Dunglish checked by ChatGPT

Dan Willingham wrote the Foreword of my/our new book Evidence Informed Teaching Practice scheduled to come out in January 2027. In it he quoted John Dewey’s 1929 book The Sources of a Science of Education about the complexity of education. I was intrigued as I knew Dewey primarily from his University [of] Chicago Laboratory School and had never even heard of this book (my bad!). I managed to find a photocopy of the book online borrowed from a public library in Kansas City, Missouri in 1950 (!) and began to read it. I then bought a reprint of the book; actually 2, one for me and one for one of my dearest colleagues Tim Surma. Here’s my take on the book.

Library lending record

The research says…

You’ve heard it often: “the research says...” It’s usually followed by something a teacher, a school, or even an entire school system should do. A study finds that an instructional approach produces better results than another, a meta-analysis reports a respectable average effect size for something, or a laboratory experiment demonstrates a reliable feature of memory, and before long the finding has travelled through a surprisingly short pipeline from single result to classroom prescription. The research has shown X; therefore teachers should do X. Add an infographic, a professional-development session, and the words “evidence based”, and voila, the transformation (and sometimes the business model) is complete.

This is understandable. Teachers want to know how to teach well, school leaders have to make responsible decisions, and researchers quite reasonably think and/or hope that their work matters. The trouble begins when we confuse the use of science in education with the direct application of individual research results to education. Those aren’t the same thing. In fact, treating them as if they are may be one of the least scientific things we can do with research.

Dewey saw the problem nearly a century ago. In The Sources of a Science of Education (1929) (notice he speaks of education and not learning), he asked what it would actually mean for education to become scientific. He didn’t define science by the presence of experiments, statistics, precise measurements, or laboratories. For him science’s defining feature lies in systematic inquiry: methods for investigating phenomena in ways that allow us to understand them more intelligently, less haphazardly, and with less dependence on routine, tradition, gut-feelings, and guesswork. That in itself is a useful corrective to much of what we hear about the Science of Education, because it shifts our attention away from whether a particular claim has a scientific pedigree towards the larger intellectual system in which that claim really earns its creds.

Science is cumulative and coherent

For Dewey, the decisive characteristic of science is that inquiry (i.e., questions and research) accumulates. One investigator doesn’t simply produce a result and hand it over to practice. Other investigators can and should repeat the work, confirm it, weaken it, or even discredit it, refine the method, discover the boundary conditions for its working, and/or reveal that what appeared to be a general effect was really the consequence of something much more specific or even something else due to confounding. Research creates new questions as often as it answers old ones and methods improve as investigators learn where earlier methods failed or misled them. Scientific knowledge grows as a connected structure – an organism - rather than as a heap of individually impressive studies – a clump of cells.

This is why a research paper or a shelf full of them isn’t science if the findings in them stay conceptually disconnected. Dewey is explicit on this point: no genuine science is formed by isolated conclusions, however exact the measurements or impeccable the technique by which those conclusions were reached. Science emerges when findings become linked into a coherent system in which they confirm, qualify, illuminate, and give meaning to one another. This takes time, particularly in fields concerned with human beings (he called them the humane sciences), where the number of relevant variables is large and the objects of study change with development, knowledge, culture, purpose, and context.

That, to my mind, is a much better way to understand the Science of Education. It’s not a toolbox containing tools like retrieval practice, spacing, worked examples, cognitive load, feedback, prior knowledge, metacognition, multimedia principles, and whatever other practices happen to enjoy empirical support at any moment. Nor is it a league-table of interventions sorted according to effect size. The Science of Education is the developing explanatory system that connects what we know about attention, perception, working memory, long-term memory, knowledge acquisition, practice, forgetting, retrieval, expertise, motivation, instruction, and transfer. It’s fed by psychology, sociology, and biology (in Dewey’s words, the sources). The interesting scientific questions aren’t whether an effect occurs, but why it occurs, what mechanism(s) might account for it, under which conditions it becomes stronger or weaker, how it relates to other established findings, and what evidence would force us to modify the explanation.

The dangerous jump from research to prescription

Once this is understood, another problem becomes visible. Even a robust scientific finding doesn’t tell a teacher what to do on Tuesday morning with Year 8 students. For Dewey, no conclusion from scientific research can simply be converted into a rule of educational practice. This isn’t because science is too abstract to matter, nor that classroom experience is somehow superior to empirical evidence. It’s because a research finding captures only part of the situation in which educational action takes place.

A well-designed experiment gains explanatory power partly by reducing complexity. Variables are controlled, held constant, excluded, randomised, averaged, or deliberately isolated so that the relation between a limited number of factors can be examined and a causal relation can be or not be seen. A classroom restores the lost complexity. Learners differ in prior knowledge, teachers in expertise, subjects in structure, tasks in goals. And the same instructional intervention may function differently depending on what came before it and what is expected to come after. The finding hasn’t become false when it leaves the laboratory, but neither has the laboratory supplied a complete specification for action.

Dewey further distinguishes between laws and rules. Scientific findings, he argues, should function as intellectual instruments for the practitioner. They should alter what the practitioner notices [think of my motto: what you know determines what you see], what relationships seem important, which explanations are plausible, and which alternatives deserve consideration. They are rules for better observation and inquiry, not fixed rules for overt action. Put differently, science should improve the quality of a teacher’s judgement and not replace judgement with compliance.

He even goes a step further writing that following research as a prescription:

…happens not because of scientific method but because of departure from it. lt is not the capable engineer who treats scientific findings as imposing upon him [sic] a certain course which is to be rigidly adhered to: it is the third or fourth-rate man [sic] who adopts this course. Even more, it is the unskilled day laborer who follows it.

Those are his words, not mine!

This matters because educators understandably like recipes. Teachers in training, Dewey observed, want to know how to do things with the greatest prospect of success. Under those circumstances, science is easily valued because it appears to put a stamp of approval on a particular procedure. Instead of becoming a means of intellectual illumination, it becomes a guarantee attached to a product: science says this works, so use it.

Original (1909) Good Housekeeping Seal of Approval

The irony is that once a scientifically informed idea is turned into a uniform procedure to be followed regardless of circumstances, it ceases to be an expression of scientific thinking and becomes just another rule-of-thumb, albeit one wearing a lab coat.

Cognitive psychology with a classroom sticker?

For Dewey, there’s a world of difference between the sources of an educational science and educational science itself. Psychology, sociology, biology, statistics, economics, etc. (the sources of science) can produce excellent science, but their findings don’t become educational science merely because someone adds a paragraph about classroom implications. Dewey argues that scientific results remain psychology, sociology, statistics, or whatever their parent discipline is until they become operative in the observation, judgement, and planning of those engaged in education.

That’s a useful warning against treating the Science of Education as cognitive psychology with a classroom sticker attached. Cognitive psychology is indispensable to our understanding of learning, and much of what we now know about memory, knowledge, practice, and expertise would be impossible without it. Yet understanding a memory phenomenon during an experiment and deciding how that understanding should be used when teaching fractions, historical causation, reading, or electrical circuits are different intellectual problems. The first informs the second, but it can’t substitute for it.

Dewey’s analogy is often engineering. Math and mechanics aren’t sciences of bridge building, yet are indispensable sources for it when selected and organised around the problems that arise when you build bridges. Bridge building, meanwhile, supplies the problems that stimulate inquiry and the practical tests through which theoretical conclusions are confirmed, modified, or exposed as incomplete. He writes, “Theory is in the end…the most practical of all things”. In education, the relation should also be reciprocal. The classroom isn’t just a place at the end of a research pipeline. Educational practice generates the problems worth explaining, and is also where our explanations encounter the complexity they must eventually help us navigate. It’s where science meets craft/art.

Practice comes first and last

Dewey, therefore, rejects the neat one-way picture where researchers discover truths, translators (Edfluencers; my term) turn them into recommendations, and teachers implement them. Educational practice, he says, comes first and last: it supplies the problems that give educational inquiry its point, and provides the arena where the usefulness of scientific conclusions is tested, revised, and developed. Scientific conclusions occupy an intermediate and auxiliary position. This doesn’t reduce science to whatever happens to work in a classroom. It makes the relationship between theory and practice more demanding, because neither side is allowed to ignore the other.

This also changes the role of the teacher. If teachers are treated merely as the delivery system for research conclusions, it’s not surprising when those conclusions are simplified into slogans, checklists, and recipes. Dewey thought classroom teachers should contribute to the development of educational knowledge because they’re in direct contact with the phenomena the science is trying so hard to understand (see his university school!). He described their potential contribution as an almost unworked mine and warned that if teachers are reduced to channels of reception and transmission, scientific conclusions will be distorted before they reach pupils. Teacher expertise isn’t an inconvenient source of variation that a scientific system should eliminate; it’s part of the intellectual machinery through which science becomes educative. That’s why teacher professional experience is one of the three pillars of evidence-informed working.

From Evidence Informed Teaching Practice

What this means for the Science of Education

A mature Science of Education should consequently make us both more confident and more cautious. More confident, because cumulative research can give us far better explanations than tradition, fashion, intuition, or the persuasive charisma of a particular educational guru. More cautious, because the more seriously we take science, the less willing we should be to elevate one study, one effect size, or one attractive mechanism into a universal classroom commandment. Scientific understanding becomes valuable in practice not when it tells teachers what to do, but when it helps them see more of the situation in front of them and react more effectively to it.

This is also why the cumulative character of science matters so much. The question should rarely be, “Is there research supporting this?” There is research supporting almost anything if the evidential bar is low enough and the search sufficiently selective. As friend and colleague Pedro De Bruyckere often says, “Everything works somewhere”. Better questions ask how a finding fits with a broader body of knowledge, whether it has survived replication and qualification, whether plausible mechanisms connect it with neighbouring findings, whether those mechanisms are relevant to the educational problem at hand, and what contextual features might change the expected outcome. The move from evidence to action is an act of informed reasoning, not a recipe.

Dewey’s most attractive claim may therefore be his insistence that science and professional artistry aren’t opponents. Education in concrete operation is an art, he says, but an art increasingly informed by science in much the same way as engineering and medicine. Scientific knowledge doesn’t make the capable engineer or doctor less capable of judgement, limit their options, or prescribe what they should do. It gives them more relationships to see, more consequences to anticipate, and more possibilities to choose from. Greater knowledge makes practice more flexible because the practitioner is liberated from blind tradition and isolated precedent.

A solid Science of Education can help us produce better informed teachers with better professional judgement and better intellectual tools who understand the systems of knowledge from which findings derive their significance. It’s not a conveyor belt running from laboratory to classroom, but a two-way street that travels in both directions with science informing practice and practice informing science.

So the next time someone posits “the research says ...”, our response shouldn’t be scepticism nor blind faith, but rather scientific curiosity: What, exactly, did the research find? How does that finding fit into a cumulative body of knowledge? What theory or mechanism gives it meaning? Under what conditions was it established? Which parts of the educational situation were absent from the study? And, most importantly, how did we get from the finding to this particular prescription for practice?

Dewey was asking versions of these questions in 1929. Nearly a century later, we have vastly better methods, far more data, and hugely richer sciences on which education can draw. Whether we’ve also become better at resisting the temptation to turn science into recipes is another matter.

Dewey, J. (1929). The sources of a science of education. Horace Liveright.

Subscribe now

Share

Leave a comment

Generated by ChatGPT

There’s a certain something that always occurs during my keynotes even though I ask the audience not to use their WMDs (Weapons of Mass Distraction). I present a slide containing something the audience apparently finds important (a model, a quote, a diagram, a summary of important results) and almost immediately, a slew of smartphones rises from the audience and with a few quick taps, the slide has been captured and the phones momentarily disappear.

If I were generous – not really my style - I could interpret this as: (1) they find the slide important or interesting, (2) they want to remember it, and (3) they think that the best way to do this is to take a picture. Of course it’s a great way to post something on social media, and since I’m often in the picture too it’s great PR for me, but for remembering? I don’t think so, but I also, until now, had no proof that they were just kidding themselves.

A fascinating new paper by Sophia Fabrizio, Rebecca Lurie, and Deanne Westerman, Digital amnesia: The aftermath of a screenshot, supports that doubt (thanks Carl Hendrick for the tip). Across seven preregistered experiments with 892 participants, they investigated what happens to memory when people shoot something rather than just look at it. [Pity that they didn’t compare it to taking notes on the image.] The result is remarkably consistent: people remembered the screenshotted material worse than just viewing, even when the researchers manipulated whether participants expected a test.

Generated by ChatGPT

Generated by ChatGPT

The finding itself isn’t really new. We’ve known for some time about what’s called the photo-taking impairment effect (Henkel, 2014): photograph something and, under many circumstances, you subsequently remember less of it than if you had simply looked at it. Think about this the next time you’re on vacation and are constantly photographing what you see instead of enjoying the view. What makes Fabrizio et al’s study especially interesting is that the effect extends to one of our most mundane digital behaviours. There’s no complicated camera to operate, no fiddling with focus or exposure or depth of field. The participants used their own smartphones, and carried out something they were utterly familiar with. Yet the simple act of making the digital ‘memory’ was associated with poorer remembering.

The obvious explanation is cognitive outsourcing (Fabrizio et al refer to this as offloading, but if you’ve read my recent blogs, you’ll see that it’s actually outsourcing). Why remember something yourself if your telephone can remember it for you? You’ve outsourced the memory to your phone! We do this all the time, and often very sensibly. I don’t memorise every appointment because my calendar does that for me. But I do have to enter it myself, just like I did with a paper calendar in my vest pocket. We’ve always externalised different forms of cognition; notebooks/notepads, diagrams, to-do-lists, shopping lists are all tools for cognitive offloading. We don’t have to maintain them in our working memories allowing us to use that space for other things. There’s nothing wrong with that, as long was we’re originally doing the thinking (i.e., cognitively processing the information). If we aren’t then we’ve outsourced that processing.

And here’s where the paper becomes particularly interesting. If taking the screenshot was really offloading, allowing us to stop remembering the captured information, perhaps the cognitive resources thereby freed become available for something else. That would at least give us a trade-off: I remember this less, but as a consequence I can think about or remember something else better.

Fabrizio and colleagues went looking for precisely those benefits but had difficulty finding them. Screenshots didn’t improve memory for where information had been encountered, didn’t improve performance on a subsequent cognitively demanding arithmetic task, and didn’t improve memory for things encountered afterwards. One experiment produced a modest benefit for later learning, but only under rather specific conditions and when recall was scored leniently. As the authors conclude, benefits appear to be the exception rather than the rule. In other words, “I don’t need to remember this because I have photographed it” may be a very bad bargain. You sacrifice the memory but don’t reliably get the supposed cognitive dividend.

Even more intriguing is that the authors aren’t convinced that deliberate offloading [their words] explains the phenomenon at all. In later experiments - remember they had seven - they explicitly encouraged situations where strategic offloading should help. Participants knew they needed to remember the material; some were even told that they would be able to consult their screenshots before the test and that the test would not take place for a month. If people were consciously deciding “my phone remembers, therefore my brain doesn’t have to”, these manipulations should have mattered. But they didn’t; the memory deficit remained.

The authors therefore consider another, possibly more plausible explanation, namely attentional disengagement. Taking the screenshot may itself signal, at some level, that our cognitive work is done. We saw the slide, heard what the presenter had to say, so brain: stop processing. We’ve captured the information, and psychologically (cognitively) we move on. Especially telling is that participants weren’t only worse at remembering the screenshotted pictures; they were also worse at remembering their source, suggesting poorer encoding of the surrounding context as well.

And now we return to the presentation I’m giving. I’m speaking about something that I hope people find interesting and will want to remember. A diagram appears on the screen. Someone thinks, “Hey that’s important.” Out comes the phone. They frame the slide, press the button, inspect the little image for perhaps half a second and put the phone away. From the audience member’s point of view: Mission accomplished. Except that I assume that the mission wasn’t to possess a JPEG of my PowerPoint slide, but rather to learn something.

This distinction is important because learning requires changes in long-term memory (Kirschner, Sweller, & Clark, 2006). To learn, information must be attended to, interpreted, connected to what a person already knows, elaborated upon, and ultimately, made retrievable. Taking a photo accomplishes none of these things. It produces an extraordinarily faithful external representation while at the same time producing a really poor internal one (if at all). Capturing information ≠ acquiring knowledge.

There is a certain comic irony here. When someone takes a photograph of one of my slides, they may feel that they have been especially diligent. They’re not ignoring the presentation; quite the contrary, they’ve identified something that they think is valuable enough to preserve. But that very act of preservation may have give their cognitive system permission to disengage from it. The photograph then joins hundreds or thousands of others in a digital graveyard, as the authors say, “buried somewhere in their camera roll”.

And this is where an important qualification is necessary. The researchers didn’t study people photographing PowerPoint slides during keynote lectures. Their participants mostly screenshotted visual material under controlled lab conditions. This means we can’t turn an interesting set of experiments into another breathless proclamation that “research shows hat photographing slides destroys learning”. Research showed no such thing.

Nor do the authors argue that screenshots are always harmful. Their experiments concern digitally captured material that is subsequently not revisited. A carefully chosen screenshot that you later retrieve, study, annotate, compare, discuss, or use for retrieval practice can obviously be useful. [Pardon my cynicism here, but I really don’t think that many of those taking the photos do that, but I haven’t studied that so I may be wrong here.] External memory aids work when they are actually used as memory aids. The authors explicitly make this distinction: a small number of intentional screenshots followed by genuine re-engagement may support memory through rehearsal, while accumulating unreviewed screenshots imposes the cost without obtaining that benefit.

So maybe I should say to my audience when they point and click, “Before you click, think about the slide. Ask yourself why you want to take the photo, write that down, and if you still want the photo, then take it”, but I probably won’t.

Yes, your phone has an excellent (photographic) memory but unfortunately, you don’t get to borrow it. There’s no usb-cable or Bluetooth connection from your phone to your brain.

Fabrizio, S. P., Lurie, R., & Westerman, D. L. (2026). Digital amnesia: The aftermath of a screenshot. Memory & Cognition. https://doi.org/10.3758/s13421-026-01921-2

Henkel, L. A. (2014). Point-and-shoot memories: The influence of taking photos on memory for a museum tour. Psychological Science, 25(2), 396-402. https://doi.org/10.1177/0956797613504438

Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1

Subscribe now

Share

Leave a comment

Generated by ChatGPT

You’ve seen and heard it before. We’re now being told, usually with great confidence and little to no evidence, that Generative AI (GenAI) will “transform” education. This is the umpteenth instance of the same old claim that some technological tool will revolutionise teaching and learning^[1]. The newest claim is that GenAI will save teachers time, personalise instruction for the student, produce lesson plans, generate exercises, differentiate materials, and write feedback. The problem is that producing text quickly isn’t the same as teaching well, and saving teacher time isn’t the same as improving student learning.

A randomised field experiment by Alp Sungu, Benjamin Lira, and Angela Duckworth^[2] (Generative AI Can Harm Teaching) provides a much needed reality check. Across fourteen middle and secondary schools in Turkey, teachers were randomly assigned either to continue their teaching as usual or to receive access to a GPT-4o-based teaching tool, with some also receiving reminders^[3]. The analysis covered 193 teachers, 2,816 students and more than 14,000 student-course observations; not bad for a RCT in a real school under real learning conditions. Note: this wasn’t a self-report survey nor a lab study using an artificial task. It examined what happened when teachers used GenAI in real schools.

Students whose teachers received access to AI reported lower intrinsic motivation. They found their courses less enjoyable, less interesting and less important. The overall decline was 0.11 standard deviations, and was larger when teachers already were heavier AI users. Student confidence also declined, though the average effect was less statistically certain. On average, examination performance didn’t change significantly, which will no doubt allow enthusiasts to say that AI “did no harm”. But averages are often the nooks and crannies where educationally important effects go to hide.

Among students taught by lower-performing teachers, the use of GenAI reduced academic achievement by 0.129 standard deviations and their confidence by 0.183 standard deviations. Among students with stronger teachers, there was no statistically reliable harm and perhaps even a small positive effect. In other words, the technology sold as a way to compensate for differences in teacher quality appears, in this study, to have widened them. The Matthew Effect strikes again! Why? Stronger teachers may have been better able to judge, reject, adapt, and integrate what Chat produced. Weaker teachers, on the other hand, seemed to be more likely to let the tool substitute for the thinking that good teaching requires.

This distinction is essential. A lesson plan isn’t teaching, a worksheet isn’t instruction, and a set of questions isn’t formative assessment let alone retrieval practice. These products of GenAI only become educationally useful when a knowledgeable teacher selects examples, anticipates misconceptions, sequences explanations, decides what their students already know, judges where support is needed, and adapts what happens next. That work isn’t simply ‘clerical preparation’ surrounding teaching. It’s a/the central part of teaching! When we outsource it to GenAI or wherever (the teacher edition of a teacher-proof learning method, a colleague who has already done all the groundwork…), we are removing the cognitive activity through which we as teachers understand the content and the students we are about to teach. In Donald Schön’s words: our reflection on action and our reflection in action.

