Make It Stick: The Science of Successful Learning

The summary

Eight-year-olds spent twelve weeks in gym class tossing beanbags into buckets. Half threw at a bucket three feet away, every time. The other half practiced only at two and four feet, never three. Then everyone was tested at three feet, and the kids who had never practiced that throw won easily. That is Make It Stick in miniature: the practice that feels like it’s working usually isn’t. Peter Brown, a storyteller, wrote it with the cognitive psychologists Henry Roediger and Mark McDaniel, who spent a decade carrying memory research into classrooms.

Fluency is not mastery

Rereading is the most popular study strategy there is — more than 80 percent of college students in some surveys — and it barely works. In the mid-1960s Endel Tulving found that people who had read a list of words six times went on to memorize them no faster than people meeting them for the first time. Decades later a Washington University team gave 148 students five passages from textbooks and Scientific American; some read once, some twice in a row. There was no rereading benefit at all.

What rereading delivers is a feeling. Familiarity with a text gets mistaken for command of its ideas — the student who highlights his notes, reviews the highlights until it all looks obvious, and pulls a D. A UCLA professor of twenty-five years failed a quiz on the nearest fire extinguisher, then found it inches from his own doorknob.

Retrieval ties the knot

Pull a fact out of memory and you change it: you make it easier to pull out next time. Psychologists call this the testing effect — a test is not a measuring stick but the practice itself.

The authors talked a middle school in Columbia, Illinois into a trial. Short no-stakes quizzes covered a third of the material: one before the lesson, one after, one the day before the exam. Another third was simply restated for review. In eighth grade science, students averaged 92 percent on the quizzed material and 79 on the rest — and the restated third scored no better than material nobody revisited. The gap was still there eight months later. Yet in one survey only 11 percent of college students said they quiz themselves.

Space it, mix it, vary it

Thirty-eight surgical residents took four short lessons in microsurgery. Half did all four in one day, the normal schedule; half took one a week. A month later the spaced group was better on every measure — time, hand movements, success at reattaching the severed aortas of live rats. Sixteen percent of the crammed group damaged the vessels beyond repair and couldn’t finish.

Interleaving works the same way and feels worse. Students learning the volumes of four unfamiliar solids either practiced one type at a time or mixed the problems up. During practice the blocked group averaged 89 percent and the mixed group 60. A week later it inverted: 20 against 63. Cal Poly’s batters who faced fastballs, curveballs and changeups in random order struggled for six weeks and came out ahead of teammates who took fifteen of each in blocks — while insisting, even after seeing the scores, that the blocks had served them better.

The difficulties are the point

Elizabeth and Robert Bjork named these obstacles desirable difficulties, and the Army’s jump school at Fort Benning is built out of them. You carry no notebook. You practice the parachute landing fall in a gravel pit, then at escalating heights, and at the end of the harness run the instructor — not you — pulls the release, so you land from a direction you didn’t choose. When Mia Blundetto, a Marine lieutenant who hated falling, came down tangled in another jumper’s chute on her third jump, she simply swam out of it.

Effort does the work. Recalling something after a lapse forces you to rebuild it from long-term memory instead of echoing it back from short-term, and the rebuilding is what strengthens it. It is also why struggling with a problem before anyone teaches you the solution beats being shown — even when your attempt is wrong, if someone corrects you. B. F. Skinner once campaigned for “errorless learning,” material fed in bites too small to fail on. He had it backwards. When French sixth graders got unsolvable anagrams, the half then told that difficulty is normal used their working memory measurably better on the next test; the others had spent it agonizing.

What actually differs between learners

Not learning styles. In 2008 Harold Pashler, Mark McDaniel, Doug Rohrer and Bob Bjork reviewed the central claim — that you learn better when instruction matches your visual, auditory or kinesthetic preference — and found very few studies built to test it, virtually none validating it, and several contradicting it. What matters is matching the method to the subject: pictures for geometry, words for poetry.

The differences that survive scrutiny are more useful. Carol Dweck taught low-performing New York seventh graders that effort physically rewires the brain; without their teachers knowing, those students went on to outwork classmates who got a lesson on memory instead. James Paterson, a Welsh teacher who competes in world memory championships, took up mnemonics as a shortcut and could walk into exams with every name and date filed away — only to find he had “the mountaintops but not the mountain range.” Mnemonics organize knowledge you already own; they can’t buy it.

The bottom line

Durable learning comes from pulling knowledge out of your head rather than pushing it in: spaced far enough apart that you’ve gone rusty, mixed up enough that you must work out which problem you’re facing, effortful enough that it feels like failing. Your own sense of how well you’re learning is the least reliable instrument you own, so the fix has to be external — a quiz, a coach, a score you can’t argue with. Read it if you study, teach, coach or train anyone, yourself included.