Make it stick

Learning Is Misunderstood

Matt Brown was flying a twin-engine Cessna 401 northeast from Harlingen, Texas on a night freight run to Kentucky when the oil pressure on his right engine started bleeding away. He was at eleven thousand feet. He ran through the math without looking anything up. He shut down the right engine, feathered the prop to cut drag, pushed the left engine harder, held opposite rudder to keep the plane straight, and limped ten more miles to a runway. He landed by making a wide left-hand turn — the only direction the plane could still bank without falling out of the sky, because the dead engine was on the right.

That sequence came out of memory. Not a checklist on his lap, not a radio call to dispatch, not a manual he could thumb through. Trained reflex, retrieved cold, under stress. And that, Peter Brown, Henry Roediger, and Mark McDaniel argue in *Make It Stick*, is the entire point of learning. Knowledge that won't show up when you need it is just decoration.

The hard news in this book is that most of us are bad at producing this kind of usable knowledge, and we are bad in a specific, predictable way. The methods we trust most — rereading the chapter, highlighting the textbook, drilling one skill until it feels smooth, cramming the night before — are some of the least effective approaches cognitive science has ever measured. The methods that actually work feel slower and harder while you're doing them. That's the trick, and that's the reason we avoid them.

Roediger and McDaniel are cognitive psychologists at Washington University in St. Louis; Peter Brown is the writer who pulled their research, and the research of about a dozen colleagues, into one book. The collaboration began with a 2002 grant from the James S. McDonnell Foundation to apply cognitive psychology to real classrooms. The book is the public report.