Lab groups and peer evaluations

This past year, I have been working to develop a series of labs that focus on scientific skills, as opposed to teaching physics content. These changes are motivated in part by the pandemic: I want to have authentic laboratory experiences that students can complete at home with limited resources. However, these reforms are also motivated by the literature which suggests that lab is better suited to the teaching of such skills as opposed to content:

  • Holmes, Natasha G., and Carl E. Wieman. “Introductory Physics Labs: We Can Do Better.” Physics Today 71, no. 1 (January 1, 2018): 38–45. https://doi.org/10.1063/PT.3.3816.
  • MacIsaac, Dan. “Report: AAPT Recommendations for the Undergraduate Physics Laboratory Curriculum.” The Physics Teacher 53, no. 4 (April 1, 2015): 253–253. https://doi.org/10.1119/1.4914580.

Lab groups are one of the necessities of such a large class. In order to respect the TA’s time and keep the grading load manageable, students must turn in reports as groups. Fortunately, I also think that learning to work in a scientific team is also an important goal of the lab experience.

This past semester, I have been trying to use Moodle to manage the lab groups and CATME to do peer evaluations. However, this has yielded two problems:

  1. The TAs must keep the lists in Moodle up to date and there is an unclear chain of command with regards to group management. Also, this requires a rather sophisticated understanding of Moodle and makes changing/managing groups difficult.
  2. The CATME protocol, while fantastic, is, I think, insufficiently transparent. Moreover, I must manage it. This is, frankly, too much load for me. I need a system that the TAs can successfully manage on their own.

I really like the multiplicative nature of the CATME results. A plan with which I am currently toying involves:

  • Have a number of points equal to the number of members in the team.
  • Each team member would distribute these points to their team members. Perhaps this would be done for a few different categories.
  • There would also be one optional point that could be given to someone who really deserves an extra boost. This would be a bonus: if everyone in the team neglects to do it, they will still all get ones (i.e. their score would be equal to their actual grade).
  • The result would be scaled in such a way that the final multipliers are between 0.7 or so and 1.05.

Obviously, this needs to be flushed out, but there are some key points for improvement here.