In lieu of the makeup session for the class we missed---which, it turns out, conflicts with every schedule under the sun---please complete the following short assignment. This assignment will be graded on a pass-fail basis, based on effort (moderate effort, not extreme effort---I don't want this to take more than 3 hours), and will be a part of the participation grade of the course. This will be due on the last day of class, the same time as the last problem set---and you should include it in your last PSET answers.
-
Read Friederike Mengel, Jan Sauermann and Ulf Zölitz, "Gender Bias in Teaching Evaluations," Journal of the European Economic Association 2019. You need not read it super-closely, and don't have to pay attention to the mathematical details. (If you're curious about this issue, you should also read Anne Boring, Kellie Ottoboni, and Philip B. Stark, "Student evaluations of teaching (mostly) do not measure teaching effectiveness," ScienceOpen 2016, but that isn't part of the assignment---it's just bonus!)
-
With what you learned from that article in mind, consider a hypothetical disparate impact discrimination lawsuit brought by a woman denied tenure at a university on the on the basis of a process which includes an examination of student teaching evaluations. Write a paragraph or two describing what kind of data the plaintiff might like to collect and how they might like to analyze it in order to make our a prima facie case of disparate impact discrimination. Bear in mind that student evaluations aren't the only thing that goes into these kinds of tenure processes.
-
Write a paragraph or two about what a cautious university counsel ought to do (preferably before any such litigation is brought) in order to validate the use of these evaluations. Assume the university wishes to use criterion-related validity evidence, as described in Stewart v. City of St. Louis. What data should the university collect, and how should it be analyzed? (Hint: the articles noted above have some ideas about what kinds of things would be evidence of teaching effectiveness.)