The usage data are also revealing. Roughly two-thirds of teachers’ conversations with the GenAI concerned producing teaching materials, such as lecture notes, homework, examinations, and syllabi (i.e., make this for me). Only a much smaller share involved instructional support such as addressing misconceptions or differentiation. The median conversation contained just two teacher prompts. In other words, this median of just 2 prompts means that “teachers typically accepted AI-generated outputs with minimal iteration, treating the tool as a generator of finished products rather than as an iterative collaborator”. This doesn’t look like (pardon my sarcasm) a thoughtful dialogue where a professional interrogates suggestions, tests alternatives, and refines a pedagogical approach. It looks much more like delegation or work and responsibility: Ask, Receive, Copy, Use. Hey, a new acronym ARCU 😊

And that may explain the motivational result. Students don’t experience teaching as the sterile transfer of technically adequate content from one container (the teacher) to another (the learner). They experience (and expect) a teacher; their examples, humour, stories, expectations and ways of explaining difficult ideas. GenAI tends towards being as smooth, generic, and plausible as possible. It can effectively and efficiently generate a lesson that looks like a lesson, just as it can generate an explanation that looks like an explanation. But generic instructional prose may remove precisely the personal and intellectual signature that makes a course coherent and a teacher credible. When students sense that materials haven’t really been thought through by the person standing in front of them, why should they invest themselves in them?

IMHO the equity implications of this research deserve particular attention. This study didn’t measure family income or socioeconomic disadvantage, so we won’t pretend that it directly proves greater harm for poorer pupils. What it does show is that the damage was concentrated among students taught by less effective teachers and research has taught us that lower-achieving students and those from poorer neighbourhoods and schools have poorer (i.e., lower-achieving) teachers^[4]. In addition, more advantaged students often have more resources: better educated parents, access to tutoring, a quiet study space, books, and the confidence to ask for help. Students with fewer resources depend more heavily on the quality of what school and their teachers provide. When AI weakens the teaching of those who most need support in exercising professional judgement, the pupils least able to compensate outside school are likely to carry the greatest risk.

There are limitations to this study. The intervention lasted just one semester, the examinations showed ceiling effects, and the tool was one particular implementation. Perhaps better training, stronger guardrails or better-designed systems would produce different results. But notice here how quickly the burden of proof shifts. Before evidence appears, AI is promoted as an obvious improvement. When evidence of harm appears or when it is shown to be worthless, we’re told that the implementation was imperfect or that future versions will be better. Education has seen this film before; Back to the Future XXI.

The conclusion here isn’t that teachers shouldn’t use GenAI. As a tool, it may be useful for low-stakes administrative tasks, locating resources, or generating possibilities that a teacher then evaluates critically. It can also be used as a sparring partner to stimulate one’s thinking or comment on proposals. But it should not be used to replace the intellectual preparation through which teachers build pedagogical content knowledge and make instruction their own. A tool that saves time by removing the thinking needed for good teaching isn’t efficient, it’s educationally expensive.

The question schools should ask isn’t, “How much time can AI save teachers?” It’s, “Which parts of teachers’ work must remain cognitively and professionally theirs because pupils’ learning depends on them?” Until we answer that question with evidence rather than enthusiasm or advertorials, scaling GenAI in classrooms isn’t innovation. It’s an experiment conducted on children which IMHO is morally and ethically unacceptable, with the greatest risks borne by those who can least afford another educational disadvantage.

Subscribe now

Share

Leave a comment

^[1] In actuality, I know of only two technological revolutions in education, namely the printing press and the blackboard The printing press was a revolution as it made access to and spreading of information possible by mass producing textbooks and other scholarly works. The blackboard was a revolution because it shifted teaching from individual, oral recitation to collective visual instruction, thereby enabling one teacher to guide an entire room at the same time.

^[2] Yes, Professor Grit, and as you know not one of my favourites so you can safely assume that I was critical of the article.

^[3] Teachers were provided with the same tool plus weekly reminders containing individualised usage statistics to encourage continued engagement.

^[4] Babu and Mendro report that low-achieving students are twice as likely to be assigned to ineffective teachers than high-achieving students.
Babu, S., & Mendro, R. (2003, April). Teacher accountability: HLM-based teacher effectiveness indices in the investigation of teacher effects on student achievement in a state assessment program [Conference presentation]. AERA Annual Meeting, Chicago, IL

Illustration: ChatGPT / Text: Me

My dear friend and colleague Pedro De Bruyckere attended me to this article - The End of Reading Is Here by Rose Horowitch - in the Atlantic. If I were to use two words to characterise it, they would be dark and apocalyptic. In essence it’s a saunter through the history of literacy beginning with the Library of Alexandria through artificial intelligence, the role and function of reading and writing for our thinking our brain and society in general, and the effects of our loss of literacy in what is called the post-literate society. It’s one of the darkest articles that I’ve read in a long time and isn’t really something someone should read first thing on a cheery, sunny, weekend morning. It’s also very ironic and that’s what I want to discuss here. For the dark and apocalyptic I would suggest everyone who is still literate take the time and make the effort for this longread.

The first irony that I saw was that it’s a longread full of long, thoughtful, deep sentences; a joy to read. It’s full of references both to what has earlier been said in the article as well as to history, psychological and media research forcing the reader to think and make inferences, those things that it also argues that we have and are becoming less able to do, namely read long and intricate sentences in large pieces of text. Horowitch writes, “New York Times best sellers today have sentences that are about one-third shorter than they were a century ago.”

In 1958, the English translation of Boris Pasternak’s Doctor Zhivago was the best-selling novel of the year, according to Publishers Weekly. Pasternak writes in long, complex sentences: “On that warm gray morning in the mountains, Zhivago felt sorry for the Tsar, was disturbed at the thought that such diffident reserve and shyness could be the essential characteristics of an oppressor, that a man so weak could imprison, hang, or pardon.”

The most popular novel written for adults [last year] was the romantasy adventure Onyx Storm. Whatever the book’s pleasures, it isn’t Pasternak: “A muscle in his square jaw ticks as he stares down at me, rippling the tawny-brown skin of his stubbled cheek.”

In that respect it’s preaching to the converted, something I also often tend yo do. It also limits the audience that can/will read and understand it as you need to take the uninterrupted time to read and think about it. It’s exactly the opposite of what today’s, younger, readers can and will do. Horowittch herself notes this when she writes, “But the people devoted to text, who derive cultural understanding and intellectual connection from the written word, are now part of a subculture. The fact that you are reading this article almost certainly makes you a member of it.” Oh the irony.

The second irony was that directly under the title is the possibility to not read it, but rather to listen to in (53 minutes and 24 seconds). Yes, I understand that at this moment in time all text from all sources including my own books must be partially or completely afforded the possibility of their being transformed in real time from written texts to AI-produced oral texts. Yes, it definitely makes the text more inclusive for those with visual impairments, but it’s still ironic to me. Being able to listen to the text, amongst other things, makes the text oral and, thus, allows the you – now the listener and not the reader - to do other things while listening to it removing the possibility of thinking about it while listening because our attention is limited and the possibility to reread passages and think about them is lost due to orality’s ethereal nature (i.e., once said it floats away into the ether; it is transitive). Horowitch discusses these and other differences (and limitations) of oral language in the article. She discusses the difference between literacy and oracy and the repercussions of the trend from the former to the latter, ultimately handling Ong’s discussion of ‘Literacy and Oracy’^[1] and our arrival in a post-literate society. To quote Horowitch,

Because spoken words disappear as soon as they’re uttered, oral cultures value repetition to aid memory. Bards in oral societies make use of stock phrases and mnemonics to keep track of their train of thought. They traffic in epithets and “enthusiastic description of physical violence,” in Ong’s words, because conflict is more memorable than dispassionate discussion. Speakers can’t edit their words the way writers can, so they press on without admitting their mistakes. If they later contradict themselves, they don’t expect the audience to recall their earlier statements. Meaning depends on the identity of the speaker, not on any concept of objective truth.

This is best illustrated by Donald Trump being “our first postliterate president”

It is unlikely that Donald Trump has familiarized himself with Orality and Literacy. But if he did, he might recognize himself in Ong’s description. Trump’s communication style is perfectly suited to an oral society. He employs epithets—“Low-Energy Jeb,” “Little Marco,” “Sleepy Joe”—that are easy to remember and repeat. He contradicts himself as though there is no record of his previous statements. Even his writing is almost indistinguishable from his speech. (It makes sense; Trump reportedly prefers dictation to composition.) His online posts are full of idiosyncratically placed punctuation, capital letters, and exclamation points. Many are memes with little text: One featured an image of an American warship hitting an Iranian airplane with a laser beam and included the phrase “Lasers: Bing, Bing, GONE!!!”

I would have loved to end this post in a sunny way on this sunny weekend, but can’t so I’ll just quote Horowitch here again.

Could the generations growing up with their brains hooked to endless video feeds be developing some kind of novel, as-yet-undetectable cognitive brilliance? Perhaps. But for now, the decline of reading seems to be ushering in a less rational, analytical, and sophisticated mode of thinking. It’s difficult to see any advantages in that.

Subscribe now

Share

Leave a comment

And here’s an interview with the author on Plain English with Derek Thompson:

Horowitch, R. (2026, July). The end of reading is here. Atlantic.

^[1] Ong, W. J. (1982). Orality and literacy: The technologizing of the word. Routledge.

Share

Leave a comment

Two notes before you begin reading this post. The first is that this piece is longer than it should be. It’s a synopsis of three posts by others totalling more than 5000 words with my own thoughts added. If it’s too long, then here’s a synopsis of my synopsis:

This post synthesises recent critiques of Alpha School by Robert Pondiscio, Jared Cooney Horvath, and Dylan Kane, examining its claim that students can complete academic learning in two hours a day through adaptive software and AI tutoring. While the model may offer genuine efficiencies and valuable enrichment, its assertion that students learn 2.6 times faster remains difficult to evaluate because the underlying data, methods, student selection, and wider school conditions have not been independently scrutinised. Moreover, they argue that rapid progress on software-aligned assessments does not necessarily demonstrate durable learning, long-term retention, or transfer to unfamiliar contexts. The post also questions whether screen-based, adaptive systems encourage deep understanding or merely teach students to optimise for task completion, potentially replacing human judgement by the teacher with increased monitoring by the system. More fundamentally, it asks whether Alpha is reducing academic subjects to content that should be mastered as quickly as possible, overlooking the role of teachers, other students, discussion, shared intellectual culture, and sustained engagement in developing knowledge and independence. Alpha should therefore be regarded as a possibly promising educational experiment, not yet as evidence that AI can replace teachers or make conventional schooling obsolete.

The second is that I was aided in this exercise by ChatGPT (for example in writing the above synopsis). Not the least in helping to turn my acquired Dunglish into something resembling English.

I’ve been bombarded in the last few days by blogs, articles, Substacks, etc. about Alpha School. Apparently – understatement – there’s a kerfuffle about the school with supporters, detractors, and people on the fence. This a synopsis of three recent pieces that I found in my feed. One is from Robert Pondiscio, one is from Jared Cooney Horvath, and one is from Dylan Kane. All three pose more questions than answers, and that’s why I’m combining them here for you. I definitely advise everyone to read the original pieces as well as follow these three.

Alpha School makes an extraordinary promise: children can complete their academic learning in just two hours a day, using adaptive software and AI tutors, and spend the rest of their time developing practical skills, pursuing projects, and following their interests.

It is a very appealing proposition. Traditional schools do waste time. Lessons aren’t always well-designed, kids don’t always receive the support they need, and teachers are expected to manage enormous differences in prior knowledge within a single classroom. And don’t get me started on things that are in the curriculum that shouldn’t be there or the use of evidence-uninformed instructional methods in schools!

While technology might help address some of the problems, Alpha School isn’t merely claiming that technology can make parts of schooling more efficient. It presents its model as a fundamental breakthrough (yes, the umpteenth!): students supposedly learn more than twice as fast as expected, teachers, according to many of the school’s proponents, are no longer required for academic instruction, and conventional schooling, thanks to Alpha has effectively become virtually obsolete.

But, as Carl Sagan said: “Extraordinary claims require extraordinary evidence”.^[1] According to Wikipedia, he used the phrase in his 1979 book Broca’s Brain and the 1980 television program Cosmos. What he meant is that claims of this magnitude require more than impressive testimonials, school visits, and attractive graphs. They require evidence that can survive independent scrutiny. More importantly, they require us to ask a question that Alpha’s marketing tends to avoid: what exactly is learning?

Extraordinary claims require extraordinary evidence

Robert Pondiscio’s wrote an open letter to Alpha’s founders and in it he identifies the most immediate problem. Alpha has published summaries of its MAP (Measures of Academic Progress) test results and repeatedly claims that its students learn “2.6 times faster”. Yet independent researchers haven’t been able to reconstruct how that figure was calculated.

And that matters. We must know how expected progress was defined, which pupils were included, whether any were excluded and why, how long they attended the school and when and why they stopped, and whether the assessments were administered under comparable conditions. We also need to know how much of Alpha’s performance is attributable to the software and how much to its highly selective environment.

Alpha is a very expensive private school, on a par with or even more expensive than many universities. It serves families willing and, apparently, able to choose an unconventional educational model and pay for it. That’s not a dismissal, buta fact. Due to their (family) backgrounds, its students may enter the school with greater prior knowledge, stronger parental support, or unusually high motivation. I say ”may” because I don’t have the numbers, but a lot of research shows this to be the case for children from high SES families. The Alpha-campuses also provide guides, incentives, enrichment activities, and a carefully constructed culture around performance. None of this invalidates the results, but it makes simplistic causal claims impossible.

Perhaps Alpha really does work exceptionally well. Perhaps its success depends on a combination of affluent families, strong leadership, close adult supervision, adaptive software, and a distinctive school culture. That would still be an impressive achievement, though as I’ve argued many times with respect to program-based studies as a whole (see here for example), it says relatively little about any of the separate parts of the programme (the instructional variables), including the AI. But it would be quite different from demonstrating that AI tutoring can replace teachers or transform ordinary public education.

Robert Pondiscio argues that transparency would help Alpha, not harm it. A genuine breakthrough becomes stronger when outsiders can examine the data, test alternative explanations and reproduce the findings. At present, Alpha asks the public to admire the speedometer without allowing researchers to inspect the engine.

Slow but steady wins the race

The second difficulty is related to Alpha School’s emphasis on speed. Jared Cooney Horvath’s discussion of digital tutoring and transfer provides an important warning, namely that becoming proficient within a learning environment isn’t the same as being able to use that knowledge elsewhere.

Digital systems are particularly good at producing rapid improvements on tasks that resemble the practice pupils have completed. They can identify errors, adjust difficulty, provide immediate feedback and keep learners working near their current level of competence. On assessments closely aligned with the software, the results may be very impressive. But the ‘learning’ or learning gains often become much smaller when learners encounter an independently designed assessment or must apply their knowledge in a different context. Horvath cites meta-analyses in which the effects of digital tutoring were considerably larger on locally developed tests than on independent measures. That’s one of the reasons why I’m particularly wary of the results of research that use experimenter-designed assessments. He also describes studies in which AI improved performance during supported practice but failed to improve—and in one case apparently harmed—later unaided test-performance. This one is particularly telling.

This distinction is essential when evaluating Alpha. MAP growth may tell us that kids are progressing on a standardised adaptive assessment. It doesn’t, by itself, tell us whether they can explain their reasoning, connect ideas across subjects, write sustained arguments, remember what they learned a year later, or apply their knowledge to unfamiliar problems.

Transfer is difficult because knowledge is initially connected to the circumstances in which it was learned (is episodic). Robust knowledge develops when learners encounter ideas repeatedly, over time, in varied situations and representations. They read about them, discuss them, write about them, solve problems with them and receive guidance from people who understand both the subject and the learner. The episodic becomes semantic.

That process can look inefficient. Learners revisit material. Teachers ask questions whose answers can’t be reduced to a score. A historical concept reappears in literature; a mathematical idea is used in science; a misunderstanding is exposed during discussion. Progress through the curriculum may appear slower precisely because knowledge is becoming richer and more transferable. A kind of desirable difficulty.

Alpha’s two-hour promise risks mistaking rapid movement through a teaching sequence for deep learning. Efficiency is valuable only when we’re certain that the right thing is being made more efficient.

The delivery truck again

The third problem is the assumption that the medium through which pupils learn is neutral. Dylan Kane challenges this directly. He argues that a screen isn’t simply a more convenient textbook, just as an AI tutor isn’t simply a teacher that happens to run on a computer. Different media encourage different behaviours. On screens, pupils are used to clicking, scrolling, skimming, gaming systems, and searching for the fastest route to completion. Kane describes students deliberately answering questions incorrectly so that an adaptive mathematics program would give them an easier path through the exercise. They understood and could game the software perfectly. They’d learned to optimise for finishing rather than understanding. A perfect example of the discipulus economicus (the calculating learner).

No amount of personalisation automatically removes this problem. Adaptive systems may actually even make it worse by turning learning into a continuous exchange between learner/user behaviour and machine response. The learner learns not only mathematics or reading but also how the system rewards, redirects, and evaluates them. This helps explain why AI-driven schooling so readily leads to technological monitoring. When a teacher is no longer responsible for maintaining attention, diagnosing disengagement, or noticing confusion, the software needs other ways to determine whether pupils are genuinely working. Cameras, eye tracking, response latency, and behavioural analytics become tempting solutions.

Such monitoring might improve compliance, but it also changes the nature of accountability. Being held to account by an adult who knows you, understands the subject being taught, has a history with you, and can exercise judgement isn’t the same as being continuously watched and measured by a system. One is a human relationship; the other resembles (is?!) surveillance.

The question is therefore not simply whether AI can keep kids on task; it’s what kind of learner—and ultimately what kind of person—the environment encourages them to be and become.

School isn’t an obstacle!

Alpha School’s rhetoric also reveals a very narrow conception of ‘academic’ education. Its model often treats mathematics, literature, science, and history as content to be mastered as quickly as possible, leaving the child more time for the supposedly more meaningful work of entrepreneurship, public speaking, teamwork, sports, and life skills.

But imho academic subjects aren’t just administrative hurdles that kids have to clear before their ‘real’ education and learning can begin. They’re part of that real learning and education. Reading literature isn’t merely vocabulary acquisition or the application thereof. Studying history isn’t just remembering people, places, and dates. Mathematics isn’t a ladder of isolated tricks and competencies that you have to quickly climb, one or more rungs at a time. The knowledge gained in these subjects introduces learners to forms of thought, bodies of knowledge, ongoing human conversations, and thinking! They require attention, patience, and sustained engagement and effort. And strangely enough, sometimes their value lies precisely in the fact that they can’t be compressed into an efficient transaction or an AI conversation.

School also offers more than personalised content delivery. Learners listen to other people’s ideas, formulate responses to questions and premises, encounter competing answers to questions and interpretations of artefacts, and learn that their own first answer may be incomplete or even wrong. A good teacher doesn’t merely select the next topic, exercise, or project. They represent a subject, communicate standards and values, notice and correct misconceptions, create both a shared intellectual culture and a psychologically safe learning space, and decide when a child needs more or different explanation, practice, encouragement, or challenge.

Alpha School hasn’t removed adults from education. Rather, it appears to have reassigned them to being the ultimate guides on the side. They’re there to motivate, maintain culture, and supervise the non-academic part of the day. What they aren’t there for seems to be teaching. This raises an important possibility related to what I spoke of with respect to programme-based studies: maybe, just maybe, the human relationships and institutional culture are doing far more of the work than Alpha’s claims about AI acknowledge.

An experiment isn’t a revolution

None of this demonstrates that Alpha School is ineffective. Its model may produce excellent outcomes for some/many of its pupils, its use of adaptive practice may be genuinely efficient, and its afternoon activities may offer experiences that conventional schools too often neglect or don’t have the means to carry out. Alpha School purports to be different and education needs institutions willing to question established notions and routines - within limits. [To this end, I’m completely opposed to schools using children as guinea pigs for half-baked educational philosophies and experiments.] The problem here is that Alpha is claiming victory before the central questions have been answered and before others have had the chance to examine the results.

In Dutch we have two sayings related to this. The first was an advertising slogan: Wij van WC-Eend adviseren WC-Eend; We at Toilet Duck recommend Toilet Duck. In other words, that you promote your own product or service and then reap the (financial) benefits yourself. A second is de slager die zijn eigen vlees keurt; the butcher inspecting his own meat and thus that claiming quality without independent verification. Alpha School is at this moment the proverbial toilet cleaning product or butcher.

Some of the central questions are: Do Alpha School pupils perform or really retain what they learn? Can they transfer it beyond the tutoring environment? Do they develop broad and coherent subject knowledge? How do they perform on extended, independent tasks? Which students thrive, and which leave the school? How much does the model depend on family background, funding, adult supervision, and school culture? What happens when it is introduced in less privileged settings? And can independent researchers examine the underlying data?

Until those questions receive convincing answers, Alpha should be treated as an interesting and ambitious school, not as proof that teachers, textbooks, and conventional classrooms are obsolete. AI may improve education. It may help teachers select practice, diagnose misconceptions, provide feedback, and reduce administrative work. But using technology to strengthen teaching is different from using technology to redefine education as the rapid completion of academic content.

