Skip to main content

340.603.49
Applied Epidemiologic Analyses for Causal Inference

Location
Internet
Term
Summer Institute
Department
Epidemiology
Credit(s)
2
Academic Year
2025 - 2026
Instruction Method
Synchronous Online
Start Date
Monday, June 10, 2024
End Date
Friday, June 14, 2024
Class Time(s)
M, Tu, W, Th, F, 8:30 - 11:50am
Auditors Allowed
No
Available to Undergraduate
No
Grading Restriction
Letter Grade or Pass/Fail
Course Instructor(s)
Contact Name
Frequency Schedule
Every Year
Prerequisite

At least one intermediate epidemiology course with a working knowledge of regression and other standard statistical methods in basic epidemiologic analysis (e.g., calculating risk differences, risk ratios, and odds ratios from 2x2 tables; knowing basic concepts of survival analyses like time origin, and how to read a survival curve; regression modeling for continuous and dichotomous response variables). Familiarity with R, SAS, or Stata is strongly advised although not enforced.

Description
This course introduces students to key terms and concepts widely used for principled causal effect estimation in epidemiology and biostatistics, and walks students through a hands-on introduction to estimation of causal effects using generalized (“g-“)methods (inverse probability weighting and g-computation).
Introduces concepts and applications of potential outcomes and structural causal models for the estimation of causal parameters in epidemiologic research. Familiarizes students with the assumptions underpinning modern causal inference methods and provides a conceptual understanding of standardization/g-computation and inverse probability weighting. Applies each of these methods in estimating the effect of a time-fixed exposure in a simple setting. Discusses the application of these methods in the literature.
Learning Objectives
Upon successfully completing this course, students will be able to:
  1. Identify the definition, interpretation, and limitations of the concept of a “causal effect” under the potential outcomes and structural causal models frameworks.
  2. Identify study biases such as confounding and selection and how to represent them and identify them in causal graphs.
  3. Identify epidemiologic approaches and study designs that seek to overcome these biases and the assumptions on which they rest.
  4. Apply the “target trial” framework for specifying a study design.
  5. Become familiar with inverse probability weighting and g-computation/g-formula for answering a causal question and understand the strengths and limitations of each approach.
  6. Discuss applications of modern causal inference framework and methods to observational data.
Methods of Assessment
This course is evaluated as follows:
  • 15% Participation
  • 15% LiveTalks
  • 70% Assignments
Special Comments

This is the online/virtual section of a course also held onsite. You are responsible for the modality in which you register.