Event

13 Feb 2026 09:00am to 05:00pm

Causal Inference with Hidden Confounding

Workshop
Event Location
The University of Melbourne, “The Spot” building (Room 1022), 198 Berkeley St, Carlton
Australia
Speakers
Prof Eric J. Tchetgen Tchetgen
The University of Pennsylvania
My primary area of interest is in semi-parametric efficiency theory with application to causal inference, missing data problems, statistical genetics and mixed model theory. In general, I work on the development of statistical and epidemiologic methods that make efficient use of the information in data...

Quantitative research in health and social sciences often aims to determine the causal effect of a well-defined intervention. While the placebo controlled randomised trial design represents the gold standard for causal inference, such studies are often infeasible for practical or ethical reason. Even if feasible, the validity of a randomized trial might be compromised by noncompliance, treatment switching, dropout, or other possible protocol violations.  As a result, researchers routinely turn to observational studies where they must contend with the ubiquitous challenge of confounding bias, whereby participants in the treatment arm may not be exchangeable with those in the control arm at baseline with respect to both observed and unobserved risk factors for the outcome. 

This workshop will focus on modern methods for causal identification and inference from observational data and imperfect randomized trials subject to confounding by hidden factors. After a brief review of methods to account for measured confounders, including standard g-methods, the workshop will largely focus on methods to account for confounding by hidden factors, such as difference-in-differences, instrumental variable, changes-in-changes, synthetic controls, negative controls, and modern proxy-based approaches. While the workshop focus is primarily on conceptual aspects of causal inference, numerous applications in health and social sciences will be discussed to illustrate the concepts and methods.

Pre-requisites: Basic understanding of regression analysis, statistical inference and probability theory. No laptop or programs will be required to attend this course.

Target audience: Quantitative researchers in health and social sciences, as well as clinicians and public health professionals.  Early career researchers in statistics, biostatistics, economics, political science, sociology and epidemiology seeking a rigorous first course on causal inference with observational data.

Date: Friday 13th February

Registration prices: Standard $540, Student $390

 

Registration is now open via Trybooking