If you went through school in the last forty years, you probably encountered the idea at some point. Maybe a teacher asked you to take a short survey to find out whether you were a visual learner, an auditory learner, or a kinaesthetic learner. Maybe you took the results seriously and noticed, a little self-conscious about your new identity, that yes, now that you thought about it, you did seem to learn better when there were pictures involved. Maybe a parent explained to a teacher that their child was “a visual learner” and therefore struggled with lectures. Maybe you still believe something along these lines about yourself.
The idea is remarkably widespread. Surveys in Australia, the United Kingdom and the United States consistently find that well over 90 per cent of teachers believe in learning styles. It appears in teacher-training courses, in education guidebooks, in self-help material for students, in corporate training programmes, and in the way many well-meaning adults explain their own relationship to learning. It is, by cultural popularity, one of the most successful ideas in education of the last half-century.
It is also, according to almost every serious empirical review, wrong. Not partially wrong. Not misleading in some contexts. Wrong in roughly the way that homeopathy is wrong — as a factual matter, there is essentially no evidence that matching instruction to students’ self-reported learning styles improves learning outcomes. And yet the belief persists, taught in schools, invoked in conversations, used to explain students’ struggles, and quietly draining time and effort that could go into techniques that do work.
This is an unusual moment in the gap between what research shows and what practitioners believe. Worth understanding.
The 2008 review that should have ended it
In 2008, four distinguished cognitive psychologists — Harold Pashler, Mark McDaniel, Doug Rohrer, and Robert Bjork — were commissioned by the journal Psychological Science in the Public Interest to conduct a comprehensive review of the evidence for learning styles. The question they were asked was specific: is there good evidence that matching the style of teaching to students’ preferred learning style improves learning outcomes?
The authors started from what they called the meshing hypothesis — the specific claim that instruction should be “meshed” with a student’s preferred style (visual instruction for visual learners, auditory for auditory, and so on). To test it, they needed research with a specific design: take students of various self-reported styles, randomly assign them to matched-style or mismatched-style instruction, and compare learning outcomes.
What they found, after a thorough search of the literature, was striking. Across the hundreds of papers purporting to support learning styles, very few used the rigorous design needed to actually test the meshing hypothesis. Most papers simply showed that students had preferences — which is true but uninteresting — or that some teaching methods work better than others — which is also true but has nothing to do with individual styles. The small number of rigorously designed studies, which actually tested whether matching teaching to style produced better outcomes, mostly found no effect. Some found small effects in the opposite direction of the hypothesis. None found the consistent, strong effects that the popularity of the idea would require.
Pashler and his colleagues concluded, carefully and firmly, that there was no scientific basis for the learning-styles framework as commonly taught. The conclusion was not that individual differences don’t exist — they do, and they matter — but that the specific claim, so often repeated, that teaching should be tailored to self-reported visual/auditory/kinaesthetic styles, has no good evidence behind it.
The persistence of the myth
In the fifteen years since the Pashler review, the consensus in cognitive science has only hardened. A 2020 paper by Steven Kirschner and Jeroen van Merriënboer in the Netherlands, aptly titled Do Learners Really Know Best? Urban Legends in Education, reviewed several myths common in education practice. Learning styles was prominent among them. Their review, building on Pashler’s, reached the same conclusion: there is essentially no evidence that matching instruction to self-reported styles improves outcomes, and the persistence of the belief in the face of this evidence is itself a phenomenon worth studying.
What’s striking is how fully the academic consensus has diverged from what teachers actually believe. In a 2020 international survey of teachers, over 95 per cent of respondents from multiple countries agreed with statements endorsing the learning-styles framework. In those same years, you’d have struggled to find any serious cognitive scientist who endorsed it. The two communities — classroom practitioners and research scientists — are operating on the basis of fundamentally different understandings of the same question.
Why the gap has persisted is a subject of genuine debate. Several contributing factors are worth naming.
The idea is intuitively appealing. We all notice preferences in ourselves. We do seem to enjoy some kinds of learning more than others. It’s a small step from “I prefer this kind of material” to “I learn better with this kind of material” — a step that turns out to be empirically wrong but that feels, from the inside, reasonable.
It flatters individuality. The learning-styles framework suggests that failures to learn are not about effort or technique but about a mismatch between the learner’s fixed style and the teaching they’ve been given. This is comforting. It converts educational difficulty into a neutral compatibility problem.
It supports teachers who want to differentiate. Good teachers want to respond to the different students in their classrooms. The learning-styles framework offered a concrete language for doing this, even though the specific categories it proposed had no empirical grounding.
