EPM5018 · Causal inference

Overview

This unit covers modern statistical methods for assessing the causal effect of a treatment or exposure from randomised or observational studies. The unit begins by explaining the fundamental concept of counterfactual or potential outcomes and introduces causal diagrams (or directed acyclic graphs (DAGs)) to visually identify confounding, selection and other biases that prevent unbiased estimation of causal effects. Key issues in defining causal effects that are able to be estimated in a range of contexts are presented using the concept of the “target trial” to clarify exactly what the analysis seeks to estimate. A range of statistical methods for analysing data to produce estimates of causal effects are then introduced. Propensity score and related methods for estimating the causal effect of a single time point exposure are presented, together with extensions to longitudinal data with multiple exposure measurements, and methods to assess whether the effect of an exposure on an outcome is mediated by one or more intermediate variables. Comparisons will be made throughout with “conventional” statistical methods. Emphasis will be placed on interpretation of results and understanding the assumptions required to allow causal conclusions. Stata and R software will be used to apply the methods to real study datasets.

Offerings

Assessment

The Handbook does not list a final examination among the assessment items. That is not a guarantee there is none.

Assessment details may change. Please refer to the assessment information in Moodle closer to the start of the teaching period.

Requisites

Learning outcomes

1. Use counterfactuals (potential outcomes) to precisely define causal effects
2. Describe the differences between association and causation, and the fundamental assumptions required for causation
3. Construct causal diagrams and use them to identify potential sources of bias
4. Implement causal inference methods, using software, for single time point and longitudinal exposures, and for mediation analyses
5. Interpret results of analyses in light of the causal assumptions required
6. Effectively communicate in language suitable for the scientific community the results of analyses using causal inference methods

Workload

No workload detail published.

Student experience, not official rules. Nothing here changes what the Handbook says.