A different perspective of cross-world independence assumption and the utility of natural effects versus controlled effects

by   Ian Shrier, et al.

The