Alpha School’s most important experiment is therefore not whether kids can finish their lessons in two hours. It is whether speed, personalisation and measurement can produce the durable knowledge, intellectual independence, and human growth that schooling is meant to cultivate. On that question, the evidence is not yet moving 2.6 times faster. It’s barely begun.

^[1] This is very much related to Hitchens’s razor: What can be asserted without evidence can also be dismissed without evidence. Wikipedia writes: Its roots can be traced to the Latin Quod gratis asseritur, gratis negatur (What is freely asserted is freely denied). It implies that the burden of proof regarding the truthfulness of a claim lies with the one who makes the claim; if this burden is not met, then the claim is unfounded, and its opponents need not argue further in order to dismiss it.

In de meeste gevallen nemen wij informatie niet simpelweg op en leren wij niet zomaar nieuwe dingen. We moeten die informatie ook actief interpreteren, verbinden met voorkennis, het geheugen reorganiseren en steeds verfijndere mentale representaties opbouwen. Leren is, naar zijn aard, een constructief cognitief proces. Daar zullen maar weinig cognitief wetenschappers het mee oneens zijn.

Maar in de afgelopen decennia is er iets subtiels gebeurd.

Een beschrijvende bewering over hoe leren plaatsvindt is stilletjes veranderd in een voorschrijvende en zelfs normatieve bewering over hoe leraren zouden moeten onderwijzen.

De redenering verloopt vaak als volgt:

· Leren is constructief.

· Daarom kan kennis niet worden overgedragen.

· Daarom is expliciet onderwijzen ineffectief.

· Daarom zouden leraren expliciete/directe instructie moeten minimaliseren of zelfs vermijden.

· Daarom zouden leraren niet moeten onderwijzen, maar slechts faciliteren en leerlingen hun “eigen” kennis laten creëren.

Geen van die conclusies volgt logisch uit de oorspronkelijke premisse (Leren is constructief).

Want zodra we verschuiven van “hoe mensen leren” naar “hoe leraren zouden moeten onderwijzen,” hebben we de vraag volledig veranderd.

Het tweede vereist eigen bewijs, omdat het in de wereld van het progressieve onderwijs is veranderd in een ideologisch voorschrift (is dat niet ironisch?).

Voordat ik verder ga, laat ik even een stap terug doen voor wie misschien minder vertrouwd is met wat hier speelt. Eerst: wat is constructivisme en waarom bestaat er onder onderwijsgevenden controverse over?

Hier is een algemeen aanvaarde definitie:

“Constructivisme is een leertheorie die verklaart hoe lerenden actief mentale representaties (schema’s) construeren door nieuwe informatie te integreren met hun bestaande kennis.” (Sommige definities voegen daaraan toe: via persoonlijke ervaring en interactie met de omgeving.)

Eenvoudig gezegd stelt de theorie dat wij nieuwe informatie van buitenaf opnemen, die verbinden met de voorkennis in ons geheugen, en zo nieuw begrip of nieuwe betekenis construeren.

Tot zover is er weinig controverse.

De controverse begint wanneer dit beschrijvende verhaal over hoe leren plaatsvindt stilletjes wordt omgevormd tot een voorschrift voor hoe leraren wel en niet zouden moeten onderwijzen.

Precies voor dat onderscheid waarschuwt Douglas H. Clements (1997) onderwijsgevenden.

Zoals hij stelt, is constructivisme in de kern een theorie van leren; geen theorie van instructie. Constructivisme verklaart wat er tijdens het leren in het hoofd van de lerende gebeurt. Op zichzelf bepaalt het echter niet of hoorcolleges, demonstraties, observatie, uitgewerkte voorbeelden, begeleide oefening, onderzoekend leren, experimenten of discussie de meest effectieve instructiemethoden zijn. Dat zijn empirische vragen die afzonderlijk bewijs vereisen.

(Helaas heeft dat er niet voor gezorgd dat dit -isme geen dogma is geworden.)

Drie verschillende betekenissen van “kennis construeren”

Wanneer iemand tegen mij zegt: “Leerlingen construeren hun eigen kennis”, wil ik een andere vraag stellen: Wat construeren zij precies?

Want het woord kennis doet hier minstens drie heel verschillende dingen.

Onderwijsgesprekken en -debatten behandelen die vaak alsof ze onderling uitwisselbaar zijn. Maar dat zijn ze niet.

1. Construeren leerlingen hun eigen kennis?

Als de bewering is dat leerlingen wiskunde, scheikunde, grammatica of natuurkunde zelf uitvinden, dan wordt die bewering al snel ongeloofwaardig.

Leerlingen construeren niet het feit dat water H₂O is. Zij construeren de abc-formule niet. Zij construeren het periodiek systeem niet.

Dit zijn producten van eeuwen van collectief onderzoek en van enorme gedeelde kennis.

Scholen bestaan nu juist omdat niet elke generatie de beschaving vanaf nul opnieuw hoeft te ontdekken.

2. Construeren leerlingen persoonlijk begrip?

Dan hebben we het over iets heel anders.

Een leerling verbindt de wetten van Newton met fietsen. Een ander verbindt evolutie met ecologie. Er is niets nieuws toegevoegd aan de kennis van de mensheid. In plaats daarvan wordt bestaande kennis betekenisvol voor de lerende.

Dit is zonder twijfel een onderdeel van leren.

3. Construeren leerlingen mentale schema’s?

Hier plaatst de moderne cognitieve wetenschap het constructieve proces. Nee, leraren kunnen schema’s niet in iemands brein uploaden. Geen enkele leraar kan begrip rechtstreeks overdragen.

Maar dat betekent niet dat leraren niets overdragen.

Leraren dragen kennis over via uitleg. Demonstraties. Uitgewerkte voorbeelden. Vragen. Corrigerende feedback. Begeleide oefening.

Leerlingen interpreteren die input actief, reorganiseren het geheugen, versterken retrievalpaden en construeren geleidelijk steeds verfijndere schema’s.

De constructie vindt plaats in de lerende. Maar de ruwe materialen worden grotendeels door de leraar via instructie aangeleverd (of overgedragen).

Het onderscheid dat alles verandert

Dit is de kern: leerlingen construeren geen publieke kennis. Zij construeren privébegrip van publieke kennis.

Dat zijn fundamenteel verschillende dingen.

Neem meetkunde. De stelling dat de hoeken van een Euclidische driehoek samen 180° zijn, bestaat al. Een leerling vindt die stelling niet uit en onderhandelt er ook niet over.

Wat de leerling construeert, is een interne representatie:

· hoe die aansluit bij voorkennis,

· wanneer die van toepassing is,

· waar die niet opgaat,

· en hoe je die kunt gebruiken om te redeneren.

Of neem schaken. De regels worden niet geconstrueerd. Legale zetten worden niet geconstrueerd. Schaakmat wordt niet geconstrueerd. Die dingen worden geleerd via instructie, observatie, boeken, coaching en oefening. Wat de lerende construeert, is een steeds verfijnder mentaal model van het spel.

Hetzelfde onderscheid geldt voor vrijwel elke academische discipline.

De doelbewuste valse tegenstelling

Een van de hardnekkigste mythes in het onderwijs is de aanname dat expliciete instructie op de een of andere manier verhindert dat leren constructief is.

Nou, dat doet het niet.

Leerlingen blijven betekenis construeren terwijl zij naar uitleg luisteren. Zij blijven construeren terwijl zij lezen. Terwijl zij demonstraties observeren. Terwijl zij oefenen. Terwijl zij discussiëren. Terwijl zij feedback krijgen.

Leren blijft altijd constructief, ongeacht de instructiemethode.

De vraag is niet of leerlingen kennis construeren of niet. Dat doen ze altijd.

De echte vraag is:

Onder welke instructiecondities zijn leerlingen het meest geneigd accurate, duurzame en overdraagbare kennis te construeren?

Dat is een empirische vraag. Geen ideologische.

Daarom verwerp ik de gebruikelijke conclusie dat, omdat leren constructief is, onderwijzen dus “constructivistisch” moet zijn. Dat is een valse tegenstelling.

Onderwijzen en leren zijn twee geheel verschillende processen.

Onderwijzen gaat over wat de instructeur doet. Leren gaat over wat er in de lerende gebeurt.

Dit zijn geen concurrerende verklaringen. Ze vullen elkaar aan.

• Onderwijzen levert informatie. Leren organiseert die.

• Onderwijzen geeft uitleg. Leren construeert betekenis.

• Onderwijzen biedt uitgewerkte voorbeelden. Leren bouwt schema’s op.

• Onderwijzen geeft feedback. Leren reorganiseert het geheugen.

Leerlingen kunnen niet leren zonder actief begrip te construeren. Maar daaruit volgt niet dat leraren expliciete instructie “zouden moeten” vermijden.

Constructie is een noodzakelijke eigenschap van leren. Op zichzelf is zij geen voorschrift voor onderwijzen.

Conclusie

Scholen bestaan omdat een groot deel van onze cultureel opgebouwde, biologisch secundaire kennis waarschijnlijk niet efficiënt of accuraat wordt verworven als leraren niet onderwijzen (en alleen maar faciliteren).

Leren is constructief. Onderwijzen is instructief. Deze twee uitspraken zijn geen tegenstanders. Zij beschrijven verschillende delen van hetzelfde causale proces.

Om leerlingen te laten leren, moeten leraren onderwijzen.

Samenvatting van mijn kernargument:

De ambiguïteit rond “constructivisme” is niet dat leren constructief is. De ambiguïteit is dat onderwijsgevenden stilletjes wisselen tussen verschillende betekenissen van “kennis construeren” en tussen “theorieën van leren” en “voorschriften voor onderwijzen”.

Deze blog was oorspronkelijk geschreven door Umes Sheshta en was gepubliceerd op zijn Substack.
ChatGPT heeft het vertaald.

Subscribe now

Share

Leave a comment

Enkele referenties:

Ausubel, D. P. (1968). Facilitating meaningful verbal learning in the classroom. The Arithmetic Teacher, 15(2), 126-132.

Bransford, J. D., Brown, A. L., & Cocking, R. R. (2000). How people learn (Vol. 11). Washington, DC: National academy press.

Clements, D. H. (1997). In my opinion:(Mis?) Constructing constructivism. Teaching children mathematics, 4(4), 198-200.

Clark, R. E., Kirschner, P. A., & Sweller, J. (2012). Putting students on the path to learning: The case for fully guided instruction. American educator, 36(1), 6-11.

Cobb, P. (1988). The tension between theories of learning and instruction in mathematics education. Educational psychologist, 23(2), 87-103.

Fisher, J. (2024). Long live the transmission model. Text Savvy.

Geary, D. C. (2008). An evolutionarily informed education science. Educational psychologist, 43(4), 179-195.

Kirschner, P. A., & Hendrick, C. (2024). How learning happens: Seminal works in educational psychology and what they mean in practice. Routledge.

Mayer, R. E. (2004). Should there be a three-strikes rule against pure discovery learning?. American psychologist, 59(1), 14.

Rosenshine, B. (2012). Principles of instruction: Research-based strategies that all teachers should know. American educator, 36(1), 12.

Translated from Dutch by ChatGPT, so…

Lately, I have been seeing more and more articles about arithmetic and mathematics anxiety. Why? What is going on? Well, there is a very persistent misunderstanding about mathematics anxiety, namely that it is mainly something experienced by students who are “just not good at arithmetic or mathematics”. Those who find mathematics difficult become anxious, and those who are anxious are apparently less mathematically gifted. Done. Child’s play. But it’s not that simple. A recent meta-analysis by Martin Brunner and colleagues, based on representative data from 2.5 million (!) people across 90 countries, shows that mathematics anxiety and mathematics achievement are, on average, negatively related, but also that this relationship is reciprocal. Put simply: anxiety can depress performance, and poor performance can feed anxiety. See this two-way correlation both as a warning light and a signpost.

That reciprocity matters for teachers. In the classroom, we often see only a snapshot: a student who freezes when fractions appear, hands in a blank test, or starts sighing - or worse - at the word algebra. The temptation is then great to slap a label on such a student: insecure, unmotivated, no aptitude, weak... But arithmetic or mathematics anxiety is not a character trait that is almost impossible to change, nor is it a fixed measure of aptitude. The anxiety arises in situations such as an unexpected turn to answer, a timed problem, a test, or sometimes merely the sight of a page full of symbols.

So what happens cognitively? Anxiety takes up space in working memory. The student not only has to solve the problem but also has to suppress thoughts such as: “I can’t do this”, “In a moment everyone will see that I’m stupid”, or “I’m going to fail again”. These thoughts compete for the same limited space needed to hold intermediate steps, choose procedures and retrieve knowledge from long-term memory. So a child may know more than he or she can show at that moment. Conversely, repeated experiences of failure can seem to prove that the anxiety was justified. It becomes a “self-fulfilling prophecy”.

That is why it is too simple to say: “First we have to increase the child’s confidence or create a positive mindset, then arithmetic or mathematics will come by itself.” Equally simple and wrong is: “Just practise more, then the anxiety will disappear.” The art is to do both at the same time: to build competence in a way that organises safety and experiences of success (see Rosenshine: small steps and letting students experience success). Not by making arithmetic or mathematics smaller, but by designing better access to it.

Retrieval practice, or self-/practice testing, can help here. I wrote a column about test anxiety in 2020. Self-testing NOT as a disguised test culture or as a weekly reckoning, but as a low-threshold exercise in retrieving what has already been learned. Start the lesson, for example, with three short questions on previous material. Let students first try individually, then compare, and end with good feedback. Use mini-whiteboards, mini-quizzes, exit tickets or short digital questions. The goal is not to judge, but to make available what has already been learned.

Why might this reduce anxiety? First, retrieval strengthens memory (the testing effect). What is retrieved more often becomes more readily available. As a result, students have to spend less working memory on basic steps and have more space left for reasoning. Second, making what was learned earlier available gives students a more realistic view of what they do and do not yet master. Anxiety often lives off vagueness: “I don’t understand any of this” becomes more manageable when it turns out to be: “I mainly forget step two in equations”. Third, it normalises making mistakes (to err is human and everyone makes mistakes), provided the teacher clearly positions the exercise as learning, not assessment. And finally, experiencing success. A sense of “I can do this” reduces anxiety.

But the form is crucial. Practice tests work better when they are predictable, short, frequent and have hardly any consequences, or none at all (in fact, no-stakes). Build up from simple to more complex questions. Give thinking time. Let students improve their answers. Make success visible. And always connect retrieval to explanation: “Why is this approach useful?” “Which mistake is likely here?” “How do you recognise this type of problem?”

Arithmetic or mathematics anxiety calls for didactic precision. The message has to be double: yes, your anxiety is real, but there is no evidence that you cannot learn this. For teachers, that is a professional task. We cannot talk the emotion away, but we can design lessons in which knowledge is successfully retrieved more often, mistakes become information, and competence grows. Because in the end, mathematics is not only about being able to calculate. It is also about daring to think.

Subscribe now

Share

Leave a comment

Brunner, M., Preckel, F., Götz, T., Lüdtke, O., & Keller, L. (2026). High math anxiety is associated with lower math achievement across 90 countries: An individual participant data meta-analysis of representative student and adult samples. Psychological Bulletin, 152(2), 207-253.

There’s value in this argument. Selwyn is right to be sceptical about commercial claims that a chair can cause learning, improve cognition, encourage collaboration, or optimise the brain. Schools should be wary when furniture companies dress up marketing in the language of learning science, neuroscience, active learning, and 21st-century learning. A chair may support certain classroom routines, but it isn’t a pedagogy. It can’t replace curriculum, instruction, explanation, practice, feedback, or classroom management.

The article becomes less convincing when it treats the very act of shaping pupil behaviour as inherently suspicious, or even wrong. Selwyn uses the language of “coercion” to describe furniture that encourages pupils to sit, face forward, stay in place, or work within certain limits. But in schooling, not all constraint is bad. In fact, some constraint is necessary. A classroom isn’t simply a space for individual expression. It’s a shared instructional environment in which one child’s freedom to wander, interrupt, fidget loudly, talk over others, or refuse to participate can reduce the rest of the class’s opportunity to learn.

This is where the article’s progressive pedagogical assumptions become clearly visible. Much progressive and student-centred rhetoric starts from the idea that children should have greater control over what, when, where, and how they learn (i.e., be student led). In its stronger forms, this becomes the belief that adult – read teacher - direction is oppressive, that pupil choice is inherently liberating and positive, and that any attempt to require attention, effort, silence, posture, or participation is a form of domination. Selwyn doesn’t put the argument quite so crudely, but his strongly pejorative use of “coercion” risks moving in that direction.

The problem is that children, especially novices, don’t usually know what’s best for their own learning. They may know what they prefer or what feels comfortable, enjoyable, or immediately rewarding. But preference isn’t the same as what’s good for learning. A pupil may prefer to talk, move around, browse, avoid difficult work, or choose an easier task. But that doesn’t mean these choices are educationally wise just as the choice for salty, sweet, or fatty foods is nutritionally wise. As I, along with many other colleagues, have argued, learners are often poor judges of what supports durable learning, especially when they lack the background knowledge needed to understand what the task requires or what good study strategies are.

This matters because learning often requires doing things that aren’t immediately attractive: listening carefully, reading demanding text, practising a skill, memorising facts, revising errors, waiting one’s turn, attending to an explanation, or persisting when a task feels difficult. None of this is “coercion” in any sinister sense; it’s education. A teacher’s job isn’t merely to facilitate whatever pupils already want to do. It’s to introduce them to knowledge, practices, disciplines, and standards they wouldn’t necessarily encounter or choose on their own.

The same issue appears in the language of “21st-century skills.” Selwyn rightly tells how furniture companies often use fashionable terms such as collaboration, creativity, active learning, learner-centredness, and flexibility. But he doesn’t sufficiently challenge the progressive mythology behind some of this language. My criticism of “21st-century skills” is relevant here: many so-called 21st-century skills aren’t new, aren’t free-floating, and can’t be taught apart from knowledge. Critical thinking, creativity, problem solving, and collaboration depend heavily on what pupils know. You can’t think critically about history without historical knowledge, solve scientific problems without scientific knowledge, or collaborate productively on a task you don’t understand.

From this perspective, the real problem with some innovative classroom furniture isn’t that it “coerces” pupils into order. The deeper problem is that it’s often sold as if flexible furniture, movement, collaboration, or choice is good or that it can substitute for serious thinking about curriculum and instruction. A room full of mobile chairs doesn’t produce collaboration. A stool that lets pupils swivel doesn’t produce attention. A “learning zone” doesn’t produce learning. These objects may help or hinder depending on how they’re used, but they don’t remove the need for explicit instruction, subject knowledge, guided and independent practice, feedback, and a well-managed classroom with well-polished routines.

Selwyn is right to criticise design solutionism, but his critique would be stronger if he distinguished bad coercion from legitimate educational structure. Schools shouldn’t humiliate pupils, physically restrain them, suppress their individuality, or design environments that are uncomfortable, exclusionary, or hostile. But it’s perfectly reasonable for schools to arrange rooms in ways that discourage interruption, protect attention, and make teaching possible. A classroom needs conditions in which pupils can hear the teacher, follow an explanation, concentrate on a text, complete written work, and listen to one another. These conditions don’t emerge automatically from freedom. They require routines, norms, seating plans, and adult authority.

This point is especially important for equity. In “permissive” classrooms, the costs aren’t distributed equally. The most confident, impulsive, socially dominant, or already knowledgeable pupils may cope, or even thrive. The quieter pupil, the anxious pupil, the pupil with weaker background knowledge, the pupil who needs calm conditions, or the child with special needs will lose out in such an environment. A teacher who insists on order isn’t necessarily being authoritarian. They may be protecting the learning rights of the whole class. The freedom of one child to interrupt can become a coercive force imposed on everyone else.

There’s also a difference between concentration and passivity. Progressive critiques often treat sitting, listening, and facing the teacher as signs of outdated schooling. But listening to an expert explanation can be intellectually active. Reading silently can be demanding. Practising a procedure can build fluency. Writing an answer independently can require effortful thought. Conversely, as Robert Coe noted, a classroom full of movement, talk, choice, and collaboration can appear active while producing little or no learning. The visible appearance of activity shouldn’t be confused with cognitive engagement.

Selwyn’s article is, to a certain extent, insightful, but it also suffers from overreach. Its insight is that classroom objects embody assumptions and that commercial “learning space” rhetoric deserves scrutiny. Its overreach is the tendency to treat the shaping of pupil behaviour mainly as a problem of power, rather than also as a condition for teaching and learning. Schools do indeed direct children. They require attention, effort, and participation. But the important question isn’t whether this is coercive in some abstract philosophical sense. The important question is whether the direction is educationally justified, proportionate, humane, inclusive, and effective.

