I’m a computer science PhD student in the Stanford Computational Policy Lab, supervised by Prof Sharad Goel. My research focuses on applying new computational tools from machine learning and statistics to policy problems. Most recently I’ve been interested in the US criminal justice system, studying:
- Algorithmic risk assessments for pre-trial decision making, and specifically what it means for these risk assessments to be fair; [paper, slides, talk]
- Police traffic stops, where we developed a model to determine if police officers were racially discriminating in their decision to search vehicles. [1, 2]
I’m also a member of the Stanford Open Policing Project, an effort to collect, clean, and release records from 130 million police traffic stops around the country.
Finally, I work with David Rothschild and Tobi Konitzer to measure public opinion using non-representative mobile polls. Our work was featured in the New York Times, where, of 5 pollsters, we were the only one to correctly give Trump the lead in Florida.
I’m on Twitter @scorbettdavies