If you’ve been reading this site for a while, you understand that I don’t have all the answers. I often write about things as I learn them. That’s especially true today as I try to implement group projects in one of my classes. This is something I’ve had mixed success with in the distant past, but for a variety of reasons, I’m giving them another try.

When I look back at the big econometrics lecture course I taught last Fall, I see a lot of good stuff with one gaping hole. I posed many interesting substantive questions, and my students learned the mathematical and computational tools they needed to answer them. But they didn’t get a chance to ask their own questions. In my seminars, the creative project is by far the most important thing students do during the semester. They ask questions they care about and work extremely hard to answer them. I want my econometrics students to have the same experience.

Last month I piloted the “creative empirical project” in my online econometrics class. Five weeks is not a lot of time to go deep, but it was mostly successful. One student estimated the influence of parental smoking behavior on child smoking, while another figured out how the performance of NBA players is rewarded in their contracts. They certainly worked hard.

Unfortunately, this isn’t going to scale to the 150-200 students I’ll have in the fall: Even with an army of teaching assistants, we won’t be able to provide the guidance and feedback required to every student. This is the main reason I’m bringing back the group project: 50 4-person projects (or 40 5-person projects) should be feasible for my team to manage.

As folks who do this regularly can attest, group projects have other advantages as well. Students learn how to work together and that’s pretty useful considering how many jobs in the real world require serious amounts of collaboration. And the projects themselves can be much more ambitious (and satisfying) when you have multiple contributing group members.

All that said, group projects are risky as everyone has heard about (or experienced first hand) groups where one or two students do all the work and everyone else free rides. There are ways to reduce the likelihood of this happening, and I’m implementing a few this month as my July online students work in groups on their creative projects.

Decision 1: How to form the groups?

For my online class, the most important dimension I want teams to match on is physical location/time zone. Several of my 17 students are in New Haven, and it’s pretty convenient if they’re in the same groups. Meetings are far easier to coordinate for remote students if they live in the same time zones.

In the fall, I’ll want to match based on schedule availability so they are physically able to meet. I also want the group members to be in the same discussion sections so they can talk about projects before, after, and sometimes during section. The teaching assistant for each section will be responsible for mentoring the groups in their section.

On top of this, it would be nice to have groups cluster on subject area interests and have a balance of skill levels. Ever since I heard about the CATME tools on Episode 25 of the Teaching in Higher Ed Podcast, I’ve wanted to try them. They automate the requisite surveying and group formation, and provide lots of other helpful information for managing groups. There’s a bit of a learning curve and I ran out of time this summer, but I’m going to try again in the fall when it will be a much bigger job.

Decision 2: How to incorporate peer evaluation?

I’m convinced that the optimal way to assign individual grades for project work is to combine measures of the quality of the overall project and the quality of the individual’s contribution. Putting weight on the whole project encourages cooperation while putting weight on the individual contribution prevents shirking.

With a good grading rubric, the project quality is relatively straight-forward to evaluate. The individual’s contribution is nearly impossible for the instructor to observe, but often quite clear to the group members themselves. That is where the peer evaluation comes in.

If I were less of a control freak, I might let my teams decide for themselves how they want to evaluate each other. Instead, I’m keeping it very simple my first time around. After each deliverable, they will grade each other on:

• Effort: Did this team member contribute to the best of their abilities?
• Dependability: Did this team member do what they say they would do? This includes showing up for meetings prepared, and producing code or prose.
• Collegiality: Was this team member a good listener? Did they respond well to feedback and/or criticism?

It might turn out these are terrible metrics, and I’m certainly open to changing them. But they make sense to me and I think they’re a reasonable starting point. Each team member will rate every one else on their team on a scale of one to five and I’ll compute the percentage of the maximum number of points that each person gets. This “peer score” will get averaged with the score I assign to the project to get an overall score. I’ll collect the data with a Google form and massage it into what I want with Excel or Stata. In addition to their Team Builder, CATME offers a more serious (and research-based) tool for peer evaluation, and I might try this in the Fall.

If you’re planning to try peer evaluation I suggest reading a few articles about to get different opinions on these topics. For example, folks at the University of Colorado have thought long and hard about peer evaluation and they’ve written a great article (Assessing Student Group Work using Peer Evaluation) that contains links to many other useful resources.

Decision 3: Should I have group charters?

Some of my friends with far more experience than me (Hello Bonni Stachoviak!) swear by having groups write their own “charters.” Creating these documents forces groups to think through up front what their individual goals are, how they are going to work together, and how they will to resolve conflicts. I love this idea, but even if I hand out a good template, I have no idea what I would say to students if they wanted guidance. So for now, I’m keeping things simple and instead just encouraging groups informally to think about these issues before they starting working together in earnest.

There are very few things in life that we do well the first time, but that doesn’t mean we should never try anything new. I have no doubt that this month my group project will be sub-optimal. I’ll learn from my mistakes and try again in the fall and then again next summer and hopefully by next fall, I’ll have a process that works well.