Research map / Probabilistic and scientific modeling / Treatment effects
Causal effects of intervening variables in settings with unmeasured confounding
Lan Wen, Aaron Sarvet, Mats Stensrud
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use weighted ICE instead of IPW under misspecification | exposure/mediator/outcome estimators · Simulation scenario 4 (n=500) | standardized bias (x100): 365.37 -> 1.02 | Improved |
| use weighted ICE instead of ICE under misspecification | ICE · Simulation scenario 3 (n=500) | standardized bias (x100): -83.93 -> 5.33 | Improved |
| use weighted ICE instead of ICE under misspecification | ICE · Simulation scenario 4 (n=500) | standardized bias (x100): 49.71 -> 1.02 | Improved |
| use weighted ICE instead of AIPW at small sample size | AIPW · Simulation scenario 1 (n=100) | standardized bias (x100): 25.03 -> 10.82 | Improved |
| use weighted ICE instead of AIPW at small sample size | AIPW · Simulation scenario 3 (n=100) | standardized bias (x100): 27.71 -> 8.551 | Improved |
| use weighted ICE instead of AIPW (sample boundedness) | AIPW · Simulation scenarios 1, 3, 4 (n=100) | standardized bias (x100): 90, 59 and 88 simulated data sets fell below 0 | Improved |
| increase sample size from 100 to 500 (all models correct) | weighted ICE · Simulation scenario 1 | standardized bias (x100): 10.82 -> 6.96 | Improved |
| replace IPW by ICE/AIPW/weighted ICE when models are correct | IPW · Simulation scenario 1 (n=500) | standardized bias (x100): 0.94 -> 0.68 | Improved |
| use weighted ICE instead of ICE | ICE · NHANES (3- and 5-year cumulative incidence) | cumulative incidence: 4.76 -> 4.76 | No clear effect |
| use IPW instead of weighted ICE | weighted ICE · NHANES 5-year causal contrast | E(Y) - E(Y^{a_M=0}) (%): 0.25 -> -0.03 | No clear effect |
Closest papers by meaning
- Automating the Selection of Proxy Variables of Unmeasured Confounders (ICML 2024 · similarity 0.89)
- End-To-End Causal Effect Estimation from Unstructured Natural Language Data (NeurIPS 2024 · similarity 0.87)
- A Non-parametric Direct Learning Approach to Heterogeneous Treatment Effect Estimation under Unmeasured Confounding (NeurIPS 2024 · similarity 0.87)
- Proximal Causal Inference With Text Data (NeurIPS 2024 · similarity 0.87)
- An Analysis of Causal Effect Estimation using Outcome Invariant Data Augmentation (NeurIPS 2025 · similarity 0.87)
- Inferring the Long-Term Causal Effects of Long-Term Treatments from Short-Term Experiments (ICML 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
