PBHLT 7115 - Causal Methods in Public Health

This course provides in-depth training in causal methods focused on causality claims in public health and the current study design and analytical methods that strengthen such claims. This includes the counterfactual/potential outcomes framework, causal estimands, mathematical conditions under which associational measures of effect equal causal measures, paths of confounding and selection bias, and time-dependent confounding. It includes experimental versus observational study designs and natural experiments. Analytical methods include propensity score models, inverse probability of treatment weighting, g-computation, g-estimation and associated techniques from econometrics. Selected advanced topics will be covered, such as extensions of analytic techniques to effect modification, mediation analysis, targeted learning, clustered data, and dynamic treatment regimes.