A New Way to Learn: Students Argue About Physics
If you walk into Eric Mazur's Harvard physics
class at the wrong time, you may think you're in a lounge instead of a
lecture hall. Students argue furiously. They're loud. They're unruly.
None of them pays attention to Mazur.
"It's absolute chaos," he admits. "It's great."
The students are arguing about physics. Every 10 minutes Mazur's class
has a ConcepTest--an on-the-spot, ungraded quiz that pinpoints an underlying
concept of physics.
Students record an answer of A, B, C, or D, then the arguments start. "I
tell them, turn to your neighbor and convince your neighbor of your
After a few minutes, the same question is put back up. Freshly convinced
students record their answers in a second round of testing. Mazur can
see their responses as they are entered into a computer; an occassional
show of hands works as well. If 50 percent of the students get the
answer right the first time, the second showing is closer to 90 percent.
In 1995, Mazur's ConcepTest System was chosen to be part of NSF's Profiles
of Innovative Projects, a booklet by the Directorate for Education
and Human Resources highlighting 50 of NSF's education ventures.
Each project addresses a generally accepted problem in classrooms.
In Mazur's case, the problem was getting his students to stop memorizing
and start thinking. Explaining a concept forces students to grasp the
issues, he says. "It's not hands-on, it's brains-on."
What's more, says Mazur, frequently students are better able to get
the concept across than he is. "Their answers are fresh and often more
effective than my polished explanations."
After the second round of tests, Mazur lectures for a few minutes
and sets up another ConcepTest. At the end of each class, Mazur knows
how many students have understood the material. It's a good position
for a teacher to be in, he says. "Once you start this process, you're
hooked. You're in touch with your class. I get nervous now if I lecture
more than 10 minutes."
Mazur didn't always teach like this. When he started in 1984, he worked
in the standard lecture/laboratory format. Then he read an article
claiming that students are memorizing, not conceptualizing science.
The article included a test of concepts. "I thought it was easier than
what I was testing them on," he says. But his class didn't. They did
Mazur rethought the process and used a 1993 grant from NSF's Division
of Undergraduate Education (DUE) to develop the ConcepTest and Peer
Instruction methods. Used originally at Harvard and University of Massachusetts,
Lowell, the method is now widely distributed through seminars and workshops,
free teachers' books published by Prentice Hall, and Mazur's World
Wide Web page.
The distribution and applicability are among the main reasons for
Federal funding from NSF, says Duncan McBride, Program Director in
DUE. "This project offers an improvement not only for Eric Mazur and
Harvard students, but for students all over the country."
For more information about ConcepTests, visit Mazur's education
Web site: http://galileo.harvard.edu