education doesn’t mean a testable hypothesis,” he says. “It means an opinion. In education, whatever you personally believe is science is science.”

Crowe says schools probably won’t change unless there’s some impetus from within. Still, external pressure from government entities can help. In the early 2000s, the federal government tried to encourage the teaching of phonics in elementary schools—an attempt that foundered for complex reasons. Now, though, around 20 states require teacher licensure tests that measure elementary candidates’ knowledge of the “science of reading.” According to NCTQ, 32 require elementary prep program standards to address its components.

Mississippi has led the way, largely thanks to a nonprofit there called the Barksdale Reading Institute. Its chief executive officer, Kelly Butler, says that although the state has made progress, there’s still a long way to go. A recent effort to require ed school reading faculty to pass a test in foundational reading skills stalled after ed school deans objected. Nevertheless, Barksdale has launched an initiative with six of the more than 20 states that have expressed interest in learning from Mississippi’s example, in hopes of developing a national model for change.

The cognitive science movement hasn’t yet gained enough traction in the United States to spark government action on either the federal or the state level, but the situation is different in England, where the teaching profession has been influenced by the same pedagogical theories—and the national government has more direct control over education. The Conservative government has enthusiastically embraced cognitive science principles, even codifying them in a framework for teacher training promulgated in 2019 that is part of a broader effort to shake up the field. Prospective teachers are to learn, among other things, that “committing some key facts to their long-term memory is likely to help pupils learn more complex ideas” and that they should “reduce distractions that take attention away from what is being taught.” The citations in the document are a litany of the biggest names in cognitive load and related theories, including John Sweller and Daniel Willingham. DfI’s Science of Learning document is also cited.

It’s unclear at this point what effect these government efforts will have on teacher training. Based on her experience in Mississippi, Butler says it’s important to provide faculty members with a “safe space to acknowledge what they do not know.” After all, their own training was most likely not grounded in the science of reading or cognitive science. It’s also essential, she says, to hold teacher-prep programs accountable for the performance of students taught by their graduates, as measured by test scores. That has been a touchy issue, with faculty complaining that test scores are influenced by factors over which they have no control. School policies, for example, could mandate practices that violate the principles of cognitive science. This, Butler says, is why reform efforts have to simultaneously address what’s going on in K–12 schools as well as teacher-training programs.

In recent decades, teacher-educators have drawn a dichotomy between the “cognitive” model, or framework, and the “sociocultural” one, rejecting the former and embracing the latter. In ed school usage, “cognitive” refers not to the principles advanced by DfI, which are premised on relatively recent research, but to any theory of learning that emphasizes cognition. The sociocultural model focuses on the learner’s interaction with teachers and others and the influence of culture—specifically cultures that have been historically marginalized.

In theory, the models could be combined; no cognitive scientist would deny that interaction and culture play a role in learning, and presumably most educators acknowledge that cognition has some importance. Beyond that, both models see prior knowledge as central to learning. But the sociocultural model prioritizes the knowledge students acquire from their own communities, whereas the cognitive framework emphasizes knowledge relating to whatever students are expected to learn next. Disagreements over what content to include in the curriculum have complicated and politicized the issue. Joseph Cambone, dean of the school of education at Salem State University in Massachusetts, told me that though he believes that the two models should be used together, relying on cognitive science “in the absence of deep sociocultural grounding may perpetuate racist practices.”

That perspective has made it harder than ever to bridge a basic difference in orientation: scientists tend to look for what all learners have in common—a mind, as Benjamin Riley points out—and teacher-educators emphasize that which makes students unique. But again, both things can be true: each student is unique and all students have things in common. And, as Kate Walsh argued on a podcast called Melissa and Lori Love Literacy, ed schools could raise their status in the rest of academia by embracing cognitive science “because they would be teaching very real, evidence-based work. But instead, the prevailing wisdom in many ed schools is, ‘I get to teach what I believe, and it’s called academic freedom.’ I would argue it’s academic freedom that’s being abused.”

That’s an argument echoed, more gently, by Stacy Edmonson, dean of the ed school at Sam Houston State University, which is part of DfI’s Learning by Scientific Design network. If instructors refuse to go along with requirements about what to cover in their courses, she says, “we tell them, this is not about academic freedom. We say this is a program focus, and if you’re not with the program, we can help you find another position.”

At a school with a strong leader like Edmonson and enthusiastic instructors like Hilary Dack, the incorporation of cognitive science may grow organically, as Riley hopes: Dack says a colleague who teaches courses for prospective high school teachers has asked to use her materials. Even so, novice teachers are often influenced more by the prevailing ethos in the K–12 schools where they’re placed or hired than by anything they’ve learned during their training. And as Kalenze points out, the number of new teachers in any given year represents only a small fraction of the workforce.

To be sure, it’s crucial to reach educators who are already in the field, and some efforts appear successful. There’s a growing list of popular books on cognitive science along with websites, podcasts, and a massive open online course called Learning How to Learn that has drawn nearly three million enrollments. There’s also an international movement of conferences called researchED, designed to bring teachers together with scientists, and Kalenze—now the author of two books on education—is its U.S. “ambassador.” But given demands on teachers’ time, perhaps the most promising development is the emergence in the past few years of elementary literacy curricula that focus on building academic knowledge rather than supposed comprehension skills—a development sparked by the Common Core. The number of schools using these new curricula is small but increasing. The pandemic has strained teachers’ resources and ratcheted up anxiety about the widening achievement gap, but that pressure could lead in either of two directions. Many districts are doubling down on reading comprehension skills practice in a fruitless effort to make up for “learning loss.” Others, however, might take advantage of an unprecedented infusion of around $190 billion in federal education aid to switch to a knowledge-building curriculum.

Ed schools remain a huge piece of the puzzle. Even if they can’t prepare candidates to teach specific content or a particular curriculum, they can at least equip them to understand the need to teach content and how to judge the merits of various curricula based on scientific evidence. And there are signs, at least in the area of early reading, that things are beginning to change “one funeral at a time,” as is sometimes said of academia. In the 2019 Ed Week survey, explicit, systematic phonics instruction was favored by only eight percent of instructors who had been in higher education for more than 20 years; for their presumably younger counterparts with fewer than six years’ experience, the percentage was four times as large.

And prospective teachers like Arrinna Poessnecker may exert some influence on the schools they land in, rather than vice versa. I asked her what she would do if she found herself in a school where the prevailing approach didn’t line up with cognitive science. “I think my personal philosophy will include bringing that up with the principal and saying, ‘This is how I’m going to teach,’ ” she said. “And if it’s a problem, I will either be allowed to stay and prove myself or I will happily find a place that actually wants its students to succeed. That sounds aggressive, but at the end of the day, that’s the goal.”