Find the subject in the handbook here Semester 1 - online
Subject Description
This unit introduces modern statistical methods for estimating causal effects from randomized or observational studies. It covers counterfactual outcomes, causal diagrams, and directed acyclic graphs (DAGs) to identify biases like confounding and selection. The “target trial” framework is used to define estimable causal effects. Methods such as propensity scores are taught for single time-point exposures, with extensions to longitudinal data and mediation analysis. The unit emphasizes comparing these methods with conventional approaches and interpreting results, highlighting necessary assumptions for causal inference. Practical applications are conducted using Stata and R software on real study datasets.
Assignments
Comments