What Comes After the Science of Learning?

Back-to-school is generally one of my favorite times of the year.  Teachers come back feeling—for the most part—refreshed, classrooms are freshly painted, people look tanned and rested, and there’s usually a lot of positive feeling in the air. It also is one of the busiest times of year. My school year generally starts with a series of professional learning sessions centered on district initiatives, along with meeting teachers & leaders, new and returning, to align the final bits and bobs before students return. 

Every year, I count myself lucky to stillbe in classrooms, working with teachers and students across K-12 schools. This year, I was particularly grateful, because something seems to be shifting in education. More educators, states and systems—both in the United States and abroad—are paying attention to a body of research that calls for a more disciplined, evidence-informed approach to instruction, particularly in reading. Education Week recently reported that although states continue to adopt policies requiring evidence-based approaches to reading instruction, national data show that elementary schoolers are still struggling in reading. And while enthusiasm for evidence-informed approaches continues to grow, with similar conversations now emerging around  math instruction, there are still many questions about how to take this research and translate it into improved outcomes for students.

For people like me, much of this is still encouraging. We seem to be moving beyond whether research should inform instruction at all. The harder challenge is figuring out what happens next.

For many teachers, educational research stops one step before the place they actually need it. Dylan Kane, a math teacher whose Substack I enjoy for its practical insights to teaching, makes some of these points in a recent post. He argues that the science of learning is not something to get too hyped about because, in and of itself, it does not sufficiently address the challenges of the classroom. His broader critique echoes questions teachers posed to me during professional learning this year and comes down to something deceptively simple:

How useful is a body of knowledge if it cannot tell us enough about how to reliably produce learning in actual classrooms? 

Kane makes several respectable points in his piece, though perhaps his strongest is his emphasis on a teacher’s content knowledge. The science of learning largely identifies general principles for how students acquire, retain, and retrieve knowledge. But those principles operate within the content teachers are trying to teach; they cannot tell teachers which knowledge matters most, how ideas within a discipline relate to one another, where students are likely to struggle, or how best to represent those ideas instructionally.

In a recent Education Week article, Alexander Kurz shares a similar frustration. He states that the field needs clearer information about how research findings translate into actual classroom instruction.  

There is a clear practitioner gap in taking what science tells us about learning, and what research identifies as promising instructional principles or practices, and turning those into something teachers can reliably enact to teach specific content to students in their classrooms. 

The science of reading movement offers a useful example of the practitioner gap. Research has established the importance of foundational skills, prompting states to require more evidence-based approaches to early reading instruction. But knowing that students need systematic instruction in decoding does not settle the much larger instructional question of how to develop proficient readers. How much instruction? In what sequence? How should foundational skills connect to vocabulary, knowledge, fluency, and comprehension? And what should teachers do when students are proficient on one measure but still struggle to comprehend complex text? 

This is the translational burden Kane is getting at: research can tell us a great deal about learning without fully telling teachers how to turn those findings into successful instruction. And while knowing how students learn may not tell teachers exactly what to do, it can profoundly shape the decisions they make about what to do daily.

For example, if I know that working memory is limited, I am going to think carefully about how much new information I introduce at once, how I sequence explanations, whether students possess the prerequisite knowledge, and when I need to scaffold. If I understand retrieval, I am more likely to deliberately bring previously learned materials back into instruction rather than assuming exposure equals retention. If I understand the role of prior knowledge, I am going to ask what students need to know before they can make sense of the thing I am about to teach. If I understand the novice-expert continuum, I am less likely to throw students into complex tasks without sufficient guidance.

Perhaps most importantly, having the knowledge of how students actually learn makes me a better consumer of my curriculum and discipline. It requires that I regard my resources critically to ensure that they allow for or incorporate the appropriate opportunities to engage my students in the high-impact practices that best help them learn. 

None of this negates the importance of a teacher’s disciplinary knowledge. To do any of those things well, teachers must have a good amount of understanding about the way their discipline unfolds and what specific ingredients students need to access their content. And in most disciplines, it will look very different. While the principles behind the science of learning cannot tell me what content specifically to teach, they absolutely change how intelligently I design instruction around that content.

