teaching

Health Services Research Methods I

Fall 2022-2025

This course is the first in a two-semester sequence on health services research methods, with a focus on econometric methods. After a review of probability theory, potential outcomes, and linear regression, we formalize the selection problem and discuss approaches to adjust for observables, first with linear regression and then with matching and machine learning. In the second half of the class, we transition to experimental and quasi-experimental methods: randomized controlled trials, instrumental variables, regression discontinuity, and difference-in-differences. Throughout, students learn to design, analyze, and critically evaluate causal studies, with attention to modern methods for inference, diagnostics, placebo tests and robustness tests/sensitivity analyses.

Napkin Math: Model intuition for the digital age

A short course from 1 hour to 3 days

Napkin math is an approach to analytic thinking that uses simple models of the world to provide a first-pass answer to policy questions and to inform the development and interpretation of more complex research. This course shows students how, when faced with a policy question, they can map it onto a mental model of an underlying data-generating process, break it into component pieces, generate a rough starting estimate, and know where to look next. We work through examples of how these skills help systematically catch typos, thinkos, and wrong turns throughout the research process. Napkin math helps you be an active and engaged producer and consumer of research: to understand mechanisms rather than fixating on results, and to avoid taking refuge in complexity. The course is organized around how these skills apply across stages of the research lifecycle: (1) choosing questions and designing studies, (2) coding, (3) gut checks and mid-course corrections, and (4) interpreting and comparing studies.