The Science of Successful Learning

Monday, April 21st, 2014

Peter C. Brown, Henry L. Roediger III, and Mark A. McDaniel summarize the science of successful learning — which, in modern Internet fashion, garners the headline, Ditch the 10,000 hour rule! Why Malcolm Gladwell’s famous advice falls short.

Malcolm Gladwell’s 10,000-hour rule is often mischaracterized. First, it’s not Malcolm Gladwell’s rule; he simply popularized the work of Ericsson et al. Second, the rule is not that 10,000 hours of practice will guaranty mastery of a skill; it’s that those who had mastered a skill had spent an average of 10,000 hours in deliberate practice, often on their own, while second-tier experts had spent their time playing their instrument (or whatever) in not-so-deliberate practice.

So, that 10,000-hour cut-off isn’t hard and fast.

Anyway, the science of successful learning suggests that many popular training tactics improve performance in the very short term while reducing the amount of learning over the long term. So, if you practice the exact same thing, over and over, in dedicated training sessions, you will see clear improvement from the start of each session to the end of each session — at least until you get tired and bored — but you’ll improve less than if you if interleaved multiple skills into each session, spaced out your training, varied each challenge a bit, and so on.

Robin Hanson doesn’t see modern teaching methods as a mere misunderstanding of the science of successful learning:

If school’s purpose were to develop skills, we’d teach differently.


So, a good test of a theory of school is: how long do you predict it will take teachers to learn this lesson? The article above talks about how many coaches have learned this lesson, plausibly because they really do want to win games, and face strong competitive pressures.

If you think the main function of schools is something other than learning, you might think it could take a very long time before schools adopt these practices. If you think the main function of schools is learning, but that public schools face much weaker pressures to be efficient that private schools, you might predict that private schools will adopt this much faster. If you think public schools are effective at adopting better approaches, you might predict that they adopt these quickly.

Leave a Reply