Skip to content Skip to navigation

The Importance of Failure in Learning

Tomorrow's Teaching and Learning

Message Number: 

The university creates an environment that causes fear of failure, which may be acceptable in the business world but not in academia.


The posting below, by Professor David Freyberg of Stanford University looks at the positive role of failure in the learning process. It is from the Stanford Report, May 1, 2002, and is reprinted with permission.


Rick Reis

UP NEXT: Building Intellectual Community Through Courses for Staff

Tomorrow's Teaching and Learning

----------------------------- 784 words ----------------------------


Learning is most effective when it incorporates and embraces failure, scholar says


Failure is an experience that most students and faculty try to avoid -- but not Dave Freyberg. A member of the Stanford faculty for 22 years, Freyberg tackled the subject head-on when he delivered the "Award-Winning Teachers on Teaching" lecture last Thursday.

"I recall my parents saying, as I went off to school, 'Learn from your mistakes,'" said Freyberg, an associate professor of civil and environmental engineering. "That's a shibboleth we all hear, and what happens is, we get to school and we're not supposed to fail."

In reality, he argued, failure is an integral part of the learning process -- or should be -- just as failure is an essential component of engineering.

"Engineers learn how to use failure as part of the design process," Freyberg noted. "We don't want to have dams and bridges that fail -- that's not good. But failure happens. That's why we have R & D [research and development]. The 'D' part is recognition that we will have failure."

The process of design -- failure followed by improvement -- is the essence of engineering, he observed. Likewise, the process of learning is most effective if it incorporates failure: You try something out, and if it doesn't work, you fix it. That approach seems to work best in writing courses where the student continually revises and improves each draft. However, the technique works in very few other courses, said Freyberg, the former associate dean of engineering for undergraduate education and a past recipient of School of Engineering's Tau Beta Pi award for excellence in undergraduate education.

Fear of failure

The university creates an environment that causes fear of failure, which may be acceptable in the business world but not in academia, he added.

"The notion that learning is a one-pass process -- I think the data don't support that. It's a multi-pass process," Freyberg noted.

As an example, he pointed to the "traditional problem set cycle" that occurs in most undergraduate engineering courses. The teacher lectures for a week or two, then hands out a problem set that is due back in a week. It takes another week for the instructor to grade the assignment, and if the student does well, he or she will file it away. But, according to Freyberg, if the problem set returns full of red ink, the student is usually too busy with new assignments to correct it, so the short-term feedback loop between failing and learning to correct your failures breaks down.

"The timing isn't right," he argued, "so I'm not convinced that problem sets lead to a lot of learning."

To create a "positive feedback loop" in his classroom, Freyberg tried an experiment. He decided not to grade problem sets, thus giving students the opportunity to fail on their homework exercises. However, they were required to take exams that covered the same material in the exercises. You're free to fail once, Freyberg told his students, but see if you can learn the material the second time through.

The experiment yielded some surprising results. In preparing for an exam, many students assumed that Freyberg was trying to trick them and chose instead to study material not covered in the problem sets.

"I've had a striking number of conversations with students after the final who've said, 'We didn't believe you.' Perhaps I'm inherently untrustworthy," Freyberg quipped. "A more likely explanation is that I haven't been successful in explaining it."

According to Freyberg, the experiment worked well with freshmen, but failed with seniors -- an "interesting temporal pattern" that he was at a loss to explain.

Esprit de corps

Freyberg eagerly sought advice on how to create a classroom environment that uses failure as a learning tool -- an environment that is supportive enough not to damage self-esteem but that simultaneously discourages long-term failure.

One student suggested that, because Stanford undergraduates have little experience with failure, Freyberg might consider conducting a kind of failure consciousness-raising session in class. Although the comment drew much laughter, Freyberg rejected the myth that Stanford students rarely fail. He also noted that nothing creates classroom esprit de corps better than group projects that fail -- followed by group projects that succeed.

"My own attempts to deal with this fear of failure have largely failed," he conceded. "I'm continuing to believe that the basic notion is right. I'm just not implementing it well."

The "Award-Winning Teachers on Teaching" series is sponsored by the Center for Teaching and Learning. The next lecture in the series by chemistry Professor Eric Kool will be held at noon May 16 at the Hartley Conference Center in the Mitchell Earth Sciences Building. Kool's topic will be "How Can Organic Chemistry Possibly Be Interesting? One Teacher's Answer."