In keeping with my recent posts about final exams, I'm curious about whether anyone has tried anything like a collaborative exam. I've been interested in the idea for a few years now -- somehow letting students work together on the exam but grading them individually. Another colleague of mine is also interested, and we recently batted around some ideas.
What interests me most about the idea is that it's probably most like practice in a lot of ways. In the practice world, when you have a problem, you are free to seek assistance even if you have to be solely responsible for the output. You can talk with an associate in your firm, an old law school friend, the assigning partner, experts in the field, etc. Talking with others about what you're working on is a normal part of law practice.
It's very appealing to me to try to create a final exam that can work in a similar way.
But I'm having trouble coming up with workable models for this. One idea is something one of my science teachers in high school did. We worked in lab groups and were able to talk with our lab partners about our problem/report. But, we had to write the report ourselves.
Can a similar model work for a final exam? Can I tell my students that they're allowed to talk with their classmates about the problem but that their writing has to be their own? I vastly prefer take-home exams for a variety of reasons, but is this model too rife with opportunity for cheating in a take-home format? Maybe it would work better in an in-class format, with students allowed to talk with one another for the first hour of the exam but then the rest of it has to be done by themselves. The proctor would be able to enforce these rules in a way that couldn't be done for a take-home exam.
But I'm not even sure this model is a good one. How would students pick others to work with? What about the student who doesn't know anyone well? Do I assign groups? If so, how? Do the students go off into their groups in separate rooms and then come back and write in the exam room? What about the good student who doesn't want to share ideas? Or might share wrong ideas to trick others?
Basically, I like the idea but am at a loss about implementation, so I'm looking for thoughts from others out there on this. All responses -- about the wisdom of the idea, about experiences with it, about implementation -- welcome!
Your model fails to mimic practice in a critical respect. In firm practice environments, people do go to one another for help spotting issues, finding resources, etc. But there are mechanisms for monitoring, over the longer term, who is contributing and who is free-riding. To be sure, these mechanisms are imperfect. But your exam scenario is a one-round game in which free riders are rewarded equally to contributors. Lousy way to grade.
Posted by: Bernie Burk | December 15, 2010 at 02:08 PM
I have to agree that this would be a nightmare from the students' perspective. It reduces the class to a popularity/connections contest. It might be true that practice is like that, but most employers do not think law school classes measure those skills/attributes. Another risk you face as a professor is inconsistent grading. There was a story of this happening where I went to school where a group of 4 students turned in the same exam and got grades ranging from A- to C+ because the professor graded inconsistently.
Posted by: anon | December 15, 2010 at 03:57 PM
We've been thinking about this for our new Global Professional LLM program, which is an executive-style masters degree in business law, open to lawyers and a small number of non-lawyers with an interest in law. We would love to incorporate team assignments as in Executive MBA programs because this is such an integral part of providing legal services in the new global business environment. It is challenging to make team assignments part of individual evaluations/assessment, so our faculty are thinking about starting with some group work (presentations, problems or cases, etc.) in the classroom.
Posted by: Archana | December 15, 2010 at 04:01 PM
I agree that this is a bad idea.
In addition to the prior comments, I think the analogy fails. It is true that "talking with others about what you're working on is a normal part of law practice." But it is also a normal part of *preparing* for law school exams. A student meets in a study group, consults with her professor, looks at other outlines, and uses whatever collaborative techniques she favors. Preparing for exams is already a collaborative exercise, just like legal practice. And when you walk into court and a judge asks you questions, you answer yourself in real-time, just like in individual exams.
Posted by: Orin Kerr | December 15, 2010 at 05:26 PM
I have assigned collaborative projects for students in small seminars. I don't make it a very large portion of the overall grade, but they do receive a 'group grade'. That said, I have them write individual narratives recounting their group experience and I reserve the right to adjust individual grades based on my cumulative reading of the narratives. In class I stress the importance of learning to work in groups and how this has 'real world' applicability. We also spend some time discussing (as a group) how to handle free rider problems, working in groups, and personal accountability. Thus far, students have seemed receptive to the idea.
Posted by: Jeff Yates | December 16, 2010 at 09:40 AM
To follow up with a response I got via phone (yes, people still use that!): the prof who called me said that he regularly gives collaborative exams in his conflicts of law course. He said that he asks students to work in groups of 3 as a 3-judge panel writing an opinion. They can write one opinion for the group or individuals can write concurrences/dissents (while joining part of the group opinion). He said that his students have really liked this type of exam and have not complained about any of the fairness issues that I raised in my post. He said he also likes it for a variety of reasons too.
Posted by: David S. Cohen | December 16, 2010 at 11:44 AM