This IU study of 70,000+ undergrads found that taking more in-person credit hours was associated with a *lower* likelihood of testing positive for COVID (with robust mitigation testing). This is good news, but it raises questions about why. (thread) 1/ https://news.iu.edu/stories/2020/11/iu/releases/20-analysis-supports-safety-of-in-person-courses.html

To understand these results, it's important to know that IU classes with 50+ students were online. Classes with less than 50 students defaulted to in-person. These "small" classes include labs, arts classes, language classes, seminars, some upper-level classes. 2/

One possibility is that there's selection into in-person classes based on factors related to how seriously students take COVID. Pre-med students, for example, take lots of lab classes and likely trust the science around COVID and take more steps to stay safe (e.g. party less). 3/

Another possibility is that taking classes in-person leads students to take more steps to stay safe. If students see their classmates and professors wearing masks and social distancing, they might be more inclined to keep up those practices outside of class. 4/

Now, these quantitative findings could be taken as evidence that universities should increase in-person class offerings. But that only makes sense if taking in-person classes actually causes students to take more steps to keep themselves and others safe. 5/

We don't have time to run an RCT and find out whether in-person classes actually reduce infection rates among students. But we could do qualitative research - talk to students about why they decided to take in-person classes and how taking them affected their behavior. 6/

That's why qualitative research is critical, especially right now. It can reveal what motivates people's behavior, explaining the patterns in quantitative data. In doing so, it can also make clear what policies are likely to be effective and which might do more harm than good. //