This 4-day workshop (May 27-30, 2025) covers foundational concepts, theory, and methods for causal inference from observational data and imperfect randomized experiments. Attend in person at the University of Pennsylvania or online, with full or single-day options, and gain skills to critically engage with modern causal inference methods.
This 4-day workshop introduces foundational concepts, theory and methods for causal inference from observational data, and imperfect randomized experiments. Participants will acquire the requisite set of skills to become sophisticated and critical consumers of emerging modern causal inference methodology. Topics covered include the g-formula, inverse probability weighting of marginal structural models, causal mediation analysis, trial emulation and methods to handle unmeasured confounding via instrumental variables, difference-in-differences techniques and proxy-based methods. The workshop will include several areas of applications in the population health sciences with special emphasis given to policy evaluation and comparison effectiveness in pharmacoepidemiology.
| Date | AM Session | PM Session |
|---|---|---|
| Tuesday, 5/27/2025 | Welcome and Introduction Historical Perspective on Causal Inference Introduction to Identification of Causal Effects | Estimation of Causal Effects Quasi-Experimental Designs |
| Wednesday, 5/28/2025 | Mediation Analysis Instrumental Variables | Sensitivity Analysis Causal Discovery |
| Thursday, 5/29/2025 | Bayesian Causal Inference | Trial Emulation Practicum |
| Friday, 5/30/2025 | Real World Evidence and Research Design Avoiding Collider and Time-Related Biases in Pharmacoepidemiology | Proxies and Negative Controls |