Research map / Probabilistic and scientific modeling / Treatment effects
Data-driven Automated Negative Control Estimation (DANCE): Search for, Validation of, and Causal Inference with Negative Controls
Erich Kummerfeld, Jaewon Lim, Xu Shi
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 |
|---|---|---|---|
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, complex graph (n=3000) | proportion bias (%): 14.61 -> 0.18 | Improved |
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, simple graph (n=3000) | proportion bias (%): 3.29 -> 0.01 | Improved |
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, complex graph (n=3000) | 95% CI coverage: 0.05 -> 0.94 | Improved |
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, simple graph (n=3000) | 95% CI coverage: 0.70 -> 0.96 | Improved |
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, complex graph, stronger edges (n=3000) | 95% CI coverage: 0.03 -> 0.96 | Improved |
| add DANCE-selected negative controls instead of random NC pair | double-NC estimator · Simulation, simple graph, stronger edges (n=3000) | 95% CI coverage: 0.27 -> 0.96 | Improved |
| use DANCE search for NCs, which adds Monte Carlo variability | double-NC estimator · Simulation, complex graph (n=1000) | Monte Carlo SE (1e-3): 40.59 -> 47.46 | Hurt |
| use DANCE search for NCs, which adds Monte Carlo variability | double-NC estimator · Simulation, simple graph (n=1000) | Monte Carlo SE (1e-3): 21.38 -> 23.88 | Hurt |
| increase sample size from 1000 to 3000 for DANCE | DANCE · Simulation, complex graph | Monte Carlo SE (1e-3): 47.46 -> 28.06 | Improved |
| increase sample size to 300 or more for NC validation | FindNC validation test · Simulation, simple and complex graphs | ROC curve: sample size 300 | Improved |
| use stronger edge strength in the simulated SEMs (ablation) | DANCE · Simulation, complex graph (n=1000) | proportion bias (%): -0.87 -> -0.57 | No clear effect |
| generate all variables as Bernoulli instead of Gaussian (ablation) | DANCE · Simulation, binary variables | ATE bias | No clear effect |
Closest papers by meaning
- Data Fusion for Partial Identification of Causal Effects (NeurIPS 2025 · similarity 0.86)
- Proximal Causal Inference With Text Data (NeurIPS 2024 · similarity 0.86)
- Doubly robust identification of treatment effects from multiple environments (ICLR 2025 · similarity 0.85)
- Detecting and Measuring Confounding Using Causal Mechanism Shifts (NeurIPS 2024 · similarity 0.85)
- Do Contemporary Causal Inference Models Capture Real-World Heterogeneity? Findings from a Large-Scale Benchmark (ICLR 2025 · similarity 0.85)
- Benchmarking Estimators for Natural Experiments: A Novel Dataset and a Doubly Robust Algorithm (NeurIPS 2024 · similarity 0.85)
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.
