Epi 625 Doctoral Seminar in Causal Inference Methods
This advanced doctoral-level course builds upon students’ prior training in epidemiology, biostatistics, and applied research, to deepen their exposure to and experience with advanced analytical methods for causal inference. Working within the potential outcomes (counterfactual) causal framework, this course situates this framework within a broader context of causality, introduces advanced causal methods, and provides students an opportunity to implement them in practice. Specific methods taught include propensity scores, inverse probability weights, g-computation, and related approaches. Methods are taught with a mixture of theoretical background, in-class activities, and at-home coding assignments using simulated data provided by the instructor. Throughout, students are encouraged to apply these approaches to their planned or in-progress doctoral research. The principal goals of this seminar class are to: 1) Familiarize students with counterfactual model for causal inference and how it fits in with other causal inference frameworks from epidemiology and beyond. 2) Provide students with the theoretical underpinning of each of several methods for causal inference, including g-computation and inverse probability weighting. 3) Apply these advanced causal methods in-class via assignments and through coding assignments. 4) Encourage students to consider how these advanced epidemiologic methods may be applied in their dissertation research.
Prerequisite
Graduate training in epidemiologic methods and biostatistics.
Epi 512,
Epi 513,
Epi 514, and
BSTA 612 (linear models) or equivalent, or permission of instructor.