Monday, April 10, 2006

Experimenting with new ways of teaching

In yesterday's post, I talked a bit about trying to rely on lecture less and other classroom methodologies more. Today I tried out a new-to-me methodology: an in-class role-playing simulation. And surprisingly, it worked out extraordinarily well! In case you're interested, here's what I did. (I'll try to explain this without giving away too many details about the class itself.)

First, I came up with a problem that was large enough for the students to spend most or all of the class grappling with it, yet small enough to be manageable (and of course, related to the material they're currently learning). The problem was of the "suppose we were to design a system to do X" variety.

Second, I came up with three distinct roles for the students to play. The roles corresponded to three groups of people who would be interested in utilizing a system to do X, each of whom has its own vested interest in the system design. I typed up a description of each role, along with basic instructions for how the students should discuss the problem, and handed these out randomly to the class.

Once the students got their roles, they divided themselves into groups according to role and discussed the problem while "in character". This was the part I was most nervous about---would computer science students play along, or would they roll their eyes and half-heartedly participate? But the students totally got into their roles--and stayed in their roles throughout the class.

After discussing the problem in the role-groups for a while, I had one person from each group write the group's system requirements on the board. This was so the students could see the discrepancies in requirements between the groups, since each group has different goals in using this proposed system.

Then the fun began! Arguments erupted between the groups as to which requirements should take precedence. I tried to stay out of the way as much as possible, and in doing so found that the groups were eventually able to reach consensus on most of the points. After much back-and-forth discussion, we came up with a common list of requirements, along with a list of technical limitations that should factor into the system's design.

It just so happens that my school is in fact looking into systems that do X, and so I ended class by showing the class the requirements document from that real-life project. And we had a frank and interesting discussion about what the real-life requirements document implied about the real-life software selection process, and the fundamental differences between our requirements and their requirements (and thus between the processes).

I have to say that I was completely blown away by what my students came up with in such a short time. It was clear that, by doing this exercise, they had a much more accurate and complete understanding of the material than they would have had I spent the period lecturing about it instead. Moreover, they even tied in material from previous weeks, and did so seamlessly and naturally. I'm a believer now!

4 comments:

ScienceWoman said...

This sounds like a great class.

Anonymous said...

Research shows that activity based learning is more effective than lectures. There is a limit, though. If students have not prepared properly, or don't have any background preparation, then the activity can become a waste of time. It is also pretty tough for the instructor to come up with good activites. However, when it works, it works wonderfully!

Anonymous said...

I've enjoyed reading your posts about teaching. Hopefully my students will benefit too! (I'm teaching my first course soon, Programming Languages and Paradigms. It should be fun!)

Jane said...

Astroprof, I totally agree. I think one of the reasons that this exercise worked so well is because we did spend some time on background/related info in previous classes, and (miraculously) the students seem to be actually doing the reading I'm assigning. But I have also been fairly successful in using activities to introduce new concepts--it's a bit trickier identifying the right kind of activity for that scenario, and sometimes it just falls flat, but it's fun to experiment!

One other interesting thing I've found is that my students get *really* uncomfortable with the idea of doing anything in class other than lecture, so part of my job is to "sell" the activity to the students and take care of their "performance anxiety". I swear they break out in hives when they see the phrase "discussion day" on the syllabus!

ScienceWoman and plam, thanks for your kind words! plam, good luck with your first course--sounds like a fun one!