While I can appreciate Kane’s frustration that the science cannot tell him precisely how to teach a particular concept to these particular students, that’s not really the standard we use for professional knowledge in other fields. Useful knowledge doesn’t have to eliminate professional judgement to improve professional judgement. The purpose of the science of learning is to make teacher expertise better informed. A teacher needs deep content knowledge because she must know the structure of what she is teaching. She needs pedagogical content knowledge because she must know how that particular content is learned, where misconceptions occur, and which representations help. She needs responsiveness because she must interpret what the students in front of her are actually doing. And she also benefits enormously from knowing something about learning itself.

Perhaps most importantly, there are several findings from the science of learning that are not merely interesting descriptions of cognition. They have genuine instructional implications.

Take for instance sequencing. If new learning depends on existing schemas and prerequisite knowledge, sequencing content isn’t a trivial curricular decision. It becomes fundamental to instruction. If working memory is limited, the amount and complexity of information presented at once matters. If retrieval strengthens later accessibility, then previously learned materials must be deliberately brought back rather than taught once and abandoned. If durable learning depends on spaced and deliberate practice, then curriculum cannot be designed solely around getting students through content on schedule. 

So, while the science of learning does not provide a script for teaching, we shouldn’t really expect it to. Its value lies in providing teachers with principles that constrain, inform and improve the thousands of instructional decisions we make within our content every day. It also gives teachers a stronger lens through which to evaluate the ideas routinely presented as best practice—particularly when their preparation has offered limited grounding in how learning actually occurs. Understanding learning makes teachers less susceptible to claims that confuse learner variability with learning preferences, equate choice with motivation, or mistake engagement for learning. 

Still, Kane’s overarching point is important because it speaks to the disconnect many educators feel when they hear this information in a professional learning session. Most want practical advice of how to take this knowledge and then apply it immediately into their classroom. How do we get better at translating research into the actual craft of teaching, without becoming less enthusiastic about what we know about learning? The science of learning is not enough. Neither is content knowledge, pedagogical expertise, or experience in isolation. Good teaching requires all of these skills. 

In writing Teaching for Cognitive Engagement, I found myself returning to this same problem: the research may identify an important principle, but teachers still need to know what that principle looks like in practice. The research-to-practice gap may be precisely where high-quality instructional coaching matters most. Coaches are ideally positioned to help teachers translate general principles about learning into content-specific instructional decisions, while using classroom evidence to determine whether those decisions are actually working. 

But coaching can suffer from the same translational problem. Too often, it is treated as a generic practice: coaches facilitate reflection, support initiatives, or introduce instructional strategies without necessarily having the depth of content, curriculum, and learning-science knowledge needed to help teachers apply those strategies to the specific material they are teaching. And not every school has access to sufficient coaching capacity—or coaches who have been adequately prepared for this kind of work.

A coach who can tell a teacher to “use retrieval” is far less useful than one who can help determine what students need to retrieve, why that knowledge matters for tomorrow’s learning, and how to build that retrieval into the lesson.

The answer, then, is to get much better at building the structures that help teachers translate research into practice. That means investing in deep content and pedagogical knowledge, developing coaches who can connect research to specific curriculum and student needs, and creating systems where effective instructional knowledge is tested, refined, and shared rather than left to individual teachers to discover through trial and error. Kane is right that much of the best practical knowledge still lives in classrooms. Our task should be to make that knowledge easier to develop, easier to examine, and far easier to spread.

References

Amplify. (2026, June). Progress Paused: What end-of-year data reveal about early literacy in grades K-2 [Research brief]. https://amplify.com/pdf/uploads/2026/06/mCLASS_EOY_Report_072426.pdf

Kane, D. (2026 August 11). I’m ambivalent about “the science of learning”: What we know, and what we don’t. Five Twelve Thirteen. https://fivetwelvethirteen.substack.com/p/im-ambivalent-about-the-science-of/comments

Schwartz, S. (2026 August 3). The ‘science of reading’ movement is evolving. Is math next? Education Week. https://www.edweek.org/teaching-learning/the-science-of-reading-movement-is-evolving-is-math-next/2026/08

The Science of Math. (n.d.).  The Science of Math. https://www.thescienceofmath.com

van Merrienboer, J. J. G., Kirschner, P. A., & Kester, L. (2003). Taking the Load Off a Learner’s Mind: Instructional Design for Complex Learning. Educational Psychologist, 38(1), 5–13. https://doi.org/10.1207/S15326985EP3801_2