The training system perpetuates it. Teacher-education programmes, for decades, have included learning styles as established knowledge. Teachers who have been taught this pass it on. Changing what’s in teacher training is slow, and retracting something once it’s been taught for a generation is even slower.
What the research does support
Debunking learning styles isn’t the same as saying that individual differences don’t matter. Cognitive science has plenty of evidence for real differences in how people learn — just not the specific differences the learning-styles framework proposes.
Students differ in prior knowledge. This matters enormously for what they can learn next and how quickly. A student who already knows the prerequisite material will learn a new topic much faster than one who doesn’t.
Students differ in working memory capacity. This is harder to change, and it affects how much new information can be held and manipulated at once. Students with lower working memory capacity benefit more from instruction that carefully limits cognitive load.
Students differ in motivation and interest. A student who cares about a topic will engage more deeply with it and learn more. This is partly trait and partly context.
Students differ in the particular gaps in their understanding of any given topic. Good teaching identifies and addresses those specific gaps — which is what differentiation ought to mean — rather than matching some putative learning style.
Beyond these individual differences, there’s a second finding that the learning-styles enthusiasts sometimes accidentally stumble toward: content should be taught in the mode that best fits the content. Spatial material — maps, molecular structures, geometry — should be taught with diagrams. Verbal material — argumentation, poetry, historical narrative — should be taught with words. Motor skills — handwriting, surgery, musical performance — need to be practised physically. This matching is real, and it’s important, but it has nothing to do with individual student styles. It has to do with matching the mode of instruction to the subject being taught. The student who needs to learn anatomy benefits from diagrams regardless of whether they’re a “visual learner”. The student who needs to learn a language benefits from speaking practice regardless of whether they’re an “auditory learner”.
This is the kernel of truth that the learning-styles framework accidentally contains and then spoils by attaching it to the wrong variable. Teach content in the mode the content needs. Vary your teaching to cover different kinds of content. Attend to individual students’ actual difficulties. Don’t pretend to know their “style” from a short questionnaire.
What to actually do — for students, for teachers
If you’re a student who has identified yourself as a particular kind of learner and have been acting on that identification, the honest thing to do is set the identification aside. Not because your preferences aren’t real — they are — but because treating them as fixed learning needs has probably cost you more than it’s gained you. Some material you find uncomfortable to learn from lectures will be learned through lectures anyway, because that’s how the material is presented. Some material you find uncomfortable in text form has to be learned through text, because that’s the most efficient way. Training yourself to learn from modes you don’t prefer is part of becoming a more capable learner, not a violation of your true nature.
Practical study advice, consistent with the research: focus on techniques that have genuine evidence behind them — retrieval practice, spaced repetition, interleaved practice, self-explanation, elaboration — rather than on matching your study technique to your supposed style. These evidence-based techniques work for everyone, regardless of their self-reported preferences.
If you’re a teacher — or a parent, or anyone else who works with learners — the implication is similar. The energy that goes into learning-styles accommodations can be redirected into more effective things: identifying students’ actual prior knowledge, varying instruction modes across content types, using retrieval practice and spaced review, checking understanding often, and responding to specific gaps in comprehension rather than supposed style mismatches.
The broader lesson
The story of learning styles is, in one sense, a cautionary tale about the persistence of comforting ideas. The framework feels right. It fits our intuitions. It gives us useful vocabulary for differences we really do notice. And yet it turns out to be wrong, and continuing to act on it wastes the time and effort of millions of teachers and students every year.
Education is a field where beliefs travel faster than evidence. Something that sounds plausible gets taught, gets repeated, gets baked into teacher training, and by the time better research comes along, a generation has built careers on the earlier idea. Updating takes decades, and in the meantime, actual students are being taught worse than they could be.
The question that remains
The learning-styles myth is a good exercise in a particular kind of intellectual humility. You probably believed something about yourself or others that the best evidence doesn’t support. You probably still feel, even after reading this, that you really are more visual than auditory, or that your son really is a kinaesthetic learner. That feeling is real. It’s also probably telling you about your preferences, not your learning needs. The two aren’t the same.
The question worth carrying, not just about learning styles but about many comforting ideas:
Of the things you confidently believe about yourself or the people close to you, how many have you ever genuinely checked — and how many have you simply inherited, because they were handed to you by people you trusted?
Key research referenced: Harold Pashler, Mark McDaniel, Doug Rohrer and Robert Bjork’s 2008 review for Psychological Science in the Public Interest; Steven Kirschner and Jeroen van Merriënboer’s 2013 Educational Psychologist paper on urban legends in education; ongoing survey research on teacher beliefs about learning styles.