Estimation guide#
All estimators are presented in the publication associated with this repository. .. admonition:: citation
Judith Abécassis, Houssam Zenati, Sami Boumaïza, Julie Josse, Bertrand Thirion. Causal mediation analysis with one or multiple mediators: a comparative study. 2025. hal-05060162
More details about the estimators, and how and when to use them, will come!
References#
Coefficient Product
The coefficient product method was proposed in
Valeri, L. and VanderWeele, T. J. `Mediation analysis allowing for exposure–mediator interactions and causal interpretation: Theoretical assumptions and implementation with SAS and SPSS macros`_.Psychological Methods, 18(2), 137–150. 2013.
Valeri, L. and VanderWeele, T. J. `SAS Macro for Causal Mediation Analysis with Survival Data`_. Epidemiology 26(2), 2015.
IPW estimator
The Importance Weighting (IPW) method originally was introduced in
and then presented in mediation analysis in
Huber, Martin, `Identifying causal mechanisms (primarily) based on inverse probability weighting`_. Journal of Applied Econometrics, 29, issue 6, p. 920-943, 2014.
G-computation estimator
The G-computation method for the mediation was introduced in
Robust estimators
The Double machine learning method originally was introduced in
Chernozhukov, D. Chetverikov, M. Demirer, E. Duflo, C. Hansen, W. Newey and J. Robins, Double/debiased machine learning for treatment and structural parameters, The Econometrics Journal, Volume 21, Issue 1, 2018.
and then adapted to the mediation task in
Farbmacher, M. Huber, L. Lafférs, H. Langen and M. Spindler, `Causal mediation analysis with double machine learning`_, The Econometrics Journal, Volume 25, Issue 2, Pages 277–300, 2022.
The Multiply Robust (MR) estimator was introduced in
Tchetgen EJ, Shpitser I. `Semiparametric Theory for Causal Mediation Analysis: efficiency bounds, multiple robustness, and sensitivity analysis`_. Ann Stat. 40(3):1816-1845. 2012.