Multi-Method Causal Evidence Synthesis: Ranking Candidate Drivers by Convergent Cross-Method Evidence from Observational Data

Manish Gupta, Dipanjan De|August 20, 2026arXiv

Key Takeaway

When inferring causality from observational data, combining evidence across multiple methods—even those with different assumptions—identifies likely causal relationships more reliably than trusting any single method alone.

Summary

This paper presents MCES, a framework that combines outputs from eleven causal inference methods across different mathematical traditions to rank which variables most likely cause outcomes in observational data.

evaluation data

Key Terms

causal-inference observational-data structural-causal-model linear-opinion-pool convergent-evidence