Part of a UNO Economics seminar series. Research seminars run from 3PM-4PM. Appetizers at Inner Rail afterwards.

Abstract:

It is common for courses to employ some form of peer grading. For instance, internal group interactions are unobserved by the instructor. Thus, instructors often use peer grades to assess the work produced by each of the group participants. Without this peer grading, free riding might become more intense as it further disincentivizes students to fairly split the workload. However, a significant portion of the Student Evaluation of Teaching literature suggests that SET scores are biased against female faculty. Notably, this bias is exhibited by the students as they are rating the faculty members. Therefore, it is reasonable to suspect that these students might exhibit this same bias against their female classmates when evaluating their work. This work has three goals. First, we extend the rubric-based estimation technique first introduced by Smith and Wooten (2023) to support arbitrary additional variables. Second, we incorporate the arbitrary additional variables capability into a software package so that others can use this technique. Finally, with the support for arbitrary additional variables, we estimate the degree of bias students exhibit when grading their female colleagues.