Brain Training Doesn’t Work

Will mastering chess make you more strategic? Does playing Sudoku speed up your mind? Do brain teasers help you think more logically?

Sadly the answer is: probably not.

From a 2016 review by Simons et al.:

“[W]e find extensive evidence that brain-training interventions improve performance on the trained tasks, less evidence that such interventions improve performance on closely related tasks, and little evidence that training enhances performance on distantly related tasks or that training improves everyday cognitive performance.”

Another study tracked participants over two years of working memory training. They found that the training had no impact on measured intelligence. The authors concluded, “These results question the utility and validity of [working memory] training as means of improving cognitive ability.”

Likewise, Giovanni Sala and Fernand Gobet performed a meta-analysis on whether studying chess and music affects academic or cognitive skills. They found only minimal effects. Some studies supported benefits from training, but the higher quality the study, the weaker the effect. Summarizing their review, the authors remark that, “this pattern of results casts serious doubt on the effectiveness of chess, music and working memory training.”

It’s easy to see why people are attracted to the idea of brain training. Intelligence is associated with nearly every positive life outcome people experience. A procedure that increases intelligence with only a small amount of daily effort would be life-altering.

Brain training also makes sense if you hold a false (but seductive) view of the mind—the idea that the mind is like a muscle.

The Mind-Muscle Myth

The idea that the mind is like a muscle has a long history. In his famous treatise on education, John Dewey linked the belief to the English philosopher John Locke. But it’s likely much older. The idea is so deeply interwoven in our folk psychology that few people even question it.

The mind-muscle metaphor goes something like this:

• Muscles improve through training.
• Strengthening your biceps by lifting dumbbells, for instance, will make you stronger at lifting groceries, luggage or rocks.
• Mental abilities improve through training.
• Therefore, strengthening your mind through puzzles, for instance, will make you smarter in business, school and life.

Items 1-3 are unproblematic. It’s #4 that gives the mind-muscle metaphor its appeal. It’s also where the analogy breaks down. Unfortunately, training on specific tasks doesn’t make you generally better at many different things.

The earliest takedown of the mind-muscle metaphor dates to Edward Thorndike. In 1901, he began a series of studies that showed practice on quite similar tasks didn’t lead to improvement in unrelated tasks. Thorndike interpreted his results in terms of identical elements: post training, performance improves on tasks that overlap in the stimulus or response required, but not beyond this.

Summarizing his view, Thorndike wrote, “the mind is so specialized that we alter human nature in small spots.”

Psychology has progressed considerably since Thorndike’s day. Yet the idea that skills are specific is a consistent finding in psychological research. In their 1989 monograph, The Transfer of Cognitive Skill, John Anderson and Mark Singley argued for what amounts to an updated version of Thorndike’s identical elements model. Skills transfer to the extent that the knowledge and procedures used between tasks are the same. If skills rely on different methods or ideas, training in one won’t help with another.

Thorndike’s identical elements model, and modern theories such as Anderson’s ACT-R, show why brain training doesn’t work. But are there any other ways to get smarter?

Does Education Boost Intelligence?

Brain-training fails because it focuses on a very narrow kind of task. Judging a good chess position and good business decision don’t use the same procedure. Thus learning strategy in chess doesn’t make you more strategic or effective in business.

Education doesn’t necessarily suffer the same shortcoming because it aims to impart a much broader set of skills. Algebra might only be suitable for problems that use algebra, according to the identical elements model. But there are lots of problems you can solve with algebra! Similarly, learning to read may not transfer (directly) to other skills, but reading can be a gateway to acquiring knowledge in practically any field.

Stuart Ritchie reviewed studies on the impact of additional years of education. He found that an extra year of schooling was typically associated with 1-5 more IQ points. These studies often rely on a quasi-experimental design. The authors studied situations where a sudden, unexpected change in policy resulted in some people getting more education than others. Testing people just before and after the cutoff let them tease out the effect of education without a formal experiment.

The optimistic take on this research would be that education improves general thinking by equipping people with diverse cognitive tools. This breadth has power. Even if a particular task is only helped by a subset of school training, many years of schooling make an overlap between skills and tasks increasingly likely.

The pessimistic stance would be that education trains you at narrow tricks that work for passing tests—sitting still for a prolonged period, guessing well when you don’t know the answer, watching out for trick questions, etc.—and these tricks also help on IQ tests.

Applications for an Identical Elements View of Learning

My perspective is that the only way to become smarter is by learning. The basic units of learning are specific, but when added together, these specific chunks can become impressive proficiency.

A concrete analogy would be language learning. Fluency isn’t a muscle you improve. It results from knowing many words, grammar, and pronunciations and using that knowledge quickly and unhesitatingly. It can be impressive to watch someone at a mastery level converse in a language you struggle with. Still, there is nothing more to it than this—if you knew everything she did, you too would be fluent.

Similarly, intelligence in real life is about having the vocabulary of methods and knowledge to deal with a wide variety of problems. Each unit of learning may seem unimpressive on its own, but combine enough of those units, and the accumulation is wisdom.

But to achieve this possibility, we must let go of the false promise that broad-ranging skills can come from practice on narrow tasks. Brain training is a dead-end, but learning is timeless.