kitchen table math, the sequel: college teaching - class discussion

Monday, July 19, 2010

college teaching - class discussion

Terrific write-up of a paper on college teaching:
Dr McCallum describes a common classroom situation: trying to get the students to actually read a scientific paper. You'd be surprised to learn that this is not as easy as it sounds, because reading and understanding a scientific paper is difficult. In these classes, the professor would assign a paper or series of papers and one student would act as the presentor of the paper while the others were the student "presentees." To ensure that the "presentees" also read the paper instead of passively sitting in the classroom or texting their friends, the professor would either reward or force student questions and discussion. To do this, some faculty awarded points for each question asked. Others asked questions about the paper on exams. Frequently, "participation points" would be awarded to force individual participation.

"In fact, the many different angles used by professors in my many classes all ended the same way," writes Dr McCallum in his paper. "Inevitably, a growing number of students did not read the papers unless they were the presenter."

To engage all his students, Dr McCallum devised and tested an innovative method in the two-hour lab section for his senior-level environmental physiology class. The design worked like this: Dr McCallum brought two copies of 10 different manuscripts on critical thermal maxima. In this case, almost all of these manuscripts were by Victor Hutchison, and they were very similar except for the organism involved. Only one of the manuscripts was a review paper. In a classroom of 20 students or less, each student was given a different manuscript. Only a few students had a duplicate paper, so almost everyone was responsible for their own article. Then, the students were given roughly 15 minutes to read their paper in class. At the end of that time, Dr McCallum asked if everyone was done. If anyone was not done reading, they were given a little more time to finish.

After everyone had finished reading their paper, Dr McCallum randomly asked one student to briefly describe what their paper was about. After they had done this, Dr McCallum asked the other student who had that same paper if the first student's iteration followed their understanding. Invariably, the second student had things to add or ask. Then Dr McCallum randomly asked a student with a different paper to compare what they read in their manuscript to what the other two students presented. If a second student had read the same paper, Dr McCallum then asked that individual if s/he had anything to add. Then, Dr McCallum asked the presenters of the first paper if they felt that the comparison was accurate and to explain why (or why not). Dr McCallum found that continuing this scenario through approximately four papers led to a fluid discussion where only an occasional question from the professor was necessary to stimulate further participation Figure 1:


The result of this model was not only to circumvent many lazy student behaviors, but also to improve reading comprehension by familiarizing students with how to read, process, and evaluate complex scientific manuscripts in a short period of time. To test this hypothesis, Dr McCallum included a 10-point short-answer question on an exam that asked students to discuss the topic of the papers that were discussed in class. He found that almost all of his students had at least a working knowledge of the topic and 65% earned at least 7 points on this question. When compared to the more traditional "presenter-presentee" scenario used during the previous semester, only 33% of the students earned at least 7 points on a similar question.

But even more interesting was that Dr McCallum's students' GRE, MCAT, and ETS major field exam scores increased significantly. For example, the class averages for raw section scores in the ETS major field test in biology increased 12% from the previous year and 50% relative to students taking the same classes with other instructors within one year of introducing this technique. Additionally, the number of students entering graduate and medical school also rose.

Parrots, People and Pedagogies: A Look at Teaching and Education
McCallum, M. (2010). A Method for Encouraging Classroom Discussion of Scientific Papers. Bulletin of the Ecological Society of America, 91 (3), 363-366 

5 comments:

ChemProf said...

It's interesting, but I wonder how extensible it is. It depends on having a group of papers that are distinct but related, so that everyone is picking up the same concepts. I'd have a hard time coming up with a set of articles like this in my classes.

It also seems like a better idea in an area like ecology where the papers are easier to read, in general. I wouldn't expect students to have a good grasp of a physics article in 15 minutes, honestly. I do know that people have had good luck in ecology and developmental biology with having students read scientific articles quite early in their college careers.

cranberry said...

You may need a group of students ready to quickly digest a relatively complex reading assignment, applying existing background knowledge. College is the right time; middle school would be too early?

In a way, it reminds me of the classroom group discussions of translation exercises from Latin texts, back in high school.

Allison said...

I can't imagine this working in many fields.

I don't think I ever read a single paper in my field (quantum computing in grad school) *where I didn't already know the result* that I was capable of really comprehending in 15 minutes. Yes, if I already knew the result, and had heard the argument before, then fine. But otherwise? No. Now, lots of people are brighter than me, but in grad school, there were entire research groups where the prof spent years teaching grad students how to read a paper.

Now, expanding out of my subfield, it depended on the field. I could never have done this for a subfield of physics even slightly different than my own. For some broad algorithm papers in CSland, sure, but all of the techniques presented in the paper would have been known to me, or nearly known. So what are the elements of this kind of paper in these discipline?

I guess this depends on what is meant by "explains". But maybe he was just very good at teaching them how to read a paper in 15 minutes, and how to ignore the stuff that wasn't relevant to his comparison.

Allison said...

On second thought, it may be more about style of the facilitator than the subject matter.

I can't see this working with many professors that I had, certainly. My experience was that most such discussions were painful for nearly everyone in the room, so no one really wanted to take a risk and say anything, whether they'd read ahead of time or not.

That people in the above example were willing to create an actual discussion of what they read without fear of being humiliated says a lot about how much the students trusted the teacher and each other. That's difficult to come by.

Catherine Johnson said...

I had the same thought re: 15 minutes.