Broadly speaking, my research has three themes: 1) Causal inference for studying "large", unprecedented policy changes like the introduction of the Affordable Care Act or the creation (or break-up) of a major free-trade area (such as NAFTA, Mercosur, or the European common market, 2) Statistical computing, including issues related to porting older custom statistical programs into Python, and comparing both efficiency and usability of the implementations of statistical techniques across different programming languages, and 3) Bringing geospatial analysis, quantile regression, and multilevel/hierarchical modeling to areas of the social, behavioral, and medical/health sciences where these techniques are less-commonly used.
My recent academic work, especially since I became an MPH student in the University of Chicago Biological Sciences Division in 2023, focuses on improving how we do research and evaluation relating to the US health care system. This includes applications of hierarchical models to healthcare settings, GIS and geospatial methods for studying health access and equity, analysis of patient-provider interactions, statistical computing for biomedical applications, and the evaluation of machine learning methods for both clinical and policy uses in medicine.
As someone who entered academia in the aftermath of the Replication Crisis, I view reproducible research not just as a core value or a set of best practices, but as the beginning of a very important conversation about how researchers actually do their work. I see this as something we should not only study and interrogate, but something we should think more carefully about how we might best communicate, explicate, and share with stakeholders, the general public, and our younger peers, especially those without personal or family experience in academia. Cross-disciplinary exchange of ideas, direct collaboration between methodologists and subject-matter experts, and placing greater value on the perspectives and experiences of the people and communities we study are all approaches which have not only risen in priority in recent years, but have played major roles in my own research career.
I have also worked on both pure and applied research projects with the founders of a medical technology startup (whose backgrounds include MD psychiatrists, clinical psychologists, and experienced professionals in both hardware and software engineering) working to make apps and tools that we hope will vastly reduce the cost and increase the speed of diagnosing movement disorders like Tardive Dyskinesia and Parkinson’s.
Publications, Invited Presentations, Grants:
Provider Level Length-of-Stay: Multivariate Model vs. Standard Metrics (joint with Matthew T. Cerasale and Andrew W. Schram). Presented at the Society of Hospital Medicine national meeting March 2023 (SHM Converge, Austin, TX), final manuscript in preparation.
The Impact of Providers on Duration of Hospitalization: Multilevel Regression Modeling (joint with Matthew T. Cerasale) Poster presented at the Institute for Public Health & Medicine Population Health Forum April 2024 (Northwestern-Feinberg, Chicago, IL)
Leveraging High Stress Months to Better Identify Inpatient Provider Performance (joint with Matthew T. Cerasale) Poster presented at the Institute for Public Health & Medicine Population Health Forum April 2025 (Northwestern-Feinberg, Chicago, IL)
Tardive Dyskinesia Detection Device: A Machine-Learning System for Movement Disorder Diagnosis (joint with W. Scott Tobey, Richard A. Markin, Bill Paulson, and Leonard Carr). NIH grant application, in-process.
Mobile AI-powered imaging technology to screen biomarkers for patients with movement disorders (joint with Leonard Carr, W. Scott Tobey, and Rizwan Akhtar). NIH, under scientific review.
Methods for Evaluating Inpatient Provider Efficiency: Comparing Standard, Multilevel, and Machine Learning Approaches (Master’s Thesis and MPH Capstone Project, accepted by the Department of Public Health Sciences May 2025, Faculty Advisor: Don Hedeker)
Provider‑Level Heterogeneity in Inpatient Efficiency Under Resource Strain: A Multilevel Quasi‑Experimental Analysis (joint with Matthew T. Cerasale) Under preparation, abstract and poster presented at the 2026 American Causal Inference Conference (Salt Lake City, UT May 2026)
Comparing Standard, Multilevel, and Machine Learning Approaches for Evaluating Inpatient Provider Efficiency(joint with Matthew T. Cerasale) Abstract and poster accepted for presentation at the AcademyHealth 2026 Annual Research Meeting (Seattle, WA May-June 2026)