The better critique of modern classroom furniture isn’t that it constrains pupils, because all classroom designs compel pupils in some way. An open-plan learning space limits most children's ability to concentrate on what they’re doing and to listen to the teacher. A better critique is that schools should ask what kind of constraint is being created, and for what purpose. Does the furniture help pupils attend, think, read, write, practise, and participate? Does it reduce unnecessary distraction? Does it support the teacher’s ability to teach clearly? Does it help pupils with different needs access the lesson? Or does it merely signal fashionable innovation while making sustained attention and learning harder?

In conclusion, Selwyn’s article is useful because it asks teachers and school leaders to look more critically at the material design of classrooms. It rightly warns against exaggerated claims made by educational furniture companies. But its suspicion of “coercion” risks echoing a permissive progressive idea: that children’s freedom from adult direction is inherently educationally good. That idea isn’t supported by what we know about novices (basically all pupils), attention, cognitive load, classroom management, or the role of knowledge and concentration in learning.

A more reasonable conclusion is this: schools shouldn’t be ashamed of creating conditions for concentration. They should do so intelligently, humanely, and with attention to pupils’ comfort and dignity. But classrooms are places for learning, and learning often requires structure. Furniture should be judged not by whether it maximises pupil choice or avoids all constraint, but by whether it helps teachers teach and pupils learn.

[1] Langdon Winner’s classic article Do Artifacts Have Politics? (Daedalus, 1980) discusses Robert Moses’s low overpasses on Long Island parkways. Winner argued that these overpasses were built too low for buses to pass underneath, which meant that people who depended on public transport – read disproportionately poorer and Black New Yorkers - had much less access to public places like Jones Beach. In that sense, the bridges weren’t politically neutral pieces of infrastructure; they embodied social and political choices in concrete and steel.

If you want to invite me to talk to you and your fellow teachers about what good learning and instruction is, please contact me.

Subscribe now

Share

Leave a comment

She drew up a table that I unfortunately don’t know how to format in Substack, so here as text, based on what published guidelines (see for example here, here and here) advise on what is important in classrooms for children with trauma experiences and how education based on cognitive science can help:

What trauma-affected children need: How explicit teaching informed by learning science/cognitive load theory addresses this need

A reliable sense that the classroom is a safe, supportive environment in which they do not need to be hyper-vigilant to threat (perceived or real):
Learning spaces should be stand-alone classrooms, not open-plan environments with inadequate sound dampening and constant visual distraction from other classrooms.
Children should be seated in rows, facing the teacher when they are being taught new knowledge and skills. Social seating is for social activities. It lends itself to distraction of all forms.
Environmental visual distractions (colourful posters, banners, bunting, etc.) should be minimised and placed at the back of the room.

A sense of safety and trust that adults in the room are “in charge” of learning and behaviour:
Classrooms should be calm, orderly and predictable learning spaces, as a reflection of taught behaviour expectations and adult-led limit-setting with clear, consistent and fair consequences.
Positive behaviours are noted and praised.
Instruction should be educator-led. Educators take responsibility for the teaching of new content. This is not left to chance via discovery learning.

A sense that, as far as possible, they are not going to be caught off-guard by shifts in tasks and activities:
Classroom routines should be predictable within and between teaching spaces.
Educators use known attention signals, e.g., a “clap-back”, a bell, or a key phrase, such as “one two three, eyes on me” to gain students’ attention.

A sense that the teacher expects them to learn:
High expectations of and for all learners, regardless of starting point, are a driving force for all instruction and support.

Optimal and ethical use of their learning time:
Device use is minimal to non-existent. Except in timetabled IT classes, technology has little to no intrinsic learning benefit in classrooms.
Pseudoscientific tools and practices are not in use (e.g., Learning Styles / Multiple Intelligences Inventories; Brain Gym; Whole Language/Balanced Literacy approaches to reading instruction; discovery learning for maths).
Well-meaning but potentially harmful whole-school mental health interventions are not used.
Instructional time is valued, respected, and optimised. Learning objectives are known and checked against.
Busy work is not accepted as a substitute for learning.

A sense of self-efficacy when encountering and consolidating new knowledge and skills:
New tasks are broken into manageable “chunks” so as not to overload the novice learner. Instruction is educator-led, via the gradual release of responsibility (I do; we do; you do).
Spaced retrieval practice and daily review provide opportunities for students to consolidate new knowledge and skills into long-term memory.

A sense of safety in situations that involve transitions:
Transitions into and out of the classroom are quiet, predictable and calm. Children are taught how to collect their belongings, line up, and move in and out of teaching spaces. In some schools, this is done in silence, to reduce disruptions associated with noise and minimise surreptitious bullying.

A sense of self-efficacy for contributing to classroom activities and discourse:
Mini-whiteboards are used for formative assessment and to trigger re-teaching where indicated. Every child is expected and supported to participate, with or without accommodations and additional support.
Turn-and-talk activities are routinely practised and used, with careful thought given to seating arrangements and student pairing.
Non-volunteers are called on (so-called “cold-calling” – better described as “warm calling”) to respond to carefully framed questions about material that has been taught. This ensures that all children participate in classroom learning, with adjustments and accommodations if needed, to support this.

Subscribe now

Share

Screens are wonderful things. I’m writing this on a laptop with one. You’re probably - as this is a blog published online - reading this on one. Screens give us access to more text, more quickly, than anything we could’ve imagined just a generation ago. But in education, “more quickly” isn’t always a virtue. Actually. sometimes it’s the problem.

A recent study by Angelica Ronconi and Lucia Mason, wonderfully titled Too Fast, Too Confident: How shallow processing and miscalibrated understanding undermine digital reading in adolescent learners, looked at how seventh- and eighth-grade students read expository science texts in print (one on earthquakes and one on volcanoes) and on screen.

Students in a within-participants design[1] read one text on paper and one digitally. The researchers measured how much time they spent reading, how much cognitive load they reported as having experienced, how confident they were that they’d understood the text, and how well they actually understood it. The headline finding is simple and uncomfortable: When students read digitally, they read faster, became more overconfident, and understood less.

This isn’t a small educational detail. It’s a flashing RED warning light. The problem isn’t that students can’t read on screens. They, and we, can and do it every day and very often. The problem is that the screen seems to invite a different way of reading. Less “I need to understand this” and more “I’ve seen this, I recognise this, so I know it” (the illusion of competence[2]). Ronconi and Mason connect this to what they call the shallowing hypothesis: “habitual exposure to digital media reshapes reading behaviors in ways that promote surface-level processing over deep comprehension” (p. 3). Digital environments often encourage skimming, scanning, jumping, and fast movement through information rather than sustained, reflective processing. In their study, digital reading impaired comprehension through a sequence: less time spent reading led to poorer monitoring of understanding, which led to lower comprehension.

This is where Salomon’s classic article, Television is “easy” and print is “tough”, becomes highly relevant. Salomon wasn’t studying tablets or laptops. The article was published in 1984 and replicated by Hans Beentjes in 1989. But what he was studying is something that hasn’t gone away, namely how learners’ perceptions of a medium influence how much mental effort they invest. He introduced the idea of Amount of Invested Mental Effort, which he defined as the number of nonautomatic mental elaborations learners apply to material. In other words: not how much effort the task objectively demands, but how much effort the learner actually chooses to put in. And that distinction matters enormously.

Cognitive load theory reminds us that working memory is limited. If we overload learners with unnecessary demands by introducing poor layout, spatial or temporal split attention, irrelevant information, scrolling, distractions (seductive details), then learning suffers. But Salomon reminds us that there’s also an opposite danger, namely that learners may underinvest mental effort when a medium feels easy, familiar, entertaining, or low stakes. In his study, children felt more competent with television and perceived it as easier and more realistic, while print was perceived as demanding more effort; yet in the end, print led to better learning and better inference making.

And that’s the rub. We often think of effort as a problem: too much effort, too much load, too much cognitive demand. And yes, unnecessary load is a problem. But effort is also the price of learning. Not all effort, as we’ve learned from Elizabeth and Robert Bjork’s work on desirable difficulties is bad. The wrong effort is bad. Wasted effort, such as effort caused by poor design or dysfunctional pedagogies/instructional interventions like discovery or productive failure, is bad. But the effort of thinking, connecting, elaborating, checking, rereading, and making inferences aren’t bugs or a problems. That’s the whole darn point of studying and are prerequisites to learning!

Salomon’s article helps us see digital reading differently. The issue isn’t that screens impose more cognitive load. Ronconi and Mason didn’t find an overall difference in students’ reported cognitive load between paper and screen. The problem is more subtle: digital reading feels easier, and because it feels easier, students invest less effort reading it. They glide through the text, don’t pause, don’t reread, don’t test themselves, and don’t notice confusion. They mistake fluency for understanding; metacognition gone wrong.

Metacognition is often described as “thinking about thinking,” but that phrase has been lost all meaning by simplification and overuse. In reading, it means something very concrete. It means asking and answering questions like: What have I actually read? Do I know whether I’ve understood what I read? Do I notice when I’m confused? Do I slow down when the text becomes difficult? Do I reread and think? Do I check whether I can explain the idea I just read without looking? Or do I just continue to get it over with?

Good readers monitor themselves while reading. They don’t just move their eyes across words, lines, and pages. They think about what they’re reading, build mental models, compare new information with what they already know, enrich or adapt their mental models, notice gaps in their knowledge and try to fill them, repair misunderstandings... Poorer readers, or in this case, readers in conditions that encourage poor habits such as on-screen reading, often don’t do these things. They confuse “I got through it” with “I got it.”

And that’s what makes Ronconi and Mason’s findings so worrying. Students reading digitally weren’t only understanding less; they were also more confident that they’d understood what they had read. They suffer from a variation of the Dunning-Kruger effect[3] and can be seen as the educational equivalent of driving faster in fog while believing visibility has improved.

Gabi Salomon gives us a useful way to explain why this might happen. In his view, learners carry assumptions about media. Television is “easy”; print is “tough.” Today, for many students, screens may be the new “easy” medium. Screens are associated with scrolling, chatting, searching, watching, gaming, swiping, and grazing. Paper, by contrast, still carries a school-like seriousness. It says: sit down, slow down, this matters.

Of course, paper doesn’t magically create understanding. A badly written text on paper is still a badly written text. A student can skim paper too. And a well-designed digital text can support learning. But while the medium may not be the message (in contrast what Marshall McCuhan wrote[4]) the medium does send signals. It cues habits and changes the learner’s expectations about what kind of mental activity is required. And this is where the Ronconi and Mason’s conclusion is both right and incomplete or even wrong. They argue that educational interventions should promote deeper engagement with digital texts and support students’ metacognitive monitoring at key points during reading. In other words: Design digital environments better and teach students to read better. That’s perfectly reasonable. If students are going to read digital texts, they need to be taught to pause, question, reread, summarise, check understanding, and resist the illusion of fluency. They must have environments that don’t stimulate surface processing.

But this is where they lose me. Why is that the first solution provided? If students read digital texts too fast, become too confident, and understand less, one solution is indeed to train them to read digital texts better and give them better texts. Fine. But another, perhaps better, solution is to ask: Why are they reading this text digitally in the first place?

This sounds almost too simple, and that’s why it’ll probably be ignored. In education, we have a habit of creating problems and then designing interventions to solve the problems we’ve created.

Students don’t concentrate well with laptops open? Let’s design attention-support tools. Students don’t monitor their digital reading well? Let’s develop metacognitive prompts. Students skim online texts? Let’s create digital anti-skimming strategies.

Or [Trigger warning: Here comes a radical thought from a grumpy old man], we could give them the text on paper. Not always. Not for everything. But when the goal is deep comprehension of a longer, complex, expository text, paper may simply be the better tool. That’s not nostalgia. It’s not technophobia nor Luddism. It’s matching the medium to the learning goal. The article itself notes that print appears to be the more reliable medium for supporting deep understanding, especially with complex expository texts.

Salomon would probably push us to ask not only “Which medium presents the information?” but also “What does the learner think this medium asks of them?” If students see digital text as quick, light, and easy, then we shouldn’t be surprised when they treat it that way. And if they treat it that way, we shouldn’t be surprised when comprehension suffers.

This doesn’t mean schools should ban digital reading. Students do need to learn to read digitally. They need to handle online sources, PDFs, search results, hyperlinks, databases, and more. But “students need digital literacy” isn’t an argument for making reading digital. Students also need to learn to speak in public and to work in teams, but that doesn’t mean every lesson should be a speech or require students to give one, or that they only work in collaborative teams. The question we should always ask is: What are we trying to achieve with our teaching?

For quick access, digital is marvellous. For search, distribution, accessibility, collaboration, and certain forms of annotation, it can be powerful. But for sustained reading and understanding of complex material, especially for children and young adolescents still developing the metacognitive skills that they need for independent study, print may reduce unnecessary problems and invite more appropriate effort.

So, yes, teach students how to read digitally. Teach them that fast isn’t the same as good, and fluent isn’t the same as understood. Teach them to pause, reread, question, summarise, and check. But also, when the task is to read a demanding text carefully and learn from it, don’t be afraid to do the simplest thing.

Print it.

Subscribe now

Share

Leave a comment

Beentjes, J. W. J. (1989). Learning from television and books: A Dutch replication study based on Solomon’s model. ETR&D, 37, 47–58. https://doi.org/10.1007/BF02298289

Bjork, E. L., & Bjork, R. A. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. In M. A. Gernsbacher et al. (Eds.), Psychology and the Real World: Essays Illustrating Fundamental Contributions to Society (pp. 56–64). Worth Publishers.

Ronconi, A., & Mason, L. (2026). “Too fast, too confident”: How shallow processing and miscalibrated understanding undermine digital reading in adolescent learners. Journal of Educational Psychology. Advance online publication. https://dx.doi.org/10.1037/edu0001013

Salomon, G. (1984). Television is “easy” and print is “tough”: The differential investment of mental effort in learning as a function of perceptions and attributions. Journal of Educational Psychology, 76(4), 647–658. https://doi.org/10.1037/0022-0663.76.4.647

[1] An experimental method where every participant is exposed to all conditions or treatments in a study. Instead of comparing different groups against each other, researchers test the same individuals multiple times

[2] The cognitive bias where you mistake familiarity for mastery. It happens when you look at information—such as a textbook or study notes—and recognize it, leading you to falsely believe you have fully learned and retained the concept when you cannot actually recall it on your own.

[3] A cognitive bias where people with limited knowledge or competence in a specific area greatly overestimate their own skills

[4] Marshall McLuhan argued that a medium’s real power lies less in the content it carries than in how the medium itself reshapes our perception, habits, social relations, and forms of thought.

Leave a comment

Share

Subscribe now

A notification on your phone, tablet, or laptop is not a trivial little thing. A recent article by Hippolyte Fournier and colleagues, with the almost clickbait title Attention Hijacked: How Social Media Notifications Disrupt Cognitive Processing, shows that social media notifications do not merely interrupt attention for a moment, but also measurably slow down the cognitive processing of the reader or recipient. In the experiment, this slowing lasted, on average, for about seven seconds. That doesn’t mean that participants looked at their phones for seven seconds. They continued performing the task but did it more slowly. And that’s exactly what matters in education: you can lose attention without anyone visibly dropping out.

In simple tasks, such a disruption shows up as a small delay. In complex tasks, recovery can take much longer. Someone who loses their place while reading, writing, calculating, or solving a problem must not only return to the task, but also return to the train of thought. The first is called a reinstatement lag: the time needed to find where you were in a text or video. The second is called a resumption lag: the time and mental effort needed to reconstruct what you were doing. A notification is therefore not just the moment something lights up, but is also the cognitive aftermath that follows the interruption.

Some of you might now be thinking: Get real, Paul. It’s just seven seconds. What’s seven seconds? Yes, it sounds trivial until you understand what it means in a classroom. There, it’s not about just one student and one notification. It’s about 25 students, multiple phones, multiple apps, and a stream of micro-interruptions. During reading, arithmetic, listening, or independent work, that’s not a little bit of distraction, but a learning environment where attention is constantly being reclaimed - or even hijacked. And yes, hijacked is an appropriate word here. The authors didn’t call their article Attention Hijacked for nothing.

Stroop Task: A neuropsychological test where participants must name the ink colour of a word while ignoring the word’s text (e.g., say “red” when seeing the word “BLUE” in red ink). This creates interference when they are incongruent, slowing response times when the word and colour don’t match.

What makes this study strong is that the researchers didn’t settle for an artificial beep in a sterile laboratory. They had 180 students perform a Stroop task while iPhone-like pop-up notifications appeared. These notifications looked like real social media messages: recognisable app icons, recognisable shape, recognisable movement. In the main condition, participants even thought the notifications came from their own phones. In two control conditions, participants saw either realistic or blurred notifications, allowing the researchers to examine more precisely what was attracting attention.

The conclusion: notifications draw attention through three different routes at once. First, through salience: something suddenly appears in your visual field and stands out, pulling attention towards it. Second, through conditioning: your brain has learned that notifications can often bring some kind of reward - even if it is only a small endorphin reward kick - and therefore you have to pay attention to them. Third, through relevance appraisal: this message might be important to me. That last route makes the disruption stronger, but it isn’t necessary. The mere shape and recognisability of a notification can be enough to shift attention. That’s not a character weakness in students; that’s cognitive science.

The researchers mainly observed a delay in reaction time, not primarily an explosion of errors. That’s’an important message for teachers. This kind of disruption often doesn’t immediately show up as visibly inappropriate behaviour or as wrong answers from your students. A student still appears to be participating, perhaps even looking at the book. screen, or board, but processes what they were doing more slowly, switches less smoothly, and loses the mental thread more easily. The damage is often not visible as chaos, but invisible loss of processing capacity. And we have far less of that than we like to think. If you’ve read my blogs or books, you already know that the capacity of our working memory is very limited, both in capacity (four to seven chunks of new information and then it is full) and storage time (three to twenty seconds without processing and then, poof, it’s gone).

This is exactly where the debate about smartphones and social media in education often goes wrong. We talk endlessly about screen time. How many hours per day? How long online? As if duration tells the whole story. This study shows something that’s at least as interesting for schools and educators: not only screen time matters, but also the number and frequency of interruptions. The size of the disruption was related to notification volume and checking frequency, not simply to total time on the device. In other words, the question is not only how much students use their smartphones, but also how often their attention is tapped on the shoulder.

That point becomes even more concrete when we look at the additional data from the study. The participating students spent, on average, almost seven hours per day on their phones, received more than 150 notifications per day, and checked their phones more than 100 times per day. These are no longer occasional interruptions. This is a constant rhythm of interruption. And someone who constantly learns to interrupt their own attention turns fragmented attention into a habit.

Of course, this was not a direct classroom study in the wild. The participants were students, the task was a Stroop task, and the design was experimental. But that is precisely why the outcome is interesting. Even in a relatively simple, clearly bounded task, the disruption was visible. In education, students usually work on tasks that demand far more from their working memory: reading comprehension, reasoning, following explanations, solving problems, writing texts, learning procedures. If attention keeps leaking away there, you don’t just lose seconds; you lose coherence.

That’s also why the argument ‘but it was not my phone or my notification’ isn’t sufficient. The study shows that non-personal or less personal notifications can also disrupt attention. Personal relevance makes it worse, but it isn’t a prerequisite. A classmate’s phone lighting up, an app icon appearing, a vibration on the desk: all of these can activate the learned expectation that something’s happening that deserves attention. The studies by Fried (2008), Sana et al. (2013), and Hall et al. (2020) all show that off-task laptop or computer use can harm not only the student doing it, but also nearby peers; visible off-task screen behaviour imposes a shared attentional cost in the classroom. In a classroom, both attention and distraction are therefore not only individual, but also contagious. This effect, known as the attention contagion, holds that the spread of attentive or inattentive states occurs among members of a group. What matters here is understanding that attention is not a purely individual state. Just like yawning, attention in a classroom is social: students can strengthen one another’s attention, but they can also spread one another’s inattention. That makes off-task screen behaviour not merely a private problem for the student behind the screen, but a risk to the attention of the entire learning environment.

For schools, this doesn’t lead to a very complicated policy conclusion: Fewer interruptions during learning. How? Notifications off during instruction. Notifications off during independent work. Phones not on the table. Fixed checking moments instead of constant availability. And where possible: Phones home or in a locker. Not because a school wants to be strict, or because technology is by definition bad, but because learning requires protected, full attention. You can’t learn what you don’t attend to!

This should be explained to students. Not: ‘Put it away because I say so.’ Rather: ‘Put it away because your attention - and your working memory, and that of your neighbours or shoulder partners - is precious and limited, and because notifications target exactly the system you need in order to learn.’ Thinking requires mental effort. Task switching is costly. We CAN’T multitask; we switch between tasks! Multitasking sounds modern, but with these kinds of tasks, it’s usually just rapid back-and-forth task-switching, with loss along the way. A notification, then, isn’t neutral because it competes with the explanation, the text, the problem, or the conversation.

The conclusion for education is fairly simple, but apparently still controversial: anyone who wants students to learn must protect their attention. Not in a totalitarian way. Not hysterically. Just professionally. Notifications are noise, and noise is the enemy of learning.

