A fair portion of schools have taken the effort to mandate/encourage course evaluations, most often submitted by students at the end of a given term. I’m aware that these reviews are used to help professors brush up on the places where they didn’t do so well, and also to given administrators some means of weeding out particularly awful lecturers.
Worthy goals, but what other nifty information could we derive from student reviews? I’d like to develop a statistical model for predicting course enrollment based on student reviews. It’s “common sense” in academia that courses with effective teachers or light workloads tend to attract more students. So, how about we use student reviews to identify classes that are going to be swamped in following terms because they are perceived by the student body as “fun” or “easy.”
This could prevent professors from being unexpectedly overwhelmed or underwhelmed by enrollement shifts and adjust their expected number of teaching assistants or course segments accordingly.
With properly-formulated course review surveys, data analysis could also pick out whether a given course was rated terribly due to the poor performance of the professor involved, or whether the class failed due to other considerations. This could help students manage their schedules to maximize their courses taken under effective professors.
If we combined enrollment and review data with course grading data, it would also be possible to parse out the relationship between workload and grades received. For example, some classes may feature light workloads yet have impossibly rigorous grading standards. Alternately, a class with a hefty workload that earns a reputation as a “hard” class might actually hand out high grades to just about everyone.
The question is whether institutions would be willing to provide such data for outside review and analysis…
