Don't lecture me

Another great radio piece by Emily Hanford (I caught the end of what I assume was just part of it on the NPR afternoon news show on Sunday) here (audio and transcript both there). She reports the research on the effectiveness of lectures in prompting actual learning: not much. Anyone reading who lectures must listen to/read it.  A long excerpt (followed by some comments):

Lecturing was the way just about everyone taught introductory physics. To think there was something wrong with the lecture meant physics instructors would "have to really change the way they do things," says Hestenes. A lot of them ignored his study and kept teaching the way they always had. They insisted their lectures were working just fine. But Eric Mazur was unusual, says Hestenes. "He was the first one who took it to heart." Mazur is a physics professor at Harvard University. He came across Hestenes's articles in 1990, five years after they'd been published. To understand why the articles had such a big impact on Mazur you have to know some things about his history. Mazur grew up dreaming of becoming an astronomer.

"When I was five years old I fell in love with the universe," he says. "I tried to get my hands on to every accessible book on astronomy. I was so excited by the world of science." But when Mazur got to university, he hated the astronomy classes."It was all sitting in the lecture, and then scribbling down notes and cramming those notes and parroting them back on the exam," he says. "Focusing on the details, focusing on memorizing and regurgitation, the whole beauty of astronomy was lost." So he switched to physics. It wasn't as heartbreaking for him to sit in a physics lecture and memorize things. Mazur eventually got a Ph.D. in physics and a job at Harvard University. Like most Ph.D.s, Mazur never got any training in how to teach.



"I just mimicked what my instructors had done to me. I think that's what we all do. So, I lectured." Turns out he loved lecturing. It's a lot more fun being on stage delivering a lecture than it is sitting in the audience watching. And that's exactly what a lecture is, says Mazur: a performance. He decided to make it fun. "Thanks to the setup we have here at Harvard, it was very flashy, like a Hollywood show," he says. "Attention-grabbing demos, me shooting through the lecture hall in a rocket car."Mazur's students apparently loved it. His classes were full and he got great evaluations from the students at the end of every semester. "For a long while, I thought I was doing a really, really good job," he says.

Then Mazur read the articles by Hestenes and Halloun. Mazur's first instinct was to dismiss the results. The test covered such basic material; he was sure his students were learning this stuff. But what if they weren't? How boring it would be to learn physics and never really understand the fundamental concepts that make physics so fascinating. Mazur thought back to his own experience with astronomy; if his students were just memorizing information and solving problems, he had to know, and he had to do something about it. So he gave them the FCI, and he was shocked. "They didn't do much better," he says. "In fact, when they looked at the test that I gave to them some students asked me, 'How should I answer these questions? According to what you taught me, or according to the way I usually think about these things?' That's when it started to dawn on me that something was really amiss."

I am a culprit in the promulgation of the lecture. But this comment by Alan Bostick, alongside the critical comments of Keith M Ellis, haunted me for a long time. I now follow Bostick;s advice whenever possible with, in recent experience, quite spectacular results. I find, though, that asking a single undergraduate to prepare a lecture, is really too much pressure. That may be because they are never asked to do so, but they aren't so I have to live with that. I have usually asked them to prepare in pairs; this past semester, mainly due to my own incompetence, I assigned them in 3s, which seems to have been ideal. In my class of 24 students, not one of the 8 classes which were run by students was less good than the average class run by me, and many were far better than my average (of course, I contributed in discussion, a lot I'd like to think, and I did select the for the first presentation a group of students whom I thought, rightly, would set a standard that would press the others to perform well).

Of course, in a class of 160 I cannot assign lectures to all the students. But I use the guidance that almost nobody can concentrate on even my lecturing for more than 20 minutes, and so break up the classes as much as possible with (I hope well considered) discussion questions, often starting out by asking them to talk to each other, then getting them to discuss in the larger format. I know this doesn't work for all the students, and I know a lot of good students find the ramblings of their peers frustrating, so I do still lecture a fair bit, and exercise a good deal of control in the discussions.

My suspicion is that there is a small subset of students who can and do concentrate for more than 20 minutes; in fact they can concentrate for hours. I was one of those, at least when the lecturer met some low threshold of quality. The reason that many academics in the US system think their own lectures are effective is that they were in that small subset (and their belief is never tested, since they assign the assessments, which are not designed to test their belief in their own effectiveness...). We select for people who are like us.

(Crossposted at CrookedTimber)