Research map / Probabilistic and scientific modeling / Treatment effects
An Online Sequential Test for Qualitative Treatment Effects
Chengchun Shi, Shikai Luo, Hongtu Zhu, Rui Song
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 BAT boundary instead of LIL boundary | BAT · QTE Scenario 1, (200,5), random | rejection probability (x100): 0.0 -> 79.2 | Improved |
| more frequent interim analyses, (n,K) from (200,5) to (20,50) | BAT · QTE Scenario 1, delta=0.15, random | rejection probability (x100): 79.2 -> 95.2 | Improved |
| increase effect size delta from 0 to 0.15 | BAT · QTE Scenario 1, (200,5), random | rejection probability (x100): 5.5 -> 79.2 | Improved |
| increase effect size delta from 0 to 0.15 | BAT · QTE Scenario 1, (200,5), random | average stopping time: 3522 -> 2554 | Improved |
| adaptive epsilon-greedy randomization instead of random design | BAT · QTE Scenario 1, (200,5), delta=0.15 | rejection probability (x100): 79.2 -> 76.2 | No clear effect |
| use BAT instead of AVT under adaptive design | BAT · ATE Scenario 1, (200,5), delta=0, adaptive | type-I error (rejection probability x100): 51.2 -> 6.2 | Improved |
| use BAT instead of AVT at small effect, random design | BAT · ATE Scenario 1, (200,5), delta=0.10 | rejection probability (x100): 6.8 -> 27.5 | Improved |
| use BAT instead of AVT at large effect, random design | BAT · ATE Scenario 1, (200,5), delta=0.30 | rejection probability (x100): 89.2 -> 88.2 | No clear effect |
| add normalization (studentization) of the supremum statistic (ablation) | BAT · QTE simulations | power | Improved |
| more frequent interim analyses, (n,K) from (200,5) to (20,50) | BAT · QTE Scenario 2, delta=0.15, random | rejection probability (x100): 43.8 -> 99.8 | Improved |
Closest papers by meaning
- Efficient Adaptive Experimentation with Noncompliance (NeurIPS 2025 · similarity 0.83)
- Active Adaptive Experimental Design for Treatment Effect Estimation with Covariate Choice (ICML 2024 · similarity 0.83)
- Benchmarking Estimators for Natural Experiments: A Novel Dataset and a Doubly Robust Algorithm (NeurIPS 2024 · similarity 0.82)
- On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization (ICML 2025 · similarity 0.82)
- Beyond the Average: Distributional Causal Inference under Imperfect Compliance (NeurIPS 2025 · similarity 0.82)
- On Online Experimentation without Device Identifiers (ICML 2024 · similarity 0.82)
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.
