Learning styles: what the evidence really says#
Ask a room of people how they like to learn and you’ll get clear answers. One reaches for diagrams, another wants to talk it through, a third has to do it with their hands. Those preferences are real, you can feel them, and almost everyone has them. So it’s tempting to assume the next step follows automatically: teach each person in their preferred style and they’ll learn better. That step is where the popular story goes wrong — and it’s worth getting right, because believing the wrong version quietly wastes effort that could go into methods that actually work.
I want to be straight with you here, even though it means correcting earlier editions of this manual. Your modality preferences are genuine. What the evidence does not support is the idea that matching instruction to your preferred style improves how much you learn. Let me unpack that, because the distinction is the whole game.
Preferences are real — “matching” is the myth#
The claim under the microscope has a name: the meshing hypothesis. It says that if you label yourself a “visual learner” and then receive material visually, you’ll learn more than if you’d received it some other way. It’s an appealing idea, and it’s testable. To confirm it, you’d need to show a crossover interaction: visual learners outperforming when taught visually and underperforming when taught verbally, with auditory learners showing the mirror-image pattern.
When Pashler, McDaniel, Rohrer and Bjork went looking for studies designed well enough to show that crossover, they found almost none — and the ones that met the bar did not produce the predicted result (Pashler et al., 2008). Later reviews reached the same place: Willingham, Hughes and Dobolyi (2015) concluded the theories lack scientific support, and a large survey by Nancekivell, Shah and Gelman (2020) showed how widely people believe the matching idea even though the experimental backing isn’t there. So the summary is narrow and precise:
- Your preferences exist. People reliably report a preferred way to take in information, and those reports are stable.
- Matching to the preference doesn’t reliably raise learning. No credible body of experiments shows the meshing crossover.
- It’s not “all a total myth,” either. The preferences are real and worth knowing. What fails is the specific instruction-matching claim built on top of them.
Why does the belief persist so stubbornly? Partly because matching feels right — material in your favoured format is more pleasant and feels easier, which we mistake for “I’m learning more.” But feeling fluent and learning durably are different things, and they often pull in opposite directions (more on that under desirable difficulties ).
Where the brain-region idea came from — and what it actually shows#
Earlier versions of this page told you that each learning style “lives” in its own region of the brain, and that brain imaging had pinned down a dedicated area per style. That over-claimed, and I’m walking it back.
Here’s the kernel of truth it was built on. Different kinds of processing do recruit characteristic regions — the occipital lobes handle vision, the temporal lobes handle sound, motor and cerebellar areas handle movement. That much is solid neuroscience. But that is a fact about content, not about learners. Everyone’s visual cortex lights up for a picture; everyone’s auditory cortex lights up for a sound. It does not follow that some people have a “visual brain” that learns better when fed pictures. The imaging shows where a modality is processed, not that matching a person to a modality improves their learning. Stretching the first fact into the second was the mistake.
I also based the original Memletic Styles partly on Howard Gardner’s “Multiple Intelligences,” and presented it as a validated taxonomy of the mind. In fairness to the evidence, I shouldn’t have. Gardner’s framework has been influential and is a useful vocabulary for talking about the many things people are good at — but it has not held up as an empirically established set of separate, brain-localised intelligences, and Gardner himself never proposed it as a recipe for matching teaching to a learner’s profile. Treat it as a helpful way to name strengths, not as proven biology.
The lever that actually works: go multimodal, then use proven methods#
If matching isn’t the win, what is? Two things, and they’re better news than the myth.
Use multiple modes — for everyone, not “your” mode for you. Most material has a best representation that has little to do with who’s looking at it. A map is better seen than described; a melody is better heard than read; a procedure sticks when you actually perform it. And combining a verbal explanation with a matching picture beats either alone, for essentially all learners, because the two channels reinforce each other — this is the well-supported principle behind dual coding . So the parent advice still stands, just for a sounder reason: if you’re explaining the water cycle, show the diagram and narrate it and have them sketch it. Not because some students are “visual,” but because the topic and human memory both reward it.
Spend your effort on methods that move the needle. The biggest gains in learning come not from the channel but from what you do with the material: retrieving it from memory rather than re-reading, spacing your practice over time, mixing related problems, and explaining ideas in your own words. These work across people and subjects, and they’re where your energy belongs. You’ll find them in proven learning methods .
So keep the seven Memletic Styles ahead — they’re a friendly way to notice your preferences and to make sure your study isn’t stuck in one gear. Just hold them with the right grip: as a prompt to vary how you engage, not as a label that dictates a single correct format for “your kind of learner.”
Quick takeaway#
Know your preferences — then don’t obey them. Pick the representation the material deserves, combine modes whenever you can, and pour your real effort into retrieval, spacing, and elaboration. That’s the version the evidence backs, and it’s the version that gets you further.