Nobody is a visual learner

01The claim

Learning styles come in dozens of flavors, but the version most of us met is VARK: Visual, Aural, Read/write and Kinesthetic. Neil Fleming built it in 1987 at Lincoln University in New Zealand. It was short, friendly and fit on a poster, which is exactly why it spread. The irony is that Fleming meant it as a prompt for people to reflect on how they study. It was never designed as a sorting system for teachers to plan lessons around. The system took it and used it that way anyway.

The claim underneath all of these models is simple. Each student has a preferred channel, and that student learns more when teaching is delivered through it. Researchers call this the "meshing hypothesis." It sounds reasonable. People clearly are different, and everyone has preferences. The problem is that a preference and a benefit are two very different things, and only one of them has ever been demonstrated.

By 2004 the field had grown so crowded that a team of British researchers led by Frank Coffield set out to review it. They counted 71 different learning-style models and found serious problems with almost all of the influential ones: questionnaires that gave different answers when the same person took them twice, categories that did not hold together statistically, and very little independent evidence that any of it improved learning.

02The fair test

Think about how you would prove that some people digest grilled fish better and others digest fried fish better, and that a questionnaire can tell you which is which. Asking diners which they prefer proves nothing. You would have to sort people by the questionnaire first, then randomly give each group either the matching or the mismatching dish, and then show a crossover. The "grilled people" must do best on grilled. The "fried people" must do best on fried. If everybody simply does better on grilled, your questionnaire told you nothing. You have discovered that grilling is better, full stop.

In 2008, four cognitive scientists, Harold Pashler, Mark McDaniel, Doug Rohrer and Robert Bjork, laid out exactly this standard for learning styles. Then they went looking for studies that met it. They found virtually none that supported the theory, and several that directly contradicted it. Their conclusion, published in Psychological Science in the Public Interest, was that there was no adequate evidence base to justify building learning styles into education, and that schools would be better off putting their resources into techniques with real evidence behind them.

People did run it properly after that, and the results kept coming back the same.

That Knoll finding explains thirty years of conviction. Studying in your preferred format feels smoother and more natural, so it feels like better learning. Your own sense of how well you learned is an unreliable witness, and learning styles exploit that witness perfectly.

I want to be straight with you about the most recent evidence, because the people who sold the posters never were. The 2024 meta-analysis did find a small average benefit, and a 2025 analysis by John Hattie and colleagues put properly designed matching studies at an effect so close to zero it rounds to nothing. So the fair summary is this. Decades of looking have turned up an effect that is tiny where it appears at all, inconsistent, and dwarfed by techniques that help every learner. If a drug performed like that, no doctor would prescribe it.

03Why the myth survives

In 2020, Philip Newton and Atharva Salvi pulled together every survey they could find on educators' beliefs. Thirty-seven studies, about 15,000 educators, eighteen countries. Across all of them, 89 percent said students learn better when teaching matches their learning style. The number showed no sign of falling, and among trainee teachers it was above 95 percent. A 2012 survey by Sanne Dekker and colleagues found 93 percent of British teachers and 96 percent of Dutch teachers agreed.

I am not going to sneer at teachers here. Most good teachers believed this at some point, for the same reasons anyone would. It was taught to them by people they trusted, it fits the everyday experience that people are different, and it carries a lovely message: every child can succeed if we just find their channel.

The failure is institutional, and it deserves to be named bluntly. When William Furey looked in 2020, learning styles appeared in teacher-licensure preparation materials in 29 American states, and a review of teacher-prep programs found two thirds required trainees to account for learning styles in lesson plans. Inspection and observation frameworks rewarded it. Publishers and consultants sold it. Schools advertised it because parents liked the sound of it. Every layer of the system that should have checked the evidence amplified a nice idea instead, and children paid for it.

They pay in two ways. First, the label becomes a ceiling. Many educators believe learning styles are inherited and fixed for life. Tell a ten-year-old they are an "auditory learner" and you have quietly told them that reading, the skill that opens every other door in school, belongs to somebody else. Second, they pay in lost time. Every hour spent making three versions of a worksheet for three imaginary kinds of brain is an hour not spent on what works for every brain in the room.