References

Forrin, N. D., Huynh, A. C., Smith, A. C., Cyr, E. N., McLean, D. B., Siklos-Whillans, J., Risko, E. F., Smilek, D., & MacLeod, C. M. (2021). Attention spreads between students in a learning environment. Journal of Experimental Psychology: Applied, 27(2), 276-291. https://doi.org/10.1037/xap0000341

Fournier, H., Fournel, A., Osiurak, F., Koenig, O., Pâris, F., Gaujoux, V., & Ringeval, F. (2026). Attention hijacked: How social media notifications disrupt cognitive processing. Computers in Human Behavior, 179, 108926. https://doi.org/10.1016/j.chb.2026.108926

Fried, C. B. (2008). In-class laptop use and its effects on student learning. Computers & Education, 50(3), 906-914. https://doi.org/10.1016/j.compedu.2006.09.006

Hall, A. C., Lineweaver, T. T., Hogan, E. E., & O’Brien, S. W. (2020). On or off task: The negative influence of laptops on neighboring students’ learning depends on how they are used. Computers & Education, 153, 103901. https://doi.org/10.1016/j.compedu.2020.103901

Sana, F., Weston, T., & Cepeda, N. J. (2013). Laptop multitasking hinders classroom learning for both users and nearby peers. Computers & Education, 62, 24-31. https://doi.org/10.1016/j.compedu.2012.10.003

Valentina Devid & Samuel Derkse

Note by me, Paul Kirschner: Self-determination theory has always been a problem for me; not what it states, but how it has been co-opted and misused. Many people - from teachers to fellow professors - say that they ae proponents of self-determination theory and apply it in the class. I often have the feeling that they have either never really read what Richard Ryan and Edward Deci wrote or have a problem understanding what the two men meant. Valentina Devid and Samuel Derkse wrote this blog in Dutch, and I immediately thought: They’ve nailed it. That’s why I asked for and got their permission to translate it into English and republish it here.

In this blog, we discuss seven ways in which self-determination theory can and has lethally mutated in schools into something that Deci and Ryan would barely recognise: Seven ‘lethal mutations’ that appear logical but hollow out the theory and undermine its workings.

Since the 1970s, Edward and Richard, better known as Deci & Ryan, have spent decades investigating the question: What motivates people to learn? Their conclusion was as elegant as it was complex: motivation is not primarily driven by external stimuli such as grades, rewards, or punishment, but by the fulfilment of three basic psychological needs¹ namely: autonomy, competence, and relatedness. And more importantly, they don’t operate independently of each another, but in interplay. When people feel they have influence over their own actions (autonomy), feel capable of carrying out a task (competence), and feel connected to others (relatedness), motivation arises that is more durable and qualitatively stronger.²

So much for the good news.

The bad news is that a theory as rich and nuanced as theirs can sometimes become almost unrecognisable in its translation to educational practice. A complex theory gets simplified and often isn’t truly understood. Rich and deep concepts turn into slogans, coherent interrelated ideas become isolated interventions, and the theory itself gradually disappears from view.

Autonomy then gets translated into ‘offering choices’ or ‘letting the learner decide’, competence into ‘positive feedback’, and relatedness into ‘having a good relationship with pupils’. Not necessarily wrong, but clearly a dilution of what the theory is actually trying to describe.³

There is more to it than that. Schools work under time pressure and with the understandable desire to achieve improvement quickly. This urgency not infrequently leads to theories being adjusted until they fit within existing beliefs and routines. Rather than changing one’s own thinking, the theory is unconsciously trimmed and shaped until it comfortably aligns with what was already being done.

And the result is predictable. The original nuance disappears, the coherence and interrelatedness between concepts recede from view and the theory loses its explanatory power. What remains resembles the original theory – the jargon stays intact – but barely functions as such. Practice then works not with the theory itself, but with a simplified and sometimes barely recognisable derivative of it. This is often the fate of complex theories.

Research into educational change has long shown that complex scientific ideas get reduced in practice to simple rules of action. For example, Cognitive Load Theory = minimise cognitive load. The abstraction of scientific knowledge clashes with the urgency of practice, so that important nuances are easily lost in the translation process.⁴ At the same time, people are cognitively inclined to reduce complex ideas to manageable rules of thumb – mental shortcuts that help in making quick decisions.⁵ Implementation research also shows that educational innovations rarely succeed when the complexity of implementation is underestimated and insufficient time is allowed for professional deepening.⁶ The result is that rich theories become unrecognisable in that translation to practice.

There are at least seven ways in which self-determination theory can mutate in schools into something that Deci and Ryan would barely recognise: Seven ‘lethal mutations’ that appear logical, but hollow out the theory and undermine its working.

1. Freedom of Choice = Experiencing Autonomy

How it has become distorted in practice
In schools, autonomy is often translated as offering choices.⁷ Pupils may choose from assignments, teaching methods, pace, or topics. Sometimes, complete choice programmes or free-choice periods are set up. The reasoning is simple: autonomy means choice, so more choice means a greater experience of autonomy.

What’s wrong with it
In the original theory, autonomy refers to the feeling that one is acting voluntarily and can identify with what one is doing.⁸ It is about experienced ownership, not an abundance of choices. A pupil can experience autonomy without any freedom of choice, for instance when they understand why a task is relevant and experience it as such.⁹ Conversely, an abundance of choices can actually undermine the experience of autonomy, especially when those choices are unclear, have little meaning, or are in essence illusory choices.¹⁰

2. Freedom of Choice = Experiencing Autonomy = Increased Motivation

How it has become distorted in practice
Once autonomy has been equated with choice, the next step follows almost automatically: the more choice you offer, the more motivated pupils become. Schools then go in search of ways to continuously expand the range of choices available, some to the point of letting pupils decide themselves what, when, and how they learn.

What’s wrong with it
Choice is only motivating under specific conditions. The options must be clear, meaningful, and aligned with the pupil’s prior knowledge and skills.¹¹ When those conditions aren’t met, precisely the opposite occurs: choice-induced stress and diminished engagement.¹² More choice is therefore no guarantee of more motivation. It’s a variable that only works under certain conditions.¹³

3. Self-Directed Learning for Everyone, Beginners and Advanced Learners Alike

How it has become distorted in practice
In some educational practices, self-directed learning and thus making one’s own choices have become dogma. It is then presented as universally good: for all pupils, in all phases of learning. Even pupils who have barely mastered the subject matter are expected to be able to choose which approach, strategy, or route works best for them. The idea is: whoever takes ownership of their learning, is more motivated and learns better.

What’s wrong with it
Research on cognitive load shows that beginners and advanced learners need fundamentally different forms of guidance.¹⁴ Pupils with little prior knowledge generally benefit more from clear explanation, worked examples, and structured practice. At this stage, they simply do not yet possess the mental model needed to make meaningful choices.¹⁵
A great deal of choice and independence at this stage primarily increases cognitive load: pupils must not only carry out the task, but also constantly decide how to approach it, what to do, what to do next...¹⁶ That requires precisely the expertise that beginners do not yet have. This not only impedes learning but may also jeopardise the three basic needs within self-determination theory itself.
Self-direction or independence can be extremely valuable when the timing is right and it matches the level of mastery.¹⁷ By failing to distinguish between beginners and advanced learners, a principle that works for advanced learners is incorrectly applied to beginners who don’t benefit from it. Indeed, freedom without knowledge isn’t autonomy – it’s getting lost.

4. Structure vs. Autonomy

How it has become distorted in practice
When autonomy is reduced to freedom and choice, a false opposition with structure emerges. Structure is then seen as restrictive, imposed, and teacher-led, and therefore demotivating. As a result, tight lessons with clear instruction are often regarded as problematic. This illustrates a broader confusion between ends and means; what we want students to develop is mistaken for how they should learn it. The same pattern appears in problem-solving: while the goal is for students to become proficient problem solvers, engaging novices in unguided problem-solving is not the most effective way to achieve that goal.

What’s wrong with it
Self-determination theory, in fact, holds that structure is essential for supporting competence. Clear goals, explicit expectations, and targeted feedback help pupils experience that they’re making progress.¹⁸ That sense of progress is in itself a powerful motivator. Structure and autonomy are therefore not mutually exclusive – the opposite of autonomy support is not structure, but a controlling style.¹⁹ Indeed, good structure can actually strengthen the sense of autonomy.²⁰

5. Autonomy Overshadows Competence

How it has become distorted in practice
In school practice, autonomy tends to receive by far the most attention. The other two basic needs within the theory – competence and relatedness – easily fall into the background.²¹ The discussion is almost always about choice and ownership. We have the impression that this is partly culturally determined: in a society in which individual freedom is highly valued, the emphasis falls on autonomy.²² At the same time, there sometimes appears to be projection from an adult perspective: what we ourselves experience as valuable, we automatically assume to be the primary need of pupils too.

What’s wrong with it
The need for competence – the experience of being capable of something and making progress – turns out in many studies to be one of the most powerful motivators.²³ Pupils become motivated when they notice that they understand something, that they’re progressing, and that they’re having experiences of success. By focusing almost exclusively on autonomy, a large part of motivational theory is left unexploited and the coherence is lost.²⁴

6. Motivation Becomes the Starting Point of Education

How it has become distorted in practice
In many schools, an apparently logical sequence emerges: first motivate, then teach. Lessons are therefore designed to make pupils enthusiastic. The idea is that once motivation is present, learning will follow naturally. Motivation thus becomes the starting point of education.

What’s wrong with it
In self-determination theory, motivation is not a separate starting mechanism that first needs to be ‘activated’. Motivation in fact often arises during learning, when pupils understand what they are doing, experience that they are making progress, and have experiences of success.²⁵

By treating motivation as a prerequisite for learning, the theory is implicitly simplified into a linear model: first motivation, then learning. In reality, motivation and learning interact mutually.²⁶ Well-designed instruction can strengthen motivation, precisely because pupils experience understanding and mastery.
In short: motivation is not only a prerequisite for learning, but also a consequence of it.

7. Motivation Is Equated with Learning

How it has become distorted in practice
Once motivation is placed at the centre, a further step often follows: motivated pupils automatically learn more. Engagement, enthusiasm, and activity in the classroom are then seen as signs that good learning is taking place.

What is wrong with it
The relationship between motivation and learning is more complex than this reasoning suggests. Research on autonomy-supportive teaching does show effects on motivation and well-being, but the effects on learning outcomes are smaller and less consistent.²⁷

A lesson can, therefore, feel very motivating without much being learnt in terms of content. In the research literature, such visible signs of engagement are also referred to as poor proxies for learning²⁸: signals that look like learning, but aren’t.

When motivation is seen as evidence that learning is taking place, attention easily shifts from content mastery to experience and sensation. This turns a nuanced motivational theory into a simple – but misleading – assumption: if pupils are enthusiastic, the learning must be going well.

Sound Instructional Design & Careful Implementation

The conclusion of this account is not that self-determination theory is incorrect. On the contrary. The theory offers valuable insights into how motivation arises and how autonomy, competence, and relatedness mutually reinforce one another.²⁹

The problem lies not in the theory itself, but in what often remains of it when it enters educational practice. What begins as a rich motivational theory ends up as a pedagogical potpourri: a collection of well-intentioned slogans about choice, ownership, and motivation.

That pattern is, incidentally, not unique to self-determination theory. Research into educational change has long shown that complex scientific ideas in practice are almost inevitably reduced to simple rules of action. Not out of ill will, but because practice demands speed, manageability, and direct applicability.³⁰ In that translation process, nuance and coherence easily fall out of view.

That is precisely where the problem arises. For when autonomy is translated as unlimited choice, motivation is treated as a prerequisite for learning, and structure is seen as the opposite of autonomy, little remains of the original theory.

The irony is that precisely what the theory tries to explain – durable motivation – thereby becomes harder to achieve. Motivation rarely grows from maximising choice or chasing enthusiasm. It far more often grows from something far less spectacular: understanding, progress, and experiences of success.³¹

The key therefore lies not in the constant expansion of autonomy in the form of freedom of choice, but in sound instructional design. Education in which goals are clear, instruction is thoughtful, pupils receive sufficient structure to experience progress, and the relevance of the subject matter is made visible.

In such learning environments, something remarkable often happens. Pupils notice that they understand something. That they are making progress. That they can do something they could not do before. And it is precisely there, in the midst of the learning process, that what schools are searching so hard for comes into being: motivation.

Not as the engine that starts learning, but as something that arises in continuous interplay with learning and subsequently reinforces it.³²

Notes

¹ Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Plenum Press. — Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 ↑

² Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01 — Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 ↑

³ Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 ↑

⁴ Pfeffer, J., & Sutton, R. I. (2000). The knowing-doing gap: How smart companies turn knowledge into action. Harvard Business School Press. ↑

⁵ Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus and Giroux. ↑

⁶ Fixsen, D. L., Naoom, S. F., Blase, K. A., Friedman, R. M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. University of South Florida, National Implementation Research Network. ↑

⁷ Katz, I., & Assor, A. (2007). When choice motivates and when it does not. Educational Psychology Review, 19(4), 429–442. https://doi.org/10.1007/s10648-006-9027-y — Patall, E. A., Cooper, H., & Robinson, J. C. (2008). The effects of choice on intrinsic motivation and related outcomes: A meta-analysis of research findings. Psychological Bulletin, 134(2), 270–300. https://doi.org/10.1037/0033-2909.134.2.270 ↑

⁸ Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01 — Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 ↑

⁹ Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 — Reeve, J. (2009). Why teachers adopt a controlling motivating style toward students and how they can become more autonomy supportive. Educational Psychologist, 44(3), 159–175. https://doi.org/10.1080/00461520903028990 ↑

¹⁰ Iyengar, S. S., & Lepper, M. R. (1999). Rethinking the value of choice: A cultural perspective on intrinsic motivation. Journal of Personality and Social Psychology, 76(3), 349–366. https://doi.org/10.1037/0022-3514.76.3.349 — Katz, I., & Assor, A. (2007). When choice motivates and when it does not. Educational Psychology Review, 19(4), 429–442. https://doi.org/10.1007/s10648-006-9027-y ↑

¹¹ Patall, E. A., Cooper, H., & Robinson, J. C. (2008). The effects of choice on intrinsic motivation and related outcomes: A meta-analysis of research findings. Psychological Bulletin, 134(2), 270–300. https://doi.org/10.1037/0033-2909.134.2.270 — Katz, I., & Assor, A. (2007). When choice motivates and when it does not. Educational Psychology Review, 19(4), 429–442. https://doi.org/10.1007/s10648-006-9027-y ↑

¹² Iyengar, S. S., & Lepper, M. R. (2000). When choice is demotivating: Can one desire too much of a good thing? Journal of Personality and Social Psychology, 79(6), 995–1006. https://doi.org/10.1037/0022-3514.79.6.995 ↑

¹³ Patall, E. A., Cooper, H., & Robinson, J. C. (2008). The effects of choice on intrinsic motivation and related outcomes: A meta-analysis of research findings. Psychological Bulletin, 134(2), 270–300. https://doi.org/10.1037/0033-2909.134.2.270 ↑

¹⁴ Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1 — Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4 ↑

¹⁵ Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. https://doi.org/10.1207/S15326985EP3801_4 — Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction, 4(4), 295–312. https://doi.org/10.1016/0959-4752(94)90003-5 ↑

¹⁶ Kirschner, P. A., Sweller, J., & Clark, R. E. (2006). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 41(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1 — Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4 ↑

¹⁷ Bjork, R. A., Dunlosky, J., & Kornell, N. (2013). Self-regulated learning: Beliefs, techniques, and illusions. Annual Review of Psychology, 64, 417–444. https://doi.org/10.1146/annurev-psych-113011-143823 — Kalyuga, S., Ayres, P., Chandler, P., & Sweller, J. (2003). The expertise reversal effect. Educational Psychologist, 38(1), 23–31. https://doi.org/10.1207/S15326985EP3801_4 — Patall, E. A., Sylvester, B. J., & Han, C. W. (2014). The role of competence in the effects of choice on motivation. Journal of Experimental Social Psychology, 50, 27–44. https://doi.org/10.1016/j.jesp.2013.09.002 ↑

¹⁸ Vansteenkiste, M., Sierens, E., Goossens, L., Soenens, B., Dochy, F., Mouratidis, A., Aelterman, N., Haerens, L., & Beyers, W. (2012). Identifying configurations of perceived teacher autonomy support and structure: Associations with self-regulated learning, motivation and problem behavior. Learning and Instruction, 22(6), 431–439. https://doi.org/10.1016/j.learninstruc.2012.04.002 — Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 ↑

¹⁹ Vansteenkiste, M. et al. (2012). Identifying configurations of perceived teacher autonomy support and structure: Associations with self-regulated learning, motivation and problem behavior. Learning and Instruction, 22(6), 431–439. https://doi.org/10.1016/j.learninstruc.2012.04.002 — Reeve, J. (2009). Why teachers adopt a controlling motivating style toward students and how they can become more autonomy supportive. Educational Psychologist, 44(3), 159–175. https://doi.org/10.1080/00461520903028990 ↑

²⁰ Jang, H., Reeve, J., & Deci, E. L. (2010). Engaging students in learning activities: It is not autonomy support or structure but autonomy support and structure. Journal of Educational Psychology, 102(3), 588–600. https://doi.org/10.1037/a0019682 — Vansteenkiste, M. et al. (2012). Identifying configurations of perceived teacher autonomy support and structure. Learning and Instruction, 22(6), 431–439. https://doi.org/10.1016/j.learninstruc.2012.04.002 ↑

²¹ Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 — Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological well-being across life’s domains. Canadian Psychology, 49(1), 14–23. https://doi.org/10.1037/0708-5591.49.1.14 ↑

²² Iyengar, S. S., & Lepper, M. R. (1999). Rethinking the value of choice: A cultural perspective on intrinsic motivation. Journal of Personality and Social Psychology, 76(3), 349–366. https://doi.org/10.1037/0022-3514.76.3.349 — Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98(2), 224–253. https://doi.org/10.1037/0033-295X.98.2.224 ↑

²³ Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01 — Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 ↑

²⁴ Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 — Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological well-being across life’s domains. Canadian Psychology, 49(1), 14–23. https://doi.org/10.1037/0708-5591.49.1.14 ↑

²⁵ Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 — Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 ↑

²⁶ Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological well-being across life’s domains. Canadian Psychology, 49(1), 14–23. https://doi.org/10.1037/0708-5591.49.1.14 ↑

²⁷ Stroet, K., Opdenakker, M.-C., & Minnaert, A. (2013). Effects of need supportive teaching on early adolescents’ motivation and engagement: A review of the literature. Educational Research Review, 9, 65–87. https://doi.org/10.1016/j.edurev.2012.11.003 ↑

²⁸ Coe, R., Aloisi, C., Higgins, S., & Major, L. E. (2014). What makes great teaching? Review of the underpinning research. Sutton Trust. https://www.suttontrust.com/wp-content/uploads/2014/10/What-Makes-Great-Teaching-REPORT.pdf ↑

²⁹ Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 — Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01 ↑

³⁰ Pfeffer, J., & Sutton, R. I. (2000). The knowing-doing gap: How smart companies turn knowledge into action. Harvard Business School Press. — Fixsen, D. L., Naoom, S. F., Blase, K. A., Friedman, R. M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. University of South Florida, National Implementation Research Network. ↑

³¹ Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01 — Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 ↑

³² Niemiec, C. P., & Ryan, R. M. (2009). Autonomy, competence, and relatedness in the classroom: Applying self-determination theory to educational practice. Theory and Research in Education, 7(2), 133–144. https://doi.org/10.1177/1477878509104318 — Deci, E. L., & Ryan, R. M. (2008). Facilitating optimal motivation and psychological well-being across life’s domains. Canadian Psychology, 49(1), 14–23. https://doi.org/10.1037/0708-5591.49.1.14 ↑

Subscribe now

Share

Leave a comment

Another day, another apparently simple (old) technique to improve learning: pre-questions. And, to be fair, this one can help. In a large ManyClasses study across 30 classes and 1,571 students (and 33! authors), pre-questions – a technique pioneered by Ernst Rothkopf and Larry Frase at Bell Telephone Labs in the 1960s and ’70s - given before short instructional videos produced an overall improvement in later performance. As a small aside, my first pieces of published research back in 1978 were on the use of pre- and post-questions in videotaped learning materials[1], and I was the not-credited second author in Christiaan Hamaker’s article on adjunct questions in Review of Educational Research.

But the interesting part is not that they worked but rather why they worked. The authors found no evidence that pre-questions made students pay more attention to the lesson. They didn’t watch more of the video after receiving the pre-questions; if anything, video viewing went down with some students not watching the video at all. The authors’ explanation was that the pre-questions seemed to work because they activated a relevant mental model in advance of instruction. In plain English, pupils learned better not because the question made them “try harder,” but because it helped them connect the new material to something already in memory.

And that, of course, is key because if the benefit depends on students already having some relevant prior knowledge, then pre-questions aren’t simply a neat little classroom trick and are not without dangers! The danger is that they may also produce a Matthew effect: the students who already know more are more likely to benefit, while those with less prior knowledge may gain little and even disengage. The paper says this quite directly: pre-questions may create a “rich-get-richer” and an accompanying poor-get-poorer pattern in authentic classrooms.

