Cross-Fitted Survey-Weighted TMLE
  • Home
  • Simulation
  • NHANES
  • Robustness
  • Methods
  • About

On this page

  • Citation
  • Code and data
  • How this site is built
  • Contact
  • License

About

Citation

This site is the interactive results companion to:

M. Ehsan Karim. Cross-Fitted Survey-Weighted TMLE with Design-Based Variance for Causal Machine Learning. Manuscript, 2026.

Code and data

  • The estimator, simulation engine, and NHANES analysis code live in the public code repository, https://github.com/ehsanx/Survey-TMLE-codes.
  • The numbers shown here are pre-computed and frozen: every plot and table is rebuilt from the CSVs in data/, which are snapshots of the manuscript’s result files (multiple imputation, \(m = 40\)).
  • The manuscript and Web Appendix are the source of record. This site is a convenience layer; if any number here ever differs from the paper, the paper governs.
  • Data source: the U.S. National Health and Nutrition Examination Survey (NHANES), publicly available from the National Center for Health Statistics.

How this site is built

A static Quarto website rendered to docs/ and served from GitHub Pages — no server, no live computation. See the README.md for the build and deploy steps.

Contact

M. Ehsan Karim — School of Population and Public Health, University of British Columbia, and Centre for Advancing Health Outcomes, St. Paul’s Hospital, Vancouver, Canada. ehsan.karim@ubc.ca

License

Code: MIT. Text and figures: CC BY 4.0.

Source Code
---
title: "About"
---

## Citation

This site is the interactive results companion to:

> M. Ehsan Karim. *Cross-Fitted Survey-Weighted TMLE with Design-Based Variance
> for Causal Machine Learning.* Manuscript, 2026.

## Code and data

- The estimator, simulation engine, and NHANES analysis code live in the public
  code repository, <https://github.com/ehsanx/Survey-TMLE-codes>.
- The numbers shown here are pre-computed and **frozen**: every plot and table is
  rebuilt from the CSVs in `data/`, which are snapshots of the manuscript's
  result files (multiple imputation, $m = 40$).
- **The manuscript and Web Appendix are the source of record.** This site is a
  convenience layer; if any number here ever differs from the paper, the paper
  governs.
- Data source: the U.S. National Health and Nutrition Examination Survey
  (NHANES), publicly available from the National Center for Health Statistics.

## How this site is built

A static [Quarto](https://quarto.org) website rendered to `docs/` and served
from GitHub Pages — no server, no live computation. See the `README.md` for the
build and deploy steps.

## Contact

M. Ehsan Karim — School of Population and Public Health, University of British
Columbia, and Centre for Advancing Health Outcomes, St. Paul's Hospital,
Vancouver, Canada. `ehsan.karim@ubc.ca`

## License

Code: MIT. Text and figures: CC BY 4.0.
 

Companion site to Cross-Fitted Survey-Weighted TMLE with Design-Based Variance for Causal Machine Learning. The manuscript and Web Appendix are the source of record.