Stratified Analysis of `Probabilities of Causation'

TL;DR

This paper introduces improved formulas and bounds for the probabilities of causation using covariate information, enhancing causal inference accuracy for policy and decision-making.

Contribution

It derives narrower bounds for probabilities of causation by incorporating covariate data and discusses covariate selection for better estimation under no-prevention assumptions.

Findings

Narrower bounds for probabilities of causation using covariates.

Identifiable case under no-prevention assumption.

Insights into covariate selection for estimation accuracy.

Abstract

This paper proposes new formulas for the probabilities of causation difined by Pearl (2000). Tian and Pearl (2000a, 2000b) showed how to bound the quantities of the probabilities of causation from experimental and observational data, under the minimal assumptions about the data-generating process. We derive narrower bounds than Tian-Pearl bounds by making use of the covariate information measured in experimental and observational studies. In addition, we provide identifiable case under no-prevention assumption and discuss the covariate selection problem from the viewpoint of estimation accuracy. These results are helpful in providing more evidence for public policy assessment and dicision making problems.