On this page
"I'm a visual learner" is one of the most common things students say about themselves. The idea behind it, that each person learns best when taught in their preferred style, is widespread in schools and coaching. It is also not supported by the evidence. Here is what researchers tested, why the idea feels true, and what personalises learning instead.
What does the learning styles idea claim?
The popular version says people can be sorted into types, such as visual, auditory or kinaesthetic, and that teaching should match each type. Psychologists call this the meshing hypothesis: instruction works best when it matches the learner's preferred format (Pashler et al. 2008).
What did Pashler and colleagues find?
In 2008, Harold Pashler, Mark McDaniel, Doug Rohrer and Robert Bjork reviewed the evidence in Psychological Science in the Public Interest. They first set out what a convincing test would need: sort students by style, randomly assign each group to different teaching methods, give everyone the same final test, and show that the best method for one style is not the best for another.
They found "virtually no evidence" for that pattern. Very few studies even used a design that could test it, and several that did found results that "flatly contradict" the meshing hypothesis. They concluded there was no adequate evidence base to justify building learning-styles assessments into general educational practice. They were also careful: many versions of learning styles have simply never been tested properly.
Has anyone run the test since?
Yes. Beth Rogowsky, Barbara Calhoun and Paula Tallal followed Pashler's design with college-educated adults. They measured each person's auditory or visual-word preference with a standardised inventory, then randomly assigned people to learn from the same nonfiction material as a digital audiobook or as an e-text. Comprehension was tested immediately and two weeks later. There was no significant relationship between preferred style and the format that worked better (Rogowsky et al. 2015).
Why does the idea feel true?
Preferences are real. Pashler's team noted plenty of evidence that people will state how they like information presented, and that people differ in specific aptitudes. A preference is not the same as a better result, though. Comfort can even mislead: material that feels easy to take in is not always material you will remember, as we explain in the illusion of fluency.
The format that works usually depends on the content, not the learner. A circuit is easier to learn from a diagram for everyone; a pronunciation needs sound for everyone. Richard Mayer's research on words and pictures, covered in our post on multimedia learning, is about designing for how all learners process information.
What does personalise learning?
What you already know. This is the difference with strong evidence behind it. Research on the expertise reversal effect shows that methods that help beginners, such as studying worked examples, can lose their benefit or even backfire for students with more experience (Kalyuga et al. 2003). Good personalisation means a different starting point and pace, not a different sense.
Methods that help almost everyone. A major review of ten study techniques rated practice testing and distributed (spaced) practice as high utility, because they benefit learners of different ages and abilities. Rereading and highlighting were rated low utility (Dunlosky et al. 2013).
Putting it together for a JEE or NEET chapter:
Take a short diagnostic quiz before studying, to find your starting point.
Learn new material in whatever format fits the content: diagrams for ray optics, equations and worked steps for kinematics.
Close the book and test yourself.
Review on spaced days, harder questions first.
Move to the next topic when you can answer without notes, not when the hours run out.
This is the idea behind mastery learning and adaptive practice.
Fermi's AI tutor starts from the same place: it works out what you already know, then decides whether to explain, ask or show, writing on a board, drawing the figure or running a simulation. Your next step: drop the label, and take a ten-question quiz on your next chapter before you open the book.