And this is where a knowledge-rich curriculum matters, though the authors don’t say anything about it. At least this is my humble inference from the research, which I feel follows directly from the findings. If pre-questions only help when students have something relevant already stored in long-term memory, then the way to tame the Matthew effect is not to arm materials such as texts, videos, and audio (e.g., podcasts which I too studied[2] with Anguelina Popova and Rich Joiner) with pre-questions. It is to ensure that ALL pupils possess the background knowledge, vocabulary, concepts, and schemas needed to profit from them in the first place. A knowledge-rich curriculum gives more students something to think with when the question arrives.

So the lesson here isn’t, “Use pre-questions because the research says so.” The message is: learning doesn’t float free from curriculum. It lives or dies by what students already know. And we know that since 1963 (David Ausubel and the Psychology of Meaningful Verbal Learning). And that means that the most powerful equity strategy often isn’t some clever activity or quick-fix, but a curriculum that systematically builds shared knowledge over time. Unfortunately, this is far less fashionable than talk of motivation, engagement, or grit. But it is probably closer to the truth.

[1] Kirschner, P. A. (1978). The effect of adjunct question position, type, and the presence or absence of feedback on learning from a videotaped lesson. D. Brook & P. Rice (eds.), Aspects of Educational Technology XII, Kogan Page.

Van den Brink, H. J., & Kirschner, P. A. (1978). Het effect van ingelaste vragen op het leren van op videoband opgenomen lesmateriaal: Veldexperiment [The effect of adjunct questions on learning van videotaped learning materials: Field experiment]. Registratie, 10(3), 13-18.

Van den Brink, H. J., & Kirschner, P. A. (1978). Het effect van ingelaste vragen op het leren van op videoband opgenomen lesmateriaal: Laboratorium experiment [The effect of adjunct questions on learning van videotaped learning materials: Laboratory experiment]. Registratie, 20(4), 4-9.

[2] Popova, A., Kirschner, P. A., & Joiner, R. (2013). Effects of primer podcasts on stimulating learning from lectures: How do students engage? British Journal of Educational Technology, 45, 330-339. doi:10.1111/bjet.12023

Popova, A., Kirschner, P. A. & Joiner, R. (2014). Enhancing learning from lectures with epistemic primer podcasts activity – a pilot study. International Journal of Learning Technology, 9, 323–337. doi:10.1504/IJLT.2014.067735

Subscribe now

Share

Leave a comment

*******************************

When I talk to educators about research, their most common complaint (by a long shot) is that they are asked to implement new interventions (a curriculum, a pedagogical technique, a software product, whatever), and are offered no reason to do so other than a breezy “all the research supports it.” The phrase is used as a blunt instrument to silence questions. As a scientist I find this infuriating because it abuses what ought to be a serious claim—research backs this—and in so doing devalues research. It’s an ongoing problem (see Jessica & Tim Lahey’s treatment here) that’s long concerned me.

In fact, the phrase “research supports it” invites questions. It implies that we can, in a small way, predict the future. It claims “if we do X, Y will happen.” If I take this medication, my ear infection will go away. If we adopt this new curriculum, kids will be more successful in learning math. Saying “research supports it” implies that you know not only what the intervention is, but you have at least a rough idea of what outcome you expect, the likelihood that it will happen, and when it will happen.

I offer the following list of rights for educators who are asked to change what they are doing in the name of research, whether it’s a mandate handed down from administrator to teacher or from lawmaker to administrator.

1. The right to know what is supposed to improve. What problem is being solved? For example, when I’ve been to schools or districts implementing a one-to-one tablet/laptop policy, I’ve always asked what it’s meant to do. The modal response is a blank look followed by the phrase “we don’t want our kids left behind.” Behind in what? In what way are kids elsewhere with devices zooming ahead?

2. The right to know the means by which improvement will be measured. How will we know things are getting better? If you’re trying to improve students’ understanding of math, for example, are you confident that you have a metric that captures that construct? Are you sure scores on that metric will be comparable in the future to those you’re looking at now? How big an increase will be deemed a success?

3. The right to know the approximate time by which this improvement is expected. A commitment to an intervention shouldn’t be open-ended. At some point we must evaluate how it’s going.

4. The right to know what will be done if the goal is or is not met. Naturally, conditions may change, but let’s have a plan. If we don’t meet our target, will we quit? Keep trying for a while? Tweak it?

5. The right to know what evidence exists that the intervention will work as expected. Is the evidence from actual classrooms or is it laboratory science (plus some guesswork)? If classrooms, were they like ours? In how many classrooms was it tried?

6. The right to have your experience and expertise acknowledged. If the intervention sounds to you and your colleagues like it cannot work, this issue should be addressed in detail, not waved away with the phrase “all the research supports it.” The fact that it sounds fishy to experienced people doesn’t mean it can’t work, but whoever is pitching it should have a deep enough understanding of the mechanisms behind the intervention to be able to say why it sounds fishy, and why that’s not a problem.

This list is not meant to dictate criteria that must be met before an intervention should be tried, but rather what information ought to be on the table. In other words, the information provided in each category need not unequivocally support the intervention for it to be legitimate. For example, I can imagine an administrator admitting that the research support for an intervention is absent, yet mounting a case for why it should be tried anyway.

Subscribe now

Share

Leave a comment

Image generated by ChatGPT

In curriculum discussions, people sometimes speak as though a knowledge-rich curriculum and an information-rich curriculum are more or less the same thing. After all, both contain content, both expose pupils to ideas and texts and topics, and both can appear broad, ambitious, and intellectually serious. But the two aren’t the same, and the distinction/difference is more than semantic. It lies at the heart of what a curriculum is and is meant to do. The purpose of a curriculum isn’t simply to ensure that pupils encounter a great deal of material. It’s to ensure that they learn things that stay with them, that connect to one another, and that make future learning easier, richer, and more meaningful. That is why the difference between knowledge-rich and information-rich matters so much.

Tuscany Harvest by Janet Kruskamp (Public Domain)

An information-rich curriculum may give pupils access to a wide range of material: articles, videos, websites, class discussions, current affairs, projects, and topical issues; a kind of cornucopia or horn of plenty. On the surface, this can look impressive because it gives the appearance of richness, variety, and responsiveness. Pupils may seem constantly engaged with fresh material, and teachers may feel they are opening up the world in all its complexity. But exposure isn’t the same as learning. Pupils can come into contact with a large amount of information without storing very much of it in their long-term memory, and they can move from one interesting topic to the next without building secure understanding of the link between them. An information-rich curriculum can therefore be full in one sense and thin in another: full of stimuli, but thin in terms of what pupils actually come to know, understand, and retain.

Image generated by ChatGPT

A knowledge-rich curriculum, on the other hand, begins from a different premise. It asks not merely what pupils might profitably encounter, but what they should know, remember, and be able to build upon later. That shift sounds small, but it changes everything. Once a school begins to think in those terms, curriculum design becomes less about assembling attractive content and more about selecting and sequencing powerful knowledge. The point isn’t simply that pupils should have access to material, but that the material should be organised in such a way that it becomes part of their intellectual inheritance. A knowledge-rich curriculum is therefore not just a curriculum with a lot in it. It’s a curriculum whose contents have been deliberately chosen and connected – both horizontally and vertically coherent - so that learning accumulates over time.

This matters because thinking doesn’t take place in a vacuum. Reading comprehension depends heavily on background knowledge. Problem solving depends on what one already knows; The forms of analysis or critical thinking that schools rightly value are rooted in domain-specific knowledge rather than in generic skill alone. Pupils don’t become better thinkers by hovering above knowledge, nor by being surrounded by information. They become better thinkers by knowing more, and by knowing more in organised ways. The weakness of an information-rich curriculum isn’t that information has no value, but that information left at the level of passing encounter too often remains fleeting. It may produce discussion in the moment without producing understanding that lasts (i.e., real learning).

To be fair, the attraction of an information-rich curriculum isn’t difficult to understand. It’s often driven by generous and admirable instincts. Schools want pupils to experience breadth, relevance, diversity, and intellectual excitement. They want the curriculum to feel alive rather than inert. All of those concerns are legitimate. No one should be defending a curriculum that is lifeless, mechanical, or indifferent to meaning. The danger here lies in mistaking abundance for education. In a world already saturated with information, the task of the school isn’t simply to mirror that abundance. It’s to select from it, shape it, and order it so that pupils don’t merely experience fragments of the world, but come to understand some of its structure.

This is why the phrase knowledge-rich is sometimes misunderstood. Critics occasionally hear it and imagine a curriculum (over)loaded with factual detail, as though the aim were simply to stuff pupils’ heads with as much content as possible. But a good knowledge-rich curriculum isn’t about indiscriminate accumulation. It’s about intellectual architecture. It’s about making sure that the knowledge pupils meet is worth knowing. It’s about introducing facts and knowledge in an order that makes sense. And it’s about preparing the ground for later learning. Knowledge-rich, in other words, doesn’t mean fact-heavy so much as carefully structured and cumulatively designed. And once we start thinking in those terms, two important concepts become central: vertical coherence and horizontal coherence.

Vertical coherence refers to the way the curriculum builds over time. In a vertically coherent curriculum, what pupils learn this year isn’t simply another set of experiences to be ticked off before moving on to the next. It becomes part of the foundation for what comes later. Knowledge is returned to, elaborated upon, deepened, and connected (think of a spiral curriculum). In history, pupils may first establish broad chronological frameworks before later engaging in more complex questions of causation, interpretation, and significance. In science, early understanding of particles, forces, or cells becomes the basis for more sophisticated explanations in later years. In English, vocabulary, familiarity with genres, and knowledge of literary and cultural reference points gradually expand pupils’ ability to read demanding texts with understanding. Vertical coherence matters because learning is cumulative by nature. You can’t explain the formation of the Himalayas without first understanding what a mountain is and how plate tectonics work. The same holds true in reading: comprehension depends on vocabulary and background knowledge. New knowledge doesn’t float free from old knowledge; it depends on it. As Christine Counsell and Steve Mastin from Opening Worlds say: ‘A knowledge-rich curriculum is a promise you make to future teachers about what their students have learned.’ So a teacher in grade 3 knows exactly what the learners had learned in 1 and 2, can build upon it, and can broaden and deepen it. So it’s not only important for the learner, but it’s also very important for the teacher and the school to have that.

This is one of the major strengths of a knowledge-rich curriculum: it takes that cumulative reality seriously. It recognises that pupils aren’t endlessly ready to begin afresh with whatever topic happens to be interesting at the time. They need a sequence that helps them move from novice understanding to more mature and expert understanding. Without that sequence, the curriculum becomes fragmented. And this is often where an information-rich curriculum runs into problems. It may offer many engaging and worthwhile units, but if those units are chosen primarily for topicality, novelty, fun, or immediate relevance rather than for their place within a broader sequence, pupils can end up with a collection of disconnected impressions rather than a growing body of knowledge. The curriculum may feel rich from week to week while remaining surprisingly thin across years. It may be horizontally busy while vertically weak.

Horizontal coherence, by contrast, concerns the way different parts of the curriculum connect within the same phase of learning. It asks whether pupils’ experiences across subjects and topics reinforce one another, and whether what they learn feels like a meaningful whole rather than a timetable of unrelated episodes. This, too, matters. A pupil studying a historical period in history class may gain from reading literature connected to that period in English or French. Vocabulary introduced in science (e.g., magnetism) may support understanding in geography (north and south poles). Work in religious education, literature, civics, and history may together enrich pupils’ sense of a civilisation, belief system, or cultural inheritance. These kinds of connections can strengthen learning because they help pupils see patterns and relationships rather than isolated facts.

Yet horizontal coherence, important though it is, can sometimes be overemphasised. Schools are often tempted by thematic planning because it creates visible links and gives the impression of integration. A project may feel elegant because several subjects appear to revolve around a common theme, but this can create a false sense of curricular strength if it isn’t accompanied by vertical coherence. A beautifully linked project isn’t necessarily a well-sequenced education. Pupils may enjoy the connections within a project while still failing to build secure knowledge over time. That is why horizontal coherence is best thought of as the connective tissue of the curriculum, not its backbone. It adds meaning and reinforcement, but cannot substitute for the careful sequencing that enables long-term intellectual growth.

The deeper problem with an information-rich curriculum, then, is that it can prioritise breadth of encounter over depth of accumulation. Pupils meet many things, but they don’t necessarily remember many of them. That matters for all pupils, but it matters especially for those who don’t already possess a great deal of background knowledge. When curriculum content is left loose, broad, and weakly sequenced, the pupils most able to benefit are often those who come to school already equipped with the knowledge and language needed to connect the dots. The others – the less privileged - are far more dependent on the curriculum itself to make those connections explicit. For that reason, the case for a knowledge-rich curriculum is also an equity argument. A strong curriculum doesn’t assume that powerful knowledge will be picked up incidentally. It teaches it deliberately, systematically, and in a way that gives all pupils access to it.

Of course, none of this means that every curriculum labelled knowledge-rich is automatically successful. Poor versions certainly exist (see Natalie Wexler’s Substack). A bad knowledge-rich curriculum can become overloaded, rigid, overly canonical, or obsessed with coverage for its own sake. It can confuse remembering with understanding and can reduce education to a checklist of things to be got through (see the different canons[1] that countries and school systems are producing). But that isn’t an argument against knowledge-richness properly understood. It’s an argument against poor curriculum design. A well-constructed knowledge-rich curriculum shouldn’t oppose inquiry, interpretation, creativity, or discussion. It should make them more possible by giving pupils something substantial to think with. Knowledge isn’t the enemy of rich education. It’s what makes rich education possible in the first place.

So the real choice isn’t between knowledge on the one hand and richness on the other. It’s between two different conceptions of richness. One (information richness) treats richness as abundance: many topics, many resources, many encounters. The other (knowledge richness) treats richness as depth, connection, and cumulative understanding. Schools should, ideally, aim for a curriculum that is knowledge-rich in structure and information-rich in texture. In other words, the spine of the curriculum should be formed by carefully selected knowledge, sequenced with strong vertical coherence and supported by sensible horizontal coherence, while the lived experience of that curriculum can still be enriched by excellent texts, examples, discussions, enquiries, and wider experiences. Information has a crucial place, but it should serve the growth of knowledge rather than stand in for it.

That’s why the distinction matters. A curriculum shouldn’t be judged mainly by how much material passes before pupils’ eyes, nor by how many interesting things they briefly touch. It should be judged by what pupils come to know, how securely they know it, how well their knowledge connects, and whether it prepares them for what comes next. A curriculum succeeds not when it’s merely full, but when it’s coherent. It succeeds not when it offers constant novelty, but when it builds understanding that endures. And it succeeds not when pupils are simply immersed in information, but when they leave school with knowledge that has been structured, strengthened, and carried forward.

If you want to learn more about knowledge-richness, please download (open access) or buy Developing Curriculum for Deep Thinking: The Knowledge Revival.

See also our blog Five Common Misconceptions About the Knowledge-Rich Curriculum

Subscribe now

Share

Leave a comment

[1] Curated, authoritative selections of events, texts, or works considered essential to a culture’s identity and understanding of its past, often serving as a “master narrative”.

Back in 1939, Alexander Flexner wrote an article in Harper’s magazine entitled The usefulness of useless knowledge. Flexner is a famous education reformer and the founder and first director of the Institute for Advanced Study in Princeton, New Jersey. Its faculty members include scholars such as Albert Einstein, Kurt Gödel, John von Neumann, and George Kennan. It was striking that the content of the article was still so timely after more than 75 years!

This is the opening paragraph:

After reading the article, there were two things that crept up and kept gnawing. First, governments express a strong urge for applied research and that society (whatever that is) should have a strong say in what is studied. Governments prioritise conversations with students, teachers, parents, schools, and other interested parties – note the absence of educational and learning scientists here - with the idea that these conversations then lead to an innovative curriculum and innovations in core objectives and learning outcomes.

This led in the Netherlands to basing policy on what Dutchies prioritise for funding and execution (there was even a Facebook® vote to determine what research project should be funded in the field of medicine!).

In other words, the content and approach of the Dutch education system as well as scientific research, would be determined through something similar to a political referendum. Imagine what would happen if we were to do the same for the setup of the national soccer team and their strategy! Oh wait, perhaps that already happened as the Dutch are out...

In one of the Dutch newspapers (Financiële Dagblad) it says: The choices made in the National Science Agenda will determine which science proposals will get funding. If knowledge institutions collaborate with commercial and community organisations as well as proposals that have an eye for valorisation (for example, translating scientific knowledge into a next level of products or services) will receive a far better rating. Hence, their chances of getting funding are much higher. According to the Dutch State Secretary for Education the societal “added value” of scientific research must be maximised, which is possible through education, valorisation, and economic applications. In other words, the science and education minister is suggesting that science must be driven by thinking in terms of efficiency and utility.

Flexner starts his article on useless knowledge with an anecdote of a conversation that he had with George Eastman, the founder of Eastman Kodak Company. Flexner asked Kodak who he thought was the most useful scientist in the world. Kodak answered “Guglielmo Marconi”. Flexner’s response to this answer was rather surprising to Kodak. Flexner stated that, although the radio has added immense value to our lives, Marconi’s contribution to this was zero to nothing; a practicality. Flexner explained that all Marconi did was add a final technical detail to what, to that point in time, was pure science. Scientists such as James Maxwell (mathematical calculations on magnetism and electricity) and Heinrich Hertz (theoretical physics research on electromagnetic waves) had laid the foundations for Marconi’s final piece of cake. Maxwell and Hertz did not think “utility” or “application”; they studied ‘stuff’ based on pure curiosity! According to Flexner, “Hertz and Maxwell could invent nothing, but it was their useless theoretical work which was seized upon by a clever technician and which has created new means for communication, utility, and amusement by which men whose merits are relatively slight have obtained fame and earned millions”.

Flexner lists some additional examples in his article and then concludes:

“…most of the really great discoveries which had ultimately proved to be beneficial to mankind had been made by men and women who were driven not by the desire to be useful but merely the desire to satisfy their curiosity…Institutes of learning should be devoted to the cultivation of curiosity and the less they are deflected by considerations of immediacy of application, the more likely they are to contribute…to human welfare…”

In addition to the Dutch crowd sourcing approach to research and their focus on practical application and utility, the other thing that kept gnawing away was an article in the Times Educational Supplement, written by the notorious eduquack Sugata Mitra[1], but not unique to him unfortunately. Albert Einstein was a notorious sufferer of what I call the Expertise Generalisation Syndrome[2] as was Sir Ken Robinson. Mitra wrote:

“...KNOWING IS an obsolete idea from a time when it was not possible to access or acquire knowledge at a moment of need. The idea of knowing assumes that the brain must be “primed” in advance for circumstances that may require knowledge. Just in case.”

He adds: “We don’t need to know until we need to know.” Whoever knows Johan Cruyff...would say it could have been one of his quotes. One famous one from Cruyff to get the idea: “If I wanted you to understand it, I would have explained it better!” Or Yogi Berra: “You’ve got to be very careful if you don’t know where you are going, because you might not get there.”

Let’s go back to Flexner’s argument, namely that so-called useless knowledge is actually the foundation for further, deeper and more creative thinking. Knowledge that at first glance might be perceived as useless is actually indispensable in that it enables us to book technical/technological, economic, and societal progress. Without such useless knowledge, we wouldn’t even be able to recognise problems; let alone think about them or work towards a solution to them. For example, without knowledge we wouldn’t even recognise global warming; and therefore wouldn’t even think about how we as individuals, society, scientists, and politicians could solve the matter.

The ideas expressed by Mitra and others don’t bring us further, but rather move us backwards towards the Middle Ages when – due to a lack of knowledge - people thought that the sun orbited around the earth, that diseases were caused by humours (bodily fluids), that an eclipse of the sun was caused by the Gods and that sacrificing a virgin would bring it back. Luckily, solar eclipses were and are rarities!

Without all of this “useless” knowledge[3], it would be impossible to understand anything of what happens around us, to judge it, or to develop ourselves as individuals and societies. Also, last but not least, without this “useless” knowledge, we wouldn’t be able to intellectually communicate about anything. The consequences would be devastating: stagnation and decline.

Stewart wrote[4] something to the effect of: Wisdom and knowledge are strongly connected. One is the complement of the other. Knowledge leads to wisdom and wisdom cannot exist without knowledge.

Steve Jobs said, “Creativity is nothing but connecting things.” Replace the word things with knowledge and it would read: “Creativity is nothing but connecting knowledge.” However, according to many this knowledge is superfluous. Why don’t they realise that without knowledge we will lose the ability to be creative?!

Share

Leave a comment

Maybe they lack the necessary knowledge (and thus the wisdom) to see this.

As Charles Sumner, politician and leader of the movement against slavery in the US stated:

“Without knowledge there can be no sure progress. Vice and barbarism are the inseparable companions of ignorance.”

Amen.

[1] Actually, Sugata Mitra is a very learned man and worthy of respect. He’s an Indian computer scientist who earned a PhD in Solid State Physics from the Indian Institute of Technology (IIT), Delhi and has published on this. The problem is that he suffers from what I have dubbed the Expertise Generalisation Syndrome. This syndrome is a common psychological disorder in people who have expertise in one specific area and who think that this means they also have something interesting and even valuable to say about areas they have no expertise in.

