Identification of At-risk Students

I investigate how to predict students' learning outcomes using clicker responses. This is because clicker data are easy to collect especially in Peer Instruction classroom setting. The performance of the initial model is provided on my ICER 2016 paper.
Currently, I am improving my model and applying it to different courses.

Scalability of Peer Instruction

How effective Peer Instruction is in a large classroom? Is it just as good in a small classroom? This is an interesting research question because the enrollment and the class size in Computer Science has increased. My SIGCSE 2017 paper analyzed student course evaluation data of Computer Science and Engineering department at UC San Diego. The results showed Peer Instruction scales better than traditional pedagogy.
As a next step, I am currently investigating whether this hypothesis holds in other disciplines.