What question did this study set out to answer?

This research aims to evaluate how well targeted maximum likelihood estimation (TMLE) performs for estimating treatment effects when data is missing and when positivity assumptions might be violated.

April 27, 2026Biometrical JournalOpen Access

Causal Effect Estimation With TMLE: Handling Missing Data and Near Violations of Positivity

View Full Paper

Ask AI

Bookmark

Authors

CWChristoph WiederkehrLudwig-Maximilians-Universität München CHChristian Heumann MSMichael Schomaker

Discussion

Member takes

Implication

Randomized trial evaluates TMLE's performance on treatment effect estimation in missing data contexts, indicating effective bias reduction strategies.

Key Points

This research aims to evaluate how well targeted maximum likelihood estimation (TMLE) performs for estimating treatment effects when data is missing and when positivity assumptions might be violated.
Utilized model- and design-based simulations to assess TMLE under different missing data scenarios.
Applied undersmoothed highly adaptive lasso to the WASH Benefits Bangladesh data set.
Compared eight missing data methods combined with TMLE, analyzing both single and multiple imputation techniques.
Non-MI methods, especially complete cases with TMLE using an outcome-missingness model, showed lower bias than other methods.
MI with classification and regression trees (CART) achieved lower root mean squared error while maintaining nominal coverage rates.
Trade-offs between bias and coverage were observed, with recommendations for TMLE methods.

Cite This Study

Wiederkehr et al. (2026) studied this question.

synapsesocial.com/papers/69eefde9fede9185760d4a64 https://doi.org/10.1002/bimj.70134

View Full Paper

Ask AI

Bookmark

Implication

Randomized trial evaluates TMLE's performance on treatment effect estimation in missing data contexts, indicating effective bias reduction strategies.

Causal Effect Estimation With TMLE: Handling Missing Data and Near Violations of Positivity

Discussion

Member takes

Implication

Key Points

Cite This Study

Causal Effect Estimation With TMLE: Handling Missing Data and Near Violations of Positivity

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider