Mento
Why Mento?GuidesReviewersPricingLoginSign Up
Sign Up

Guides · Evidence

You are not a visual learner

Matching teaching to a student's preferred style has been tested repeatedly and does not work. Believing it does can actively cost you.

Mento · 27 July 2026 · 7 min read

On this page

  1. 01What the claim actually is
  2. 02The evidence
  3. 03Why it survives anyway
  4. 04The part that actually costs you
  5. 05What to do instead
  6. 06What this means for Mento

In short

  1. 01People really do have preferences about how material is presented. That part has never been in dispute.
  2. 02What fails is the meshing hypothesis: that teaching to someone’s preferred style improves their learning. Studies built to detect that effect do not find it.
  3. 03The cost is not the belief. It is that a style label hands you a respectable-sounding reason to skip the two techniques that do work.

Almost everyone can tell you their learning style. Ask a room of students and the hands go up immediately: visual, auditory, hands-on. It is one of the most confidently held beliefs in education, and one of the most thoroughly tested.

It does not survive the testing.

What the claim actually is

It helps to be precise, because two different claims get bundled together and only one of them is wrong.

The first claim is that people have preferences about how material is presented. That is true and uncontroversial. Ask someone whether they would rather read a diagram or a paragraph and they will have an answer, and it will be reasonably stable over time.

The second claim is the one everything else rests on, and it is called the meshing hypothesis: that matching instruction to a learner’s preferred style produces better learning than mismatching it. Not that people have preferences — that teaching to those preferences works.

That second claim is what fails.

The evidence

Testing meshing properly requires a specific and slightly awkward study design: you have to take learners of at least two different styles, randomly assign each of them to at least two different modes of instruction, and test everyone on the same material. Only that crossover design can show the interaction the theory predicts.

Figure 1

LearnerGiven visual instructionGiven verbal instruction
Prefers visualPredicted: best result
Found: no reliable advantage
Predicted: worse result
Found: no reliable penalty
Prefers verbalPredicted: worse result
Found: no reliable penalty
Predicted: best result
Found: no reliable advantage

Swipe the table for the rest →

What the meshing hypothesis predicts, and what studies using this design report. The theory does not merely predict that some people do better with diagrams — it predicts that the diagonal beats the off-diagonal, for both groups at once. That crossing pattern is the thing that has to show up, and it is the thing that does not.

A commissioned review went looking for studies with that design. It found very few. Among those it did find, several results ran directly against the hypothesis, and the review concluded there was no adequate evidence base to justify using learning-styles assessment in general educational practice.

Pashler, McDaniel, Rohrer & Bjork (2008), Learning Styles: Concepts and Evidence, Psychological Science in the Public Interest.

The direct experimental tests since have kept coming up empty. Matching instruction to VARK-assessed type produced no comprehension benefit in one randomised study; in a chemistry-education replication, students who studied in ways aligned with their assessed type did not do better than those who did not.

Rogowsky, Calhoun & Tallal (2015), Journal of Educational Psychology; Husmann & O’Loughlin, Journal of Chemical Education.

Why it survives anyway

Three reasons, and none of them is that people are gullible.

The first is that the preference is real. You genuinely do prefer the diagram. It is a small step from “I prefer this” to “I learn better from this”, and nothing in ordinary experience corrects the step, because you never get to run the counterfactual on yourself.

The second is that it is a kind explanation. If you are struggling, a style mismatch is a far more comfortable diagnosis than the alternatives, and it comes with an obvious remedy.

The third is that it is enormously convenient to sell. A learning-style quiz is cheap to build, feels personalised, and produces a result the user recognises. Almost every study app has one.

The part that actually costs you

If this were merely a harmless myth, it would not be worth an article. The problem is what the belief licenses.

Once a student accepts a label, the label becomes a reason to avoid things. I am not a reading learner, so retrieval practice is not for me. The techniques with the strongest evidence behind them — testing yourself, spacing your practice — are effortful and feel worse in the moment than re-reading does. A style label hands you a respectable-sounding excuse to skip precisely those.

There is a second cost that is easy to miss. A style is a fixed trait, and framing ability as a fixed trait cuts against the belief that effort changes outcomes.

“I am a visual learner” is a sentence about what you are, not about what you did.

What to do instead

The useful move is to stop sorting yourself by modality and start sorting your methods by evidence. Two techniques come out well ahead of the rest, across ages, subjects and abilities.

  1. 1

    Test yourself instead of reviewing

    Close the notes and try to produce the answer. It will feel worse than re-reading and it works considerably better. There is a whole guide on how to do this properly.

  2. 2

    Spread practice out instead of massing it

    The same total hours, distributed across days, beat the same hours in one sitting — reliably enough that both techniques were rated high-utility in a large review of study strategies, while highlighting and re-reading were rated low.

Dunlosky, Rawson, Marsh, Nathan & Willingham (2013), Improving Students’ Learning With Effective Learning Techniques, Psychological Science in the Public Interest.

What this means for Mento

We do not ask what kind of learner you are, and there is no plan to. No style quiz, no visual-learner badge, no modality tag on your account.

There are real differences between learners that we do act on — how much you already know about a topic, how well-calibrated your confidence is, how much has decayed since you last practised. Those are measurable from what you actually do, they change with effort, and adapting to them has evidence behind it. A self-reported style has none of those properties.

Diagrams alongside text are good, incidentally — for everyone. Presenting material in more than one form is well supported. It is the sorting of people into types that the evidence does not license.

Mento turns your own notes into practice questions, with graduated hints instead of answers and spaced review so what you learn stays learned.

Start Free — No Card Required

Keep reading

  • Review center or self-review? How to actually decideBoard review · 8 min read
  • Reviewing for a board exam while working full-timeBoard review · 7 min read
  • How to use a LET reviewer without wasting your timeLET review · 7 min read
  • The topics your reviewer never mentionsMethod · 9 min read
  • Preparing for UPCAT, USTET and the other entrance examsEntrance exams · 8 min read
  • Retrieval practice, and how to actually do itMethod · 8 min read
  • Preparing for the Philippine Bar, without drowning in codalsBar review · 9 min read
Mento

Exam readiness and research tools for Philippine students.

Product

  • Pricing
  • Why Mento?
  • Reviewers
  • Value Proposition
  • Research Tools
  • Sample Size Calculator

Learn

  • Guides
  • About Mento
  • How Mento uses AI

Support

  • raymund@mento.cloud
  • Contact Us
  • Terms of Service
  • Privacy Policy
  • Refund Policy

Connect

  • Facebook
  • Instagram
  • LinkedIn

Mento is an independent product and is not affiliated with, endorsed by, or sponsored by OpenAI, Anthropic, Google, DeepSeek and Perplexity, or any other AI provider. Mento uses third-party AI models to generate study content, which can be inaccurate or incomplete — always check important answers against your course material. How Mento uses AI

© 2026 Mento. All rights reserved.