[2] A phenomenon where experts in a specific, narrow field mistakenly believe their expertise translates to other areas—most notably in education and instructional design—without having the necessary knowledge or training in those new fields. See this blog.

[3] Although I would love to discuss the relationship between knowledge and expertise, we won’t. After all, this is a blog and not a book chapter.

[4] John Stewart (1749-1822): The Revolution of Reason, or the Establishment of the Constitution of Things in Nature

¡Advertencia! Lectura extensa (>2000 palabras)

¡Advertencia! Traducción realizada por ChatGPT

La instrucción se parece un poco a ver a un mago o ilusionista: lo que ves parece claro y convincente, pero la apariencia superficial no explica del todo lo que realmente está ocurriendo. En un espectáculo de magia, el público se centra de manera natural en la acción visible y pasa por alto los mecanismos ocultos que producen el efecto. Algo parecido puede ocurrir en educación.

Acabo de publicar un libro con Carl Hendrick y Jim Heal titulado Instructional Illusions. No es un libro sobre mitos educativos en el sentido habitual. No nos limitamos a dar palos a viejas tonterías gastadas como los estilos de aprendizaje y dar el asunto por zanjado. El problema es más interesante que eso, y también más peligroso. Muchas de las cosas que abordamos en el libro no son absurdas. Son plausibles. Contienen un grano de verdad. Precisamente por eso son tan persistentes.

La educación es especialmente vulnerable a la ilusión porque el aprendizaje permanece en gran medida oculto a la vista. Lo que sí podemos ver es actividad, rendimiento, confianza, conformidad, disfrute, discusión, color, movimiento y producción. Lo que realmente nos importa es si los alumnos están aprendiendo algo de manera duradera, recuperable y transferible. Y esas dos cosas no siempre coinciden. De hecho, a menudo no coinciden en absoluto.

Ese es el argumento central de Instructional Illusions. Exploramos diez casos en los que la apariencia superficial de la enseñanza y del aprendizaje oculta una realidad más profunda. O, dicho de forma más sencilla: cosas que parecen aprendizaje a menudo no lo son, y cosas que no parecen especialmente impresionantes a menudo resultan ser exactamente lo que el aprendizaje requiere.

Infografía - Domingo Chica Pardo @dchicapardo

1. La ilusión del compromiso

La primera ilusión es la idea de que, si los estudiantes están implicados, entonces están aprendiendo. Por supuesto, la implicación importa. El problema es que la implicación se ha convertido en un término peligrosamente elástico. Un aula puede estar llena de discusión, movimiento, colaboración y artefactos atractivos mientras los alumnos siguen, cognitivamente hablando, exactamente donde empezaron. Están implicados a nivel conductual y/o emocional, pero NO a nivel cognitivo, y eso es lo que hace falta para aprender.

La pregunta, entonces, no es si los estudiantes están activos. La pregunta es: ¿activos pensando en qué? Si la tarea no centra la atención en los conocimientos y conceptos que deben aprenderse, la implicación se convierte en poco más que un efecto escénico. Parece impresionante. Sale bien en las fotos. Incluso puede resultar agradable. Pero sentirse bien con una actividad no es lo mismo que aprender de ella.

2. La ilusión de la pericia

La gente suele asumir que la pericia se traduce automáticamente en una buena enseñanza. No es así. Los expertos saben más, organizan el conocimiento de manera diferente y mejor, y resuelven problemas con más eficiencia que los novatos. Pero esa pericia crea un problema. Como los expertos han automatizado tanto de lo que saben, a menudo olvidan cómo es no saber. Se saltan pasos, comprimen las explicaciones y dan por supuesto un conocimiento previo que los estudiantes todavía no tienen. Lo que al experto le parece claro y obvio puede resultar desconcertante para el novato.

Los novatos, mientras tanto, tienen su propio punto ciego. A menudo sobrestiman lo que entienden, confunden familiaridad con dominio y no reconocen las lagunas de su propio conocimiento.

Así que tenemos un doble problema: los expertos pueden subestimar lo que necesita explicarse y los novatos pueden sobreestimar lo que han comprendido. Por eso, enseñar bien requiere más que conocer la materia. Requiere entender la distancia entre tu posición como experto y el conocimiento de principiante que tienen tus alumnos, y tender un puente sobre ella. Eso implica una secuenciación cuidadosa, modelado explícito, ejemplos y no ejemplos, atención a las ideas erróneas y una idea clara del conocimiento previo que los estudiantes necesitan para dar sentido al material nuevo.

3. La ilusión centrada en el alumno

Pocas falsas dicotomías en educación han hecho tanto daño como la gastada oposición entre enseñanza centrada en el profesor y centrada en el alumno. La implicación es que uno debe elegir entre una instrucción opresiva y un descubrimiento emancipador. Esto es una tontería. A los estudiantes no se les sirve bien abandonándolos a su suerte simplemente porque a alguien la enseñanza directa le parezca estéticamente poco atractiva.

Los estudiantes, por supuesto, necesitan pensar, explicar, aplicar, discutir y, con el tiempo, regular más su propio aprendizaje. Pero esa capacidad no cae del cielo. Se desarrolla mediante conocimiento, guía y práctica. Lo que a menudo se pierde en estos debates es que la independencia del alumno no es el punto de partida; es el destino. Los estudiantes se vuelven más autónomos a medida que crece su conocimiento. La agencia importa, pero funciona mejor dentro de una estructura, no en lugar de ella. La elección real no es entre enseñanza centrada en el profesor o centrada en el alumno. Es entre una instrucción que se toma en serio el aprendizaje y una instrucción que confunde libertad con eficacia.

4. La ilusión de la transferencia

Todo el mundo quiere que los estudiantes sean capaces de transferir lo aprendido a situaciones nuevas. La escuela no existe para producir alumnos que solo puedan tener éxito en la hoja de ejercicios que vieron ayer. Para ‘combatir’ esto, existe una fuerte tentación de pensar que la transferencia ocurre cuando enseñamos de maneras amplias, abiertas y conectadas con el mundo real. Pero lanzar a los estudiantes a una tarea auténtica y desordenada no produce mágicamente un pensamiento flexible. Normalmente produce confusión, porque los alumnos todavía no tienen las estructuras de conocimiento seguras que necesitan para reconocer qué es relevante y qué no.

La transferencia no es una habilidad genérica que pueda espolvorearse sobre un plan de clase. Depende del conocimiento: conocimiento seguro, bien organizado y específico del dominio, almacenado en esquemas cognitivos amplios y profundos. Los estudiantes transfieren lo que saben cuando saben lo suficiente como para reconocer similitudes y diferencias significativas entre contextos.

Esta es una de las razones por las que las vagas apelaciones al aprendizaje del mundo real resultan tan poco útiles. Si los alumnos todavía no entienden la estructura de conocimiento subyacente de un dominio, ponerlos en una tarea auténtica y desordenada no produce transferencia. Produce confusión. La transferencia y la aplicación no son lo contrario del conocimiento. Son las consecuencias de poseerlo.

5. La ilusión de los resultados fáciles

Otra idea seductora es que un buen aprendizaje debería sentirse fácil, fluido y natural. Si lo has logrado como docente, has tenido éxito. Pero el aprendizaje no carece de fricción. Si nada está siendo desafiado, cambiado, recuperado o reorganizado, hay muchas probabilidades de que no esté ocurriendo gran cosa en términos de aprendizaje. Dicho esto, no toda dificultad es deseable. Algunas dificultades, como la dificultad deseable, son productivas; otras, como el fracaso productivo (o, mejor dicho, fracaso improductivo), no son más que un mal diseño instruccional.

La práctica de recuperación es un buen ejemplo. Puede resultar más difícil que releer apuntes o repasar una solución modelo. El rendimiento incluso puede parecer peor a corto plazo. Pero precisamente ese esfuerzo forma parte de la razón por la que funciona tan bien para la retención a largo plazo. Lo mismo ocurre con la práctica espaciada y otras actividades cognitivamente exigentes que a menudo hacen que el rendimiento parezca peor a corto plazo mientras mejoran el aprendizaje a largo plazo. La lección no es hacer que aprender sea más difícil, sino dejar de confundir la fluidez a corto plazo con el aprendizaje a largo plazo.

6. La ilusión de la motivación

Quizá una de las suposiciones más apreciadas en educación sea que la motivación debe venir primero. Motiva a los niños, y el rendimiento vendrá después. La evidencia no encaja cómodamente con esa historia tranquilizadora. En muchos casos, la relación funciona al revés: el rendimiento genera motivación. El éxito, especialmente el éxito temprano y repetido, hace que los estudiantes estén más dispuestos a invertir esfuerzo.

Eso no significa que la motivación sea irrelevante. Por supuesto que importa. Pero sí significa que debemos desconfiar de tratarla como un precursor mágico que los docentes deben conjurar antes de que pueda comenzar un aprendizaje serio. Los intentos de fabricar motivación mediante eslóganes, ánimos vagos o exhortaciones genéricas a amar el aprendizaje suelen ser inútiles. Decir a los estudiantes que crean en sí mismos, que amen aprender o que desarrollen una mentalidad de crecimiento en abstracto puede sonar estimulante, pero a menudo tiene poco efecto si no están volviéndose realmente más competentes en un dominio concreto. La competencia motiva. El progreso motiva. Ser capaz de hacer algo que antes no podías hacer motiva. Por eso es tan importante descomponer el material en pasos manejables, mantener una alta tasa de éxito durante la práctica guiada y ofrecer retroalimentación útil y accionable.

7. La ilusión del descubrimiento

La creencia de que los estudiantes aprenden mejor cuando descubren las cosas por sí mismos tiene una persistencia notable. Resulta atractiva porque se alinea con valores culturales apreciados: libertad, independencia, autenticidad, creatividad. También resulta atractiva porque queda bien. Ver a los alumnos hacer cosas por sí mismos crea una poderosa impresión visual de apropiación y profundidad.

El problema es que los novatos no son pequeños expertos esperando florecer. Los novatos no abordan los problemas como lo hacen los expertos porque todavía no tienen las estructuras de conocimiento que hacen posible el pensamiento experto. Sin suficiente conocimiento previo, el descubrimiento sin guía o con guía mínima impone grandes exigencias a la memoria de trabajo. Se pide a los estudiantes que busquen, infieran, comparen y generen antes de tener suficiente en la memoria a largo plazo para hacerlo con éxito. Sin suficiente conocimiento previo, las actividades de descubrimiento se reducen al ensayo y error y suelen dejar a los alumnos con una comprensión parcial, ideas erróneas o ninguna comprensión clara. La instrucción explícita no es enemiga de la independencia. Suele ser el camino hacia ella.

8. La ilusión de la singularidad

Los docentes se resisten con razón a la afirmación burda de que todos los aprendices son iguales. No lo son. Los estudiantes difieren en conocimiento previo, dominio del lenguaje, intereses, aptitud, contexto y necesidades específicas. Cualquier explicación seria de la enseñanza tiene que tomarse eso en serio. Pero de esa verdad algunas personas dan un salto hacia una conclusión mucho menos defendible: que, como los alumnos difieren, no puede haber principios generales de enseñanza eficaz. Eso sencillamente no se sigue.

Los seres humanos comparten una arquitectura cognitiva común. La memoria de trabajo es limitada. La memoria a largo plazo importa enormemente. El conocimiento previo determina lo que puede comprenderse. La práctica cambia el rendimiento. La recuperación fortalece la memoria. Estas cosas no son añadidos opcionales. Forman parte de cómo funciona el aprendizaje humano.

Así que sí, los alumnos son únicos en aspectos importantes. Pero no son únicos de maneras que los eximan de la arquitectura básica de la cognición. Una buena enseñanza depende de sostener ambas verdades a la vez: lo que los aprendices comparten y en qué difieren. Si pierdes cualquiera de los dos lados, acabas en problemas. Ignora la diferencia, y te volverás rígido. Ignora lo común, y caerás en la fantasía de que cada alumno necesita un método completamente a medida.

Como dijo tan profundamente Dylan Wiliam: “…la enseñanza es interesante porque los alumnos son muy diferentes, pero solo es posible porque son muy parecidos”.

9. La ilusión del rendimiento

Esta es quizá la más consecuente de todas las ilusiones: el rendimiento no es lo mismo que el aprendizaje. Los estudiantes pueden rendir bien durante una clase o inmediatamente después y aun así recordar muy poco más adelante. Pueden contestar preguntas, completar tareas y producir respuestas correctas en el momento sin haber formado recuerdos duraderos ni una comprensión flexible. Es decir, sin haber aprendido realmente.

El aprendizaje es lo que permanece después, cuando la clase ha terminado, cuando las pistas han desaparecido, cuando ya se ha producido cierto olvido y cuando los estudiantes necesitan recuperar y aplicar lo que saben en condiciones algo diferentes. Por eso un rendimiento fluido en el aula puede ser un indicador tan pobre del aprendizaje a largo plazo. Los alumnos pueden parecer que ‘lo han entendido’ cuando el conocimiento sigue siendo frágil. A la inversa, pueden parecer que están luchando cuando en realidad están implicados en el tipo de procesamiento esforzado que vuelve el aprendizaje más duradero.

Esto importa porque buena parte de la escolarización sigue recompensando lo que es más fácil de observar. Vemos un rendimiento fluido e inferimos un aprendizaje seguro. Vemos esfuerzo e inferimos fracaso. A menudo, la realidad está más cerca de lo contrario. El rendimiento inmediato es con frecuencia un mal indicador de si se ha producido un aprendizaje duradero. Por eso la distinción entre aprendizaje y rendimiento no es una nota técnica al pie. Es una de las ideas más importantes que un docente puede comprender.

10. La ilusión de la innovación

Por último, está la ilusión de la innovación: la suposición de que lo más nuevo es mejor. La educación siempre ha tenido debilidad por la novedad, especialmente cuando la novedad llega envuelta en un lenguaje visionario, entusiasmo tecnológico y promesas de transformación. Sin embargo, gran parte de lo que se vende como innovación está pobremente respaldado por evidencia, mal implementado o no es más que vino viejo en odres nuevos.

Esto no significa que las escuelas deban convertirse en museos. Significa que la innovación debe estar sometida a disciplina. En el libro defendemos la identificación de prácticas con una sólida base empírica —como la instrucción explícita, la práctica de recuperación, el aprendizaje espaciado y los ejemplos resueltos— y convertirlas en el fundamento. Los nuevos enfoques deben ponerse a prueba con cuidado, implementarse bien y evaluarse con escepticismo. La innovación solo merece la pena si mejora el aprendizaje, no si simplemente renueva el envoltorio. No confundas cambio con mejora.

Levantar el velo

En la conclusión del libro, usamos la exposición que hizo James Randi de Uri Geller como metáfora. Randi no derrotó la ilusión discutiendo con más pasión. Controló las condiciones. Utilizó una indagación disciplinada para revelar lo que realmente estaba ocurriendo. Eso, en esencia, es lo que la ciencia del aprendizaje ofrece a la educación: no una iniciativa, no un truco, no un eslogan, sino una lente.

Y ese es realmente el sentido de Instructional Illusions. La educación está llena de cosas que parecen funcionar, parecer correctas y resultar persuasivas. Pero la enseñanza no debería juzgarse por lo convincente que parece en el momento. Debería juzgarse por lo que los estudiantes aún saben, entienden y pueden hacer más tarde. Los ojos y los oídos se dejan engañar con facilidad. Los docentes, si quieren ser verdaderamente profesionales, necesitan algo mejor que las apariencias. Necesitan evidencia, teoría y juicio.

Subscribe now

Share

Leave a comment

Warning!! Longread (>2000 words)

Instruction is a bit like watching a magician or illusionist: what you see feels clear and convincing, but the surface appearance doesn’t fully explain what’s really going on. In a magic show, the audience naturally focuses on the visible action and misses the hidden mechanisms that produce the effect. Something similar can happen in education.

I’ve just published a book with Carl Hendrick and Jim Heal called Instructional Illusions. It’s not a book about edu-myths in the usual sense. We’re not simply taking a bat to tired old nonsense such as learning styles and calling it a day. The problem is more interesting than that, and also more dangerous. Many of the things we address in the book aren’t absurd. They’re plausible. They contain a grain of truth. That’s precisely why they’re so persistent.

Education is especially vulnerable to illusion because learning is largely hidden from view. What we can see is activity, performance, confidence, compliance, enjoyment, discussion, colour, movement, and output. What we actually care about is whether students are learning something in ways that are durable, retrievable, and transferable. Those aren’t always the same thing. In fact, they often aren’t the same thing at all.

That’s the central argument of Instructional Illusions. We explore ten cases where the surface appearance of teaching and learning masks a deeper reality. Or, put more simply: things that look like learning often aren’t, and things that don’t look especially impressive often turn out to be exactly what learning requires.

1. The engagement illusion

The first illusion is the idea that if students are engaged, they’re learning. Of course, engagement matters. The problem is that engagement has become a dangerously elastic term. A classroom can be buzzing with discussion, movement, collaboration, and attractive artefacts while students remain, cognitively speaking, exactly where they started. They’re behaviorally and/or emotionally engaged, but NOT cognitively engaged, and that’s what’s needed to learn.

The question, then, isn’t whether students are active. The question is: active in thinking about what? If the task doesn’t focus attention on the knowledge and concepts to be learned, engagement becomes little more than a stage effect. It looks impressive. It photographs well. It may even feel good. But feeling good about an activity isn’t the same as learning from it.

2. The expertise illusion

People often assume that expertise automatically translates into good teaching. It doesn’t. Experts know more, organise knowledge differently and better, and solve problems more efficiently than novices. But that expertise creates a problem. Because experts have automatised so much of what they know, they often forget what it’s like not to know. They skip steps, compress explanations, and assume background knowledge that students don’t yet have. What feels clear and obvious to the expert can be baffling to the novice.

Novices, meanwhile, have their own blind spot. They often overestimate what they understand, mistake familiarity for mastery, and don’t recognise the gaps in their own knowledge.

So we have a double problem: experts can underestimate what needs to be explained, and novices can overestimate what they’ve grasped. Good teaching therefore requires more than knowing your subject. It requires understanding the gap between where you are as an expert and novice knowledge that your students have and building a bridge across it. That means careful sequencing, explicit modelling, examples and non-examples, attention to misconceptions, and a clear sense of what prior knowledge students need if they’re to make sense of the new material.

3. The student-centred illusion

Few false dichotomies in education have done more damage than the tired opposition between teacher-centred and student-centred. The implication is that one must choose between oppressive instruction and emancipatory discovery. This is nonsense. Students aren’t well served by being abandoned to their own devices simply because someone finds direct teaching aesthetically unappealing.

Students do, of course, need to think, explain, apply, discuss, and eventually regulate more of their own learning. But that capacity doesn’t fall from the sky. It develops through knowledge, guidance, and practice. What often gets missed in these debates is that student independence isn’t the starting point; it’s the destination. Students become more autonomous as their knowledge grows. Agency matters, but it works best within structure, not instead of it. The real choice isn’t between teacher-centred and student-centred. It’s between instruction that takes learning seriously and instruction that confuses freedom with effectiveness.

4. The transfer illusion

Everyone wants students to be able to transfer what they’ve learnt to new situations. Schools aren’t there to produce students who can only succeed on the worksheet they saw yesterday. To ‘combat’ this there’s a strong temptation to think that transfer is what happens when we teach in broad, open-ended, real-world ways. But throwing students into a messy authentic task doesn’t magically produce flexible thinking. Usually it produces confusion, because students don’t yet have the secure knowledge structures needed to recognise what’s relevant and what isn’t.

Transfer isn’t a generic skill that can simply be sprinkled over a lesson plan. It depends on knowledge: secure, well-organised, domain-specific knowledge stored in broad and deep cognitive schemata. Students transfer what they know when they know enough to recognise meaningful similarities and differences across contexts.

This is one reason why vague appeals to real-world learning are so often unhelpful. If students don’t yet understand the underlying knowledge structure of a domain, putting them into a messy authentic task doesn’t produce transfer. It produces confusion. Transfer and application aren’t the opposite of knowledge. They’re the consequences of having it.

5. The easy-wins illusion

Another seductive idea is that good learning should feel easy, smooth, and natural. If you’ve achieved that as teacher, you’ve succeeded. But learning isn’t frictionless. If nothing’s being challenged, changed, retrieved, or reorganised, there’s a fair chance not much learning is happening. That said, not all difficulty is desirable. Some difficulty like desirable difficulty is productive; some like productive failure (or better named unproductive failure) is merely bad instructional design.

Retrieval practice is a good example. It can feel harder than rereading notes or looking over a worked solution. Performance may even appear worse in the short term. But that very effort is part of why it works so well for long-term retention. The same goes for spaced practice and other conditions that make learning less smooth but more durable. The same is true for spaced practice, and other cognitively effortful activities that often make performance look worse in the short term while making learning better in the long term. The lesson isn’t make learning harder, but rather stop confusing short-term fluency with long-term learning.

