Against Learning Styles

[A] The belief that each learner has a style and does best when teaching is matched to it is held by most teachers and has been for forty years. [B] As a claim about instruction it is almost entirely unsupported. The test it requires is straightforward: sort learners by style, teach the same material two ways, and show that each group does better with its own. Studies of that design are rare, and those that exist have mostly failed to find the interaction. [C] Some material is better taught one way whoever receives it, and a diagram helps everybody with a diagram-shaped problem. [D]

The usual explanation, that teachers are credulous, is both unkind and too simple. People do differ, visibly, in what they say they prefer, and a teacher who notices that is noticing something real. The error is in the inference: a preference is not evidence that the preferred method teaches that person more, and the two come apart in the studies that measure both, where learners often prefer the presentation from which they learn least. The lesson is narrower than the debunkers suggest. What should go is the matching, not the attention. A teacher who watches how a particular student goes wrong is doing something the evidence supports; a teacher who sorts thirty of them into three boxes on the strength of a questionnaire is not.