04The kernel of truth

There is something real buried in the myth, which is part of why it is so hard to kill. Some content genuinely is best taught in a particular format. Geography needs maps. Pronunciation needs sound. A tennis serve needs a racket. But in every case the format is chosen by the content, for every learner. The map is visual for everybody.

Even better, one of the most reliable findings in the field is that combining words with well-chosen pictures helps nearly everyone. Richard Mayer's multimedia research and dozens of studies since keep landing in the same place. A labeled diagram with a clear explanation beats either one alone, for so-called verbal learners as much as visual ones.

And yes, some students genuinely do better hearing text read aloud. A student with weak decoding or dyslexia may get far more from an audiobook than from print. That is an accommodation for a measured skill gap. The right response is to provide the audio while also working on the reading. Calling that child an "auditory learner" makes a fixable gap sound permanent.

05What works for everyone

Good cooks are remarkably consistent about fundamentals. Sharp knife, hot pan, salt early, taste constantly. Nobody has a personal "sautéing style." Learning has fundamentals like that, and the evidence behind them is about as strong as anything in education gets.

In 2013, John Dunlosky and four colleagues rated ten common study techniques on how well each held up across different students, subjects and settings. Only two earned their top rating: practice testing (pulling information out of your head by quizzing yourself) and distributed practice (spreading study across several days instead of cramming). A later analysis of 242 studies by Gregory Donoghue and John Hattie put the same two at the top. A meta-analysis of 222 classroom studies, covering more than 48,000 students from primary school to university and eighteen subjects, found quizzing reliably raised achievement. In real eighth-grade science classes, low-stakes quizzes with feedback lifted unit exam scores by 13 to 25 percent, and the gains held through to the end of the year.

The techniques at the bottom of the list were rereading, highlighting and summarizing, which happen to be what most teenagers do the night before a test.

Here is why students resist the good methods. They feel worse. Spacing means you have partly forgotten things by the next session. Self-testing exposes what you do not know. Mixing problem types is slower and more frustrating than doing twenty of the same kind in a row. Robert and Elizabeth Bjork call these "desirable difficulties," and in study after study, students rate the comfortable method as more effective even when their own scores show the opposite. In a 2006 experiment by Henry Roediger and Jeffrey Karpicke, one group read a passage about fourteen times while another read it about three times and spent the rest of the session trying to recall it. A week later the recall group remembered far more, 61 percent against 40, and the rereaders had been the most confident of all.

So the two myths rhyme. Learning styles feel right because your preferred format feels smooth. Rereading feels right because familiar text feels like knowledge. In both cases, the feeling of fluency is lying.

06Five things worth measuring

Throw out the style quiz, and you are left with a better question. Children are not all the same, and every teacher knows it by the second week of term. So which differences matter? The research points to five. Every one of them is measurable, and every one of them can be changed.

Prior knowledge. What a student already knows is one of the strongest predictors of where they end up. A 2022 review by Bianca Simonsmeier and colleagues, pooling nearly nine thousand results, found prior knowledge correlates strongly with final knowledge (around r = 0.53) but barely predicts how much students gain. Students who start behind can learn just as fast. They simply start further back. In a classic 1988 study by Donna Recht and Lauren Leslie, middle schoolers who knew a lot about baseball understood a passage about a game far better than those who did not, regardless of reading ability. Weak readers who knew baseball beat strong readers who did not. One caution runs the other way: heavy step-by-step scaffolding that helps a beginner can slow down a student who already knows the material, a pattern researchers call the expertise reversal effect.

Reading skill. In secondary school almost everything arrives through text. A student who reads slowly or decodes with effort is fighting on two fronts in every subject, the content and the page. Reading is a skill, trainable at any age. It is never a style.

Executive function. This is the brain's management team: working memory, tracking deadlines, starting tasks, switching between them, resisting the phone. Some students genuinely struggle here, and for some there is an underlying condition such as ADHD that deserves a proper evaluation.

Self-regulation. Can the student plan their study, check whether it is working, and adjust? Most teenagers cannot at first, because nobody taught them. This is the most hopeful item on the list. A 2018 analysis of 48 self-regulated learning programs found gains that actually grew at follow-up, with students from low-income families benefiting most over time. The UK's Education Endowment Foundation rates teaching metacognition and self-regulation at around seven months of extra progress, at very low cost.

