Willingham: No evidence exists for learning style theories

By Daniel Willingham. The idea that we have in hand a learning styles theory that can be used to improve instruction is remarkably well ingrained. But a new journal article supports a conclusion I have made in the past: There is no evidence supporting any of the many learning style theories that have been proposed. This should raise serious questions about teaching training.

This month another article was published reviewing the scientific literature on learning styles. It appeared in a journal called Psychological Science in the Public Interest, published by the Association for Psychological Science.

This journal has an interesting premise. The editor recruits three or four top researchers to review the scientific literature on a complex topic of public import. The researchers must be knowledgeable, but not directly involved in prior research on the topic, so that they will be impartial.

The straightforward conclusion matched the one that I have drawn in the past—there is not evidence supporting any of the many learning style theories that have been proposed.

As my previous posting on this subject did, blog postings about this article attracted comments that were sometimes highly negative, and which I think constitute a good argument for the need for greater emphasis on critical thinking skills in the blogosphere.

Here are four common complaints, along with my response.

(1) Scientists are always changing their minds about everything. Just wait a few years, and they will say that learning styles do exist. Unlikely. Mainstream scientists have proposed and tested learning styles theories but there has never been a body of data that they thought reflected learning styles.

(2) No one has proven that learning styles don’t exist. We just don’t have data yet showing that they do. Of course. One can never prove a negative. Learning styles might exist. So might the Loch Ness monster and the Yeti.

For a researcher, one has to wonder whether it’s worth the expense to keep looking for something that no one can find. For a teacher, you have to ask whether “it’s not proven that it doesn’t exist” is good enough to bring a practice into a classroom.

Imagine your doctor prescribing a patient medicine, and when you ask about its effectiveness your doctor shrugs and says, “No one has proven that it doesn’t work.”

(3) I know that there are learning styles. [Insert story here about oneself, one’s child, one’s students, etc.] It’s so obvious! There is a reason that people use the scientific method to address complex questions: It’s hard to keep track of all of the variables that might be involved, or even to keep track of all the outcomes. You have to be systematic about it. That’s basically what the scientific method forces you to do.

Is that really necessary? Shouldn’t it be obvious whether or not people have learning styles? For a couple thousand years it wasn’t obvious to physicians that bloodletting didn’t work. When there are lots of factors contributing to outcomes, you really need to do research.

(4) Learning styles exist, but scientists can’t find evidence for them because they are too rigid about it.

It’s not that every child has one style that applies to every task. Everyone uses combinations of styles, and figuring out a child’s style and how it relates to their work is more of an art than a science. Scientific theories do need to be specific enough that they can generate predictions.

If you can’t write down on a piece of paper. “Under conditions X with person Y, Z ought to happen,” it’s not a scientific theory.

That’s not a problem—not every practice in a classroom needs to be based on a scientific theory–but we might as well be plain about what is scientifically supported and what is not.

The idea that we have in hand a learning styles theory that can be used to improve instruction is remarkably well ingrained. This should raise serious questions about teacher training.