A step-by-step guide for causal mediation analysis with the G-computation (Tutorial Repository)

Cite as:

Note: This repository was created in September 2022 and is periodically updated to ensure compatibility with the latest versions of R packages.

What’s in this repository

This repository contains guidance for performing causal mediation analysis (CMA) with the mediational g-formula (g-computation) in R. The workflow follows the approach described in: Burgos-Ochoa (2023) https://doi.org/10.1111/ppe.12954

The files are:

CMA_gformula_tutorial.Rmd An R Markdown notebook with step-by-step guidance, explanations, and code. Use this if you want a structured tutorial that explains each step (natural course, counterfactuals, mediation scenario, TE/NDE/NIE, bootstrap).

G-formula_BigData_Special_Issue.R A plain R script with the same code, but without explanatory text. Use this if you only want the implementation.

Who it’s for

Researchers and practitioners with experience with data analsis in R and familiarity with regression models, who want to apply CMA using the g-computation approach to their own data.

How to use

• Clone or download this repository.
• Open CMA_gformula_tutorial.Rmd in RStudio (or another R Markdown-capable editor).
• Run through the chunks step by step to reproduce the example and adapt it to your dataset.
• If you prefer a leaner version, use the R script G-formula_BigData_Special_Issue.R.