Learning Styles Don't Hold Up — Here's What Does

01The Myth That Won't Quit

Ask most teachers, trainers or managers about learning styles and you'll hear the same thing: some people are visual learners, others auditory, others kinaesthetic. Identify which type you're dealing with, match your instruction to it, and learning improves. It sounds reasonable, it feels intuitive, and it is almost certainly wrong.

The VARK model — Visual, Auditory, Read/Write, Kinaesthetic — is the most widely recognised version, developed by Neil Fleming in the late 1980s. Dozens of similar frameworks followed. By the 2000s, learning-styles theory had become one of the most widely held beliefs among educators globally, embedded in training programmes, school inspections and professional development courses.

The problem is the evidence. For learning-styles theory to be useful, two things must be true: people must reliably sort into distinct categories, and matching instruction to those categories must improve outcomes. Research has tested both claims repeatedly, and neither holds up. A landmark review published in Psychological Science in the Public Interest in 2008, led by Harold Pashler and colleagues, found that the meshing hypothesis — the idea that matched instruction outperforms mismatched instruction — had essentially no rigorous empirical support. Subsequent reviews have reached the same conclusion. People do have genuine preferences, but preference is not the same as learning advantage.

The persistence of the myth is itself a psychological story. Learning-style labels feel validating. Students and trainees enjoy receiving them, which generates positive feedback for the frameworks that produce them. Teachers find the categories practically intuitive. These social and motivational effects create the impression that the approach works, long after the experimental record says otherwise.

02What the Evidence Actually Supports

Discarding learning styles does not mean abandoning differentiation or good instructional design. It means replacing a weak framework with methods that genuinely move the dial.

Retrieval practice is among the most robust findings in the learning sciences. The act of pulling information from memory — through quizzes, self-testing, or low-stakes recall exercises — strengthens retention far more than re-reading or highlighting. The testing effect is not subtle: the advantage over passive review is large, consistent across age groups and content domains, and replicates reliably. Designing instruction so that learners retrieve frequently is one of the highest-leverage changes a teacher or trainer can make.

Spaced practice is equally well-evidenced. Distributing learning across time rather than massing it into a single session produces dramatically better long-term retention from the same total study time. The spacing effect has been documented since Hermann Ebbinghaus's work in the 1880s and has been replicated in classroom settings, workplace training and language learning. One implication: a module delivered once and never revisited will fade faster than the same material spaced across several sessions.

Interleaving — mixing different problem types or topics within a study session, rather than blocking similar items together — produces stronger transfer and longer retention, despite often feeling more difficult in the moment. The subjective difficulty is part of the mechanism: the extra cognitive effort strengthens encoding. This runs directly counter to what learners typically prefer, which is why it is frequently underused.

Worked examples and concrete encoding matter too. There is strong evidence, rooted in cognitive load theory, that novices benefit from studying fully worked examples before attempting problems independently. Starting with practice before the underlying structure is understood generates unnecessary struggle that crowds out learning. As competence grows, the balance shifts: more independent problem-solving and less direct modelling.

Finally, feedback quality shapes whether practice improves performance or merely confirms existing errors. The evidence on feedback is nuanced — timing, specificity and whether feedback targets strategy rather than self all affect outcomes — but the consistent finding is that learners need accurate, actionable information about where their thinking went wrong, not just whether the final answer was right.

None of these methods requires knowing a learner's preferred style. All of them require knowing the material, the stage of learning and the goal. That's a harder intellectual task than sorting people into categories — but it's the task that actually works.

03Who did the work