The past, present, and future of graduate admissions in physics

While graduate admissions in physics directly affects only a small number of people on an annual basis, the number of people indirectly affected is orders of magnitude greater. Those who complete graduate degrees in physics will go on to become leaders in industry, government, and academia, with the latter educating the next generation of leaders in science and engineering. Given the possibly enormous consequences of our decisions in physics graduate admissions, care should be taken to ensure that the process is working effectively. The evidence, however, suggests it is not. Many inequities exist in the admissions process, unfairly keeping potentially great scientists from pursuing graduate studies. This dissertation seeks to understand what those inequities might be and how we might address them. First, I study the admissions process in the physics department at a Midwestern, public university using the random forest algorithm, a machine learning method, to understand what drives their process. After finding test scores and grades drive the process, I show that one of those tests, the physics GRE, does not gives applicants the outsized advantage that it is claimed to provide. Given the components that drove the admissions process contain inequities, the second half of the dissertation explores whether a rubric-based holistic admissions process might be able to address those inequities. Preliminary evidence suggests that it does. Finally, to ensure that the methods used in the previous chapters were appropriate, the dissertation concludes with a simulation study, finding that the methods used might lead to false negatives in conclusions. Overall, this dissertation suggests that the current graduate admissions process in physics contains inequities and that rubric-based admissions might be able to address them. By addressing those inequities, everyone can be given a fair shot in the admissions process and physics as a discipline can work toward becoming more representative of the population. Failure to act only perpetuates the inequities that have and will continue to keep people out of physics.