Believe it or not, teaching kids it’s ok to fail may actually help them learn. It seems that trying, and failing to retrieve an answer, is actually helpful to learning. If this turns out to be true, then “failing to learn” could take on an entirely new meaning.

In the article Getting It Wrong: Surprising Tips on How to Learn, authors Henry L. Roediger and Bridgid Finn, discuss research that shows, “People remember things better, longer, if they are given very challenging tests on the material, tests at which they are bound to fail.”

The study by researchers Nate Kornell, Matthew Hays and Robert Bjork has turned the idea of errorless learning (the old drill-em-and-kill-em) on its ear. Their findings, Learning, Memory and Cognition in The Journal of Experimental Psychology, demonstrate that students learn better when conditions lead them to make errors. Their results show that, “if students make an unsuccessful attempt to retrieve information before receiving an answer, they remember the information better than in a control condition in which they simply study the information.”

So how can students and teachers use failure to improve the way we learn? One idea is for students to attempt to answer the questions at the end of a chapter before they start reading. If the book doesn’t contain questions, students can follow the SQ3R method. Since 1946, this approach has been successfully adopted by students to formulate their own questions as they preview, or survey, a chapter. Teachers can help create opportunities to fail in the classroom by giving pretests prior to lectures and assigned readings. Both previewing and pre-testing direct the students’ attention toward critical concepts.

If this somehow seems counterintuitive, consider the following two experiments. In the study, The Pretesting Effect: Do Unsuccessful Retrieval Attempts Enhance Learning? [pdf], some students were given a pre-test on an essay they had not read, while others read the essay and either had their attention directed toward critical ideas with italics and bolding or were allowed additional time to study. The students who guessed first performed 10 percent better than those who had their reading directed or who studied extra. In a different experiment by Kornell, Hays and Bjork, students were to be tested on “weak pairs” of words. The first group was given 8 seconds to guess the other word in a given pair (the probability of guessing correctly was between 2 and 3 percent). Regardless of their guess, this group was then given 5 seconds to see the correct combination, such as star-night. The other group was given 13 seconds to memorize each set of paired words. As in the first experiment, the students who were challenged to guess the answers performed about 10 percent better.

I’m starting to think my kids already had this figured out for themselves. For years, my older son has driven us crazy by trying to solve math problems his way before following the steps shown in the textbook. As for my younger son, I cringe to think of the number of times I’ve begged him to study a spelling list or review his notes before asking me to quiz him. Each time, he refuses, saying he “learns it better” when he gets the answer wrong the first time.

Following their experiments, the authors concluded that “unsuccessful tests are helpful, not hurtful,” as long as feedback is provided. Their findings suggest “educators and learners should introduce challenges into learning situations, including using tests as learning events, even if doing so increases initial error rates.”