6. The motivation illusion

Perhaps one of the most cherished assumptions in education is that motivation must come first. Get children motivated, and achievement will follow. The evidence doesn’t sit comfortably with that comforting story. In many cases, the relationship runs the other way: achievement breeds motivation. Success, especially early and repeated success, makes students more willing to invest effort.

That doesn’t mean motivation is irrelevant. Of course it matters. But it does mean we should be wary of treating it as a magical precursor that teachers must somehow conjure before serious learning can begin. Attempts to manufacture motivation through slogans, vague encouragement, or generic exhortations to love learning are often useless. Telling students to believe in themselves, to love learning, or to develop a growth mindset in the abstract may sound uplifting, but it often has little effect unless students are actually becoming more competent in a specific domain. Competence is motivating. Progress is motivating. Being able to do something you couldn’t do before is motivating. That’s why breaking material into manageable steps, maintaining a high success rate during guided practice, and giving actionable feedback matter so much.

7. The discovery illusion

The belief that students learn best when they discover things for themselves has remarkable staying power. It’s attractive because it aligns with cherished cultural values: freedom, independence, authenticity, creativity. It’s also attractive because it looks good. Students doing things for themselves creates a powerful visual impression of ownership and depth.

The problem is that novices aren’t little experts waiting to bloom. Novices don’t approach problems the way experts do because they don’t yet have the knowledge structures that make expert thinking possible. Without sufficient prior knowledge, unguided or minimally guided discovery places heavy demands on working memory. Students are asked to search, infer, compare, and generate before they’ve got enough in long-term memory to do those things successfully. Without sufficient prior knowledge, discovery activities are reduced to trial-and-error and place heavy demands on working memory and often leave students with partial understanding, misconceptions, or no clear understanding at all. Explicit instruction isn’t the enemy of independence. It’s usually the route to it.

8. The uniqueness illusion

Teachers rightly resist the crude claim that all learners are the same. They aren’t. Students differ in prior knowledge, language proficiency, interests, aptitude, background, and specific needs. Any serious account of teaching has to take that seriously. But from that truth some people leap to a much less defensible conclusion: that because learners differ, there can be no general principles of effective teaching. That simply doesn’t follow.

Human beings share a common cognitive architecture. Working memory is limited. Long-term memory matters enormously. Prior knowledge shapes what can be understood. Practice changes performance. Retrieval strengthens memory. These things are not optional extras. They’re part of how human learning works.

So yes, learners are unique in important ways. But they’re not unique in ways that exempt them from the basic architecture of cognition. Good teaching depends on holding both truths together: what learners share, and how learners differ. Lose either side, and you end up in trouble. Ignore difference, and you become rigid. Ignore commonality, and you drift into the fantasy that every learner needs an entirely bespoke method.

As Dylan Wiliam so profoundly said “…teaching is interesting because students are so different, but it is only possible because they are so similar.”[1]

9. The performance illusion

This is perhaps the most consequential illusion of the lot: performance isn’t the same as learning. Students can perform well during a lesson or immediately after it and still remember very little later on. They can answer questions, complete tasks, and produce correct responses in the moment without having formed durable memories or flexible understanding. That is, without actually learning.

Learning is what remains later, after the lesson has ended, after the cues have disappeared, after some forgetting has taken place, and after students need to retrieve and apply what they know under somewhat different conditions. That’s why smooth classroom performance can be such a poor proxy for long-term learning. Students may look like they’ve “got it” when the knowledge is still fragile. Conversely, students may appear to be struggling when they’re actually engaged in the kind of effortful processing that makes learning more durable.

This matters because much of schooling still rewards what’s easiest to observe. We see smooth performance and infer secure learning. We see struggle and infer failure. Often the reverse is closer to the truth. Immediate performance is frequently a poor proxy for whether long-term learning has taken place. That’s why the distinction between learning and performance isn’t a technical footnote. It’s one of the most important ideas teachers can understand.

10. The innovation illusion

Finally, there’s the innovation illusion: the assumption that newer means better. Education has always had a weakness for novelty, particularly when novelty arrives wrapped in visionary language, technological excitement, and promises of transformation. Yet much of what’s sold as innovation is either poorly evidenced, badly implemented, or merely old wine in a new bottle.

This doesn’t mean schools should become museums. It means innovation should be disciplined. In the book we argue for identifying practices with strong evidence behind them, such as explicit instruction, retrieval practice, spaced learning, and worked examples, and making those the foundation. New approaches should be tested carefully, implemented well, and evaluated sceptically. Innovation is only worthwhile if it improves learning, not if it merely refreshes the packaging. Don’t confuse change with improvement.

Lifting the veil

In the conclusion to the book, we use James Randi’s exposure of Uri Geller as a metaphor. Randi didn’t defeat the illusion by arguing more passionately. He controlled the conditions. He used disciplined inquiry to reveal what was actually happening. That, in essence, is what the science of learning offers education: not an initiative, not a gimmick, not a slogan, but a lens.

And that’s really the point of Instructional Illusions. Education is full of things that seem to work, look right, and feel persuasive. But teaching shouldn’t be judged by how convincing it appears in the moment. It should be judged by what students still know, understand, and can do later. The eyes and ears are easily fooled. Teachers, if they’re to be properly professional, need something better than appearances. They need evidence, theory, and judgment.

[1] Wiliam, D. (1999). Formative assessment in mathematics. Part 1: Rich questioning.Equals: Mathematics and Special Educational Needs 5(2) 15-18.

Subscribe now

Share

Leave a comment

Laat dat even bezinken. Vijf jaar. Twee komma drie miljoen euro. Een jaar bouwput. onvoldoende onderwijsresultaten, kinderen die niet goed geleerd hebben en docenten die gefrustreerd goed onderwijs probeerden te geven… En waarom? Omdat men heeft ontdekt wat je al had kunnen weten vóórdat de eerste paal de grond in ging. Zoals Lee Shulman het zo mooi zei: Het is al erg genoeg om het wiel opnieuw uit te vinden. Maar pas echt pijnlijk is het als ze de lekke band opnieuw uitvinden.

Open leerpleinen worden verkocht als modern. Als toekomstbestendig. Als flexibel. En vooral: als iets dat past bij mooie woorden als samenwerken, autonomie en “21st century skills”. Het probleem is niet dat samenwerking of autonomie slechte doelen zijn. Het probleem is dat je een schoolgebouw niet moet ontwerpen op basis van visies, slogans en esthetiek, maar op basis van de voorwaarden waaronder leren mogelijk is. En een van die voorwaarden is dat leerlingen aandacht kunnen richten en vasthouden. Dat klinkt simpel, maar het blijkt in onderwijsbeleid – zo te zien bij basisschool Wonderwijs, Onderwijsstichting Movare en de Gemeente Heerlen – verrassend ingewikkeld.

In een open leerplein gebeurt namelijk precies wat je niet wilt: er is permanent geluid, beweging, visuele prikkeling en sociale ruis. En ja, die vier horen bij elkaar. Geluid is zelden “achtergrond”; het is vaak intrusief (indringend) en betekenisvol, zeker wanneer het uit de mond van leeftijdsgenoten komt. Beweging trekt automatisch het oog. Visuele prikkels trekken aandacht zonder dat leerlingen daar een bewuste keuze voor maken. En sociale ruis is het ergst van allemaal, omdat het precies het soort prikkel is waar kinderen evolutionair en psychologisch op afgesteld zijn.

Ik verwoordde dat in een artikel over leerpleinen lekker ongepolijst: kinderen zijn “net eksters”; alles wat beweegt trekt aandacht, en op een leerplein wordt dat ekstergedrag voortdurend aangesproken. Het gevolg is voorspelbaar: concentratie wordt moeilijker, cognitieve belasting stijgt, en leren wordt slechter. Niet omdat kinderen “niet gemotiveerd” zijn, maar omdat we het brein vragen het onmogelijke te doen: focussen in een omgeving die gebouwd is om te storen.

Wie denkt dat dit een kwestie is van smaak – “sommigen houden van open, anderen van gesloten” – mist dat dit onderwerp al decennia onderzocht is. Neem de akoestische literatuur over open-plan-klaslokalen. Shield en Dockrell hebben in een review veertig jaar onderzoek samengevat over geluid in open-plan basisscholen. Hun conclusie is niet verrassend: open ruimtes zorgen voor meer storend geluid, slechtere spraakverstaanbaarheid en meer afleiding. En dit is geen ‘klein ongemak’. Als je een leerling niet goed kan laten horen wat de leraar zegt, of als leerlingen elkaar voortdurend horen, dan tast je de kern van onderwijs aan.

Nog concreter: onderzoek van Mealings en collega’s bespreekt precies het intrusieve karakter van geluid in open-plan omgevingen die verkocht worden als “21st century learning spaces”. Het gaat niet alleen om geluidsniveau, maar om de onvoorspelbaarheid en betekenis van het geluid: je hoort taal, je hoort instructie van een andere groep, je hoort feedback, je hoort klassengesprekken. Dat is geen ‘white noise’ waar je na twee minuten aan went. Dat is cognitief concurrerende input die steeds opnieuw om aandacht vraagt.

En dan is er nog de bredere literatuur over schoolgebouwen en leerprestaties, waarin ruis al jaren naar voren komt als een van de weinige factoren waar de evidence relatief robuust is. Woolner en collega’s bespreken dat ook: terwijl er discussie is over allerlei ‘design features’, is het gewicht van bewijs rond noise en afleiding opvallend consistent.

Wie dit liever in een Nederlandse, lerarenvriendelijke tekst leest: Mirjam Neelen en ik schreven er een stuk over met de veelzeggende titel “Stilte AUB”. Daarin wordt nog eens uitgelegd dat omgevingsgeluid niet alleen irritant is, maar ook samenhangt met slechtere prestaties. Geen spektakel, geen hype, gewoon onderzoek.

En toch bouwen we (basisschool Wonderwijs, Onderwijsstichting Movare en de Gemeente Heerlen) dit soort leerpleinen. Waarom? Omdat onderwijs zo nu en dan last heeft van een vorm van collectieve amnesie, gecombineerd met een verslaving aan het woord “innovatie”. Open ruimtes zien er modern uit. Ze leveren mooie foto’s op. Ze passen bij beleidsdocumenten. Ze trekken ondeskundige ouders aan. Ze suggereren flexibiliteit en efficiëntie. En ze geven bestuurders het gevoel dat ze “iets gedaan hebben”.

Maar onderwijs is geen interieurtrend. Je ontwerpt een school niet voor de brochure, maar voor de cognitieve realiteit van kinderen. En die realiteit is dat aandacht beperkt is en werkgeheugen kwetsbaar is. Vooral bij beginners. Vooral bij kinderen met minder voorkennis. Vooral bij leerlingen die thuis minder ondersteuning krijgen. Ironisch genoeg zijn dat precies de leerlingen voor wie een prikkelrijke omgeving het meest schadelijk is. Experts kunnen ruis beter filteren; beginners niet. Dus als je “inclusief” wilt zijn, is een leerplein een merkwaardige keuze.

Het wrange van het Wonderwijs-verhaal is dat de remedie nu ouderwets is: muren terug, scheiding terug, structuur terug. En dat kost volgens de Limburger dus 2,3 miljoen euro en een jaar tijd. Niet omdat men dat niet kon weten, maar omdat men het kennelijk niet wílde weten.

Wat had er moeten gebeuren? Geen bullet points, geen managementtaal. Gewoon een simpele stap die in elke andere sector normaal is bij grote investeringen: eerst kijken wat de evidence zegt. Je laat je adviseren door mensen die de literatuur kennen, vooral dichtbij zijnde onderwijswetenschappers (ik noem geen namen, maar hij staat bekend als onderwijsrealist en Grumpy Old Man), maar ook akoestiek en omgevingspsychologen. Je maakt een ontwerp dat niet afhankelijk is van het wonderlijke idee dat kinderen zichzelf wel leren concentreren te midden van constante prikkels. En je bouwt niet op basis van filosofie, hype of onderbuik, maar op basis van hoe leren daadwerkelijk werkt.

Want innovatie zonder kennis is geen innovatie. Het is gokken. En in dit geval gok je met kinderen, hun lestijd en publiek geld.

Nog meer willen weten?

Connolly, D., Dockrell, J., Shield, B., Conetta, R., Mydlarz, C., & Cox, T. (2019). The effects of classroom noise on the reading comprehension of adolescents. Journal of the Acoustical Society of America, 145(1):372. https://doi.org/10.1121/1.5087126

Jahncke, H., & Hallman, D. M. (2020). Objective measures of cognitive performance in activity based workplaces and traditional office types. Journal of Environmental Psychology,72, 101503. https://doi.org/10.1016/j.jenvp.2020.101503

Pujol, S., Levain, JP., Houot, H. et al. (2014). Association between ambient noise exposure and school performance of children living in an urban area: A cross-sectional population-based study. Journal of Urban Health, 91, 256–271. https://doi.org/10.1007/s11524-013-9843-6

Zhang, B., & Navejar, R. (2018). Effects of ambient noise on the measurement of mathematics achievement for urban high school students. Urban Education, 53, 1195-1209. https://doi.org/10.1177/0042085915613555

Extra informatie

• https://www.forbes.com/sites/nataliewexler/2022/07/18/elementary-classrooms-are-too-noisy-for-kids-to-learn/?sh=4ef8bdf63c60

• https://www.nytimes.com/1982/04/26/nyregion/student-scores-rise-after-nearby-subway-is-quieted.html

• https://www.edweek.org/teaching-learning/sounding-an-alarm-background-noise-can-hurt-student-achievement/2018/11

• https://www.edweek.org/teaching-learning/children-may-struggle-more-with-a-noisy-classroom-than-adults/2019/02

  Subscribe now

  Share

  Leave a comment

© https://picryl.com/media/alignment-psf-2f776c

Many of us were taught, sometimes explicitly and sometimes implicitly, that good instruction is about alignment. Lessons should align: learning goals should align with how things are taught, which should align with assessments of learning (i.e., Biggs, 1996; constructive alignment). Practice should resemble the test. If students are going to be examined in a particular way, then surely they should practise in that way too. The logic is both seductive and intuitive: alignment feels fair, efficient, and professional.

But most teachers also experience a persistent and frustrating pattern. Learners perform well during practice but stumble on tests. They say things like, “I knew this yesterday,” or “This looks different.” Teachers respond, understandably, with disbelief: but we covered this and it looked like everyone knew it. There’s something about alignment that seems to work while at the same time, seems not to work at all.

This tension sits at the heart of what looks like a contradiction between two well-established ideas from cognitive psychology: transfer-appropriate processing(TAP) and contextual interference. One seems to argue for matching and similarity, while the other argues for variation and disruption. In reality, they’re not opposed. They’re talking about different levels of learning (actually two different things) or processing (Craik & Lockhart, 1972; Morris, Bransford, & Franks, 1977), and together they explain why so much seemingly good teaching produces fragile learning.

Let’s start with transfer-appropriate processing (TAP) which is often summarised as “learning works best when practice matches the test.” While that slogan is catchy, it’s also misleading. TAP doesn’t claim that formats, question types, or layouts used during teaching should match the assessment. It actually claims that learning transfers best when the mental processes used during learning overlap with those required later. Biggs (p. 353) writes: “The teaching methods we choose need to engage students in activities that are likely to require them to perform in the way nominated in the curriculum objectives.” In other words, the way alignment is applied can be seen as a lethal mutation[1] of what it actually means. The emphasis isn’t on matching formats, contexts, or appearances; it’s on matching thinking.

In transfer-appropriate processing, what matters is whether the kind of thinking learners do during practice is the same kind of thinking they’re required to do during later assessment or application. Practising recognition tasks prepares you for recognition. Practising explanation tasks prepares you for explanation. Practising generation tasks prepares you for generation. When learning and later use require different kinds of thinking than during the teaching, performance drops even when the content is familiar.

Seen correctly, TAP isn’t a defence of teaching to the test in the usual sense. It’s a warning about process mismatch. Students don’t always fail because they lack the necessary knowledge, but rather because they were trained to do the wrong kind of cognitive work (i.e., processing). If students practise recognising answers but are later asked to make deductions, failure shouldn’t surprise us. I spoke about this years ago when my ‘professorial chair’ was in collaborative learning at both Maastricht University and the University of Utrecht. What often occurs in collaborative learning is that we ask students to work collaboratively (or cooperatively) but test them individually.

Contextual interference refers to varying the conditions under which learning and practice take place. The task remains the same, but the context changes: the format, the cues, the representation, and the situation in which the knowledge appears. Research shows that this variation reliably makes practice harder and performance worse in the short term, while improving retention and transfer in the long term.

At first glance, contextual interference seems to contradict TAP. If learning should match later performance, why deliberately introduce variation? Why practise under conditions that don’t resemble the test?

The answer is that contextual interference deliberately disrupts surface similarity in order to strengthen process similarity. Stable contexts allow learners to lean on cues provided by the environment when they must reproduce the information. They learn not just the content, but also the situation. This makes retrieval easier, but also fragile. If you change the situation, the performance collapses. What contextual interference does is remove those crutches. When formats change, cues disappear, and representations vary, learners can no longer rely on recognition or routine during practise. They have to retrieve the underlying knowledge, decide what applies and what doesn’t, and (re)construct a solution. In other words, they must engage in exactly the kind of thinking that most assessments and almost all real-world applications actually should demand.

And this is where contextual interference and TAP stop looking like opposites. Contextual interference is actually one of the most powerful ways to implement TAP correctly. It ensures that learners practise the right cognitive processes, rather than merely rehearsing familiar forms.

The connection becomes especially clear when we look at assessment. Many so-called assessment failures aren’t failures of learning, but failures of transfer. Tests often remove the contextual supports that were present during instruction: the familiar wording, the predictable layout, the worked-example-like cues. From the teacher’s perspective, this feels unfair. From the student’s perspective, it feels inexplicable. In reality, the assessment has revealed whether students can perform when those supports are gone.

This also explains why “teaching to the test” can feel effective and still fail. When practice mirrors the test closely at a surface level, students learn to recognise patterns and not to understand principles. Scores improve quickly. Student (and teacher) confidence rises. But this is alignment of appearance, not alignment of thinking. The moment students encounter the same ideas in a different form, for example on an exam, in another subject area, or outside school in everyday situations, the apparent mastery evaporates into thin air.

Here is where TAP allows us to rescue teaching to the test and make it something positive. Teaching to the test works when we redefine what ‘the test’ actually is. If the test requires retrieval without cues, then teaching to the test means practising retrieval without cues. If the test requires interpreting unfamiliar representations, then teaching to the test means practising with varied representations. If the test requires reasoning and adaptation, then teaching to the test means practising reasoning and adaptation; early, often, and under changing conditions. The tail is here, effectively wagging the dog!

© Ron Leishman

Good teaching to the test doesn’t copy the test. It trains the cognitive work the test demands. Contextual interference supports this kind of alignment. By varying contexts during learning, teachers ensure that students aren’t learning where or how to answer, but what to think about and why. From a TAP perspective, contextual interference is not teaching away from the test; it’s teaching to the real test hiding behind the paper.

There’s, however, an important caveat. Neither TAP nor contextual interference implies that more difficulty is always better. For novices, the cognitive processes required for transfer may themselves be too demanding. Early in learning, students need guidance, worked examples, and stable contexts to build initial cognitive schemata.

When learners lack basic schemas, variation overwhelms working memory. In that phase, stability isn’t the enemy of learning but a prerequisite. Contextual interference becomes desirable only if there’s something stable to abstract across. As learning progresses, however, stability must give way to variation, support must be reduced, and contexts must change. Otherwise, learners remain dependent on cues that won’t be present when it matters.

This brings us back to the paradox. Practice that looks neat and aligned often undermines learning, while practice that looks messy and non-aligned often supports it. Smooth lessons can produce brittle knowledge, while struggle during practice can produce deep learning and flexible understanding. Alignment isn’t the problem. Rather, aligning with the wrong things is. When we align instruction with surface features, we train recognition of those surface features, while when we align instruction with cognitive processes, we train transfer.

So no, transfer-appropriate processing and contextual interference aren’t enemies. Transfer-appropriate processing tells us what kind of thinking learning must involve. Contextual interference is one powerful way to ensure that kind of thinking actually happens. If learning looks a little worse in the moment, that may be the clearest sign that it’s finally doing what we want it to do.

Leave a comment

Share

Subscribe now

Biggs, J. (1996). Enhancing teaching through constructive alignment. Higher Education, 32, 347–364. https://doi.org/10.1007/BF00138871

Craik, F. I., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11(6), 671-684. https://doi.org/10.1016/S0022-5371(72)80001-X

Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of Verbal Learning and Verbal Behavior. 16(5),519–533. https://doi.org/10.1016/s0022-5371(77)80016-9

[1] According to TES, a lethal mutation occurs when evidence-informed practice is modified beyond recognition from the original practice. This can happen when teachers adapt evidence-informed practices and techniques to fit within their own contexts, leading them to unwittingly develop counterproductive strategies.