Flunking Exams Can Be a Good Thing

From the TI archives — originally published September 2014.

Here's a counterintuitive finding from learning science: it's not the studying that cements knowledge so much as the testing. In September 2014, the New York Times Magazine reported on the "pretesting effect" — the finding that taking a quiz before you've studied a topic, failing at it, and then studying makes the material stick far better than studying alone (NYT Magazine).

The study at the center of the piece was a classroom first: the UCLA psychologist Elizabeth Ligon Bjork ran the experiment in her own live classroom, and found that pretesting raised performance on final-exam questions by an average of about 10 percent over a control group. Wrong answers, it turns out, prime the brain — they change how we store the information in the question, so that when the correct answer arrives later, the mind is ready for it. The wrong guess isn't wasted effort; it's a hook waiting for the right answer to hang on.

There's a caveat worth noting: pretesting works when the unfamiliar concepts are embedded in language you can parse. Quiz students on a totally alien notation and the pretest is mostly noise. But within familiar territory, the flunk-first approach outperforms plain rereading — a habit that feels productive precisely because it's fluent, and fails precisely for the same reason.

The practical takeaway holds up a decade later, and it's the reason so many good coaching centers and learning apps are built the way they are: quiz yourself early, quiz yourself often. The test isn't the end of learning. Done right, it's the beginning of it.