AI, Ethics, and Geoethics (CS 5970)
Module 6: Case study on racial bias in university algorithms
- (10-15 min) Read the article
- (5 min) Read the case study
- (15-30 min) Discussion on slack
- (5 min) grading declaration
Read this recent article (also showed up on in-the-news): Major Universities are Using Race as a “High Impact Predictor” of Student Success. Warning, the quote at the end is rather horrifying (it comes from a university president who was fired for saying it. I’m rather shocked the article repeats it!). The rest of the article is just flat out disturbing.
Black students labeled less likely to succeed
This graph comes directly from the article for today
Imagine that OU is considering adopting this software and has put together a focus group of students, faculty, and staff to give input on whether they should adopt this or not. You are in one of the focus groups, giving feedback to the university. The main argument for adopting the software is that we are short-staffed and they want to improve student success in a way that does not break the bank. Note the discussion of advisor-to-student ratios discussed in the article. I don’t know our numbers but they are at least one advisor per several hundred students.
Please note, I have no idea if we are already using it as we were not mentioned by name in the article. However, it seems like something we could do! Also, note the discussion at the end about using this same model for admissions.
This discussion will happen in the #case-studies channel. Remember to use threads so that we can keep track of the conversation more easily.
Answer the following questions.
- If your perception is that your focus group is actually providing feedback that the university is listening to, what feedback do you provide about the use of this software? Do you see both pros and cons? Be sure to give your feedback to an audience that is not computer scientists. As with our earlier case studies, your target audience matters here.
- If you feel that your focus group was assembled to provide justification to say yes to adopting the software (yes, this happens here as well as at other places), how do you handle your opinion of the use of the software?
- What if you learn that they are considering using it for admissions as well as for risk prediction? What feedback would you give in this scenario?
- OU students: After you have done your reading and engaged actively in discussion, complete the grading declaration titled “Module 6: Case study on racial bias in university algorithms”