A course in emulating trials you cannot run

What this is

A free, open course on causal inference and target trial emulation, written for people who analyze observational health data: forty-five chapters in three levels, a DAG builder, an R console, and eleven calculators for the bias and design questions that come up in review.

Who it is for

Epidemiologists, biostatisticians, pharmacoepidemiologists and clinical researchers who have data and a causal question. It assumes you can read a regression model and are comfortable in R. It does not assume you have seen potential outcomes, g-methods or quantitative bias analysis before.

How it is organized

Chapters 1–15

Counterfactuals, causal graphs, confounding, selection bias and measurement error.

Chapters 16–33

Propensity scores, inverse probability weighting, standardization, the g-formula, marginal structural models and survival analysis.

Chapters 34–45

Time-varying treatment, mediation, instrumental variables, and quantitative bias analysis including probabilistic and Bayesian approaches.

The study database

The course teaches the method. Whether the method has worked in practice is a separate, empirical question, and it is answered by a living systematic review maintained at Xera OpenScience: every published target trial emulation that reports a randomized benchmark, paired with the trial it emulates.

This site reads its headline figures from that database rather than keeping a copy, so the two cannot disagree.

Accounts

Every chapter, tool and calculator is readable without signing in. An account exists only to record which chapters you have finished, keep your saved DAGs, and track quiz results.

Part of Xera.ac

Xera.ac builds open tools for research transparency. Alongside this course: the target trial emulation database and the Retraction Tracker.