BMDS237

What can we do when randomization isn't possible? How can we estimate the effect of a policy change, treatment, or intervention when we can't run an experiment? This project-based course teaches you how. You'll learn to answer causal questions using observational data - messy, real-world datasets that capture actual clinical course, policy implementations, and health outcomes. We'll work with large medical databases, survey data, and administrative records to tackle questions that randomized trials can't or won't address. What we'll do: Build the foundations of modern causal inference frameworks;Critically reproduce and learn from influential published studies that use real-world data; Complete a hands-on research project; Learn both the statistical methods and the art of designing credible quasi-experimental studies. This course is ideal for students planning research careers who want practical skills in causal inference. Students with interests in health policy, epidemiology, health economics, or data science are encouraged to enroll. Prerequisites: one or more courses in probability, and statistics or biostatistics.