Test anxiety. Some students know the material cold and still come apart in the exam room. Ray Hembree's 1988 synthesis put the relationship between test anxiety and performance around r = -0.24, substantial for a single factor, and a 2018 meta-analysis by Nathaniel von der Embse and colleagues found the same pattern, strongest for the worrying, catastrophizing side of anxiety. The best study technique doubles as the best treatment, because frequent low-stakes practice testing makes the real test feel familiar.

Look at the right-hand column. Not one row says "use a different learning method." Every row says "give more support for the same method." Retrieval, spacing and explanation stay exactly where they are. These five differences decide how much scaffolding goes around them.

That is why I refuse to accept that some children simply cannot learn. I have never met one. I have met plenty who were handed the wrong approach, or no reason to care, or a label instead of help. That is on us, not on them. When a student is struggling, the honest question is never "what kind of learner are they?" It is "which of these five have we not addressed yet?"

And when difficulty persists despite good teaching and real effort, with reading that stays labored, math facts that never become automatic, or disorganization in every setting, ask for a proper evaluation. An online quiz cannot diagnose dyslexia or ADHD. A qualified evaluator can, and in most systems the school is obliged to help you get one.

07The subject changes the recipe

Parents are right that studying should look different for math than for history. They usually just have the reason wrong. You do not cook a fillet of fish the way you cook a brisket. One needs four minutes over high heat, the other twelve hours of low smoke. The cook is the same. The ingredient dictates the technique.

Learning works the same way. The core engine applies to everything. What changes from subject to subject is what you retrieve and how you practice it.

Math is about choosing and executing methods, so practice has to be problems, and the problems need to be mixed. In a preregistered trial with 787 seventh graders, students whose homework mixed problem types scored 61 percent on a test a month later, against 38 percent for students who practiced one type at a time. Mixing forces students to work out which method a problem needs, which is exactly what an exam tests and chapter-by-chapter homework never practices.

Science is about mechanisms. Quiz yourself on the facts, but also redraw diagrams from memory and explain them out loud. A student who can draw the nephron from a blank page and talk through it understands it. A student who has highlighted the textbook diagram three times does not.

History is about evidence and argument. Spaced flashcards for dates and names, yes, but the real work is building cause-and-effect chains from memory and practicing with actual documents. Sam Wineburg's research at Stanford showed that historians read sources completely differently from students, asking who wrote it, when and why before asking what it says. The Reading Like a Historian curriculum built on that work improved historical thinking, factual knowledge and reading comprehension.

English and writing are skills built through practice and feedback. Explicit instruction in planning and revising is the strongest finding in the large Writing Next review of adolescent writing.

Languages are about retrieval and output. Daily spaced flashcards in both directions, then listening, reading and speaking every day. Nobody thinks their way to fluency.

The gaps deserve honesty. The best math mixing trial was in seventh grade, so applying it to calculus is a reasonable inference rather than direct proof. Much of the classroom quizzing evidence comes from students aged eleven to fourteen. Research on study methods in the arts is thin. Where the evidence runs out, I would rather say so than pretend.

08What to do on Monday

If you are a parent

• Retire the learning-style label without a fight. If your child says "I'm a visual learner," say "Maybe. Let's test it," and run a two-week experiment with self-quizzing.
• Ask three better questions. What does my child already know in this subject? How well do they read? Do they plan their study, or just do it?
• Help with the routine, not the homework. Talking about goals, expectations and study strategies helps. Hovering over homework often does not.
• Protect sleep and move the phone out of the room during study. Neither is glamorous. Both set the ceiling on everything else.

If you are a student

• Stop rereading. Close the book, write down everything you remember, then check. The gap between what you wrote and what was there is exactly what you need to study.
• Spread it out. Four 30-minute sessions across a week beat one two-hour session the night before, for everyone.
• Expect the good methods to feel harder. That feeling is the learning happening, not a sign you are doing it wrong.
• If reading, focus or nerves are getting in the way, tell someone. They are real, measurable and fixable.

If you are a teacher or school leader

• Strike learning styles from professional development, observation rubrics and the school website.
• Build low-stakes retrieval into every lesson. Five questions at the start, on last week's content, no grades. It is the cheapest high-impact change available.
• Screen for the five things that matter, starting with reading level and prior knowledge.
• Teach study skills explicitly, and warn students in advance that the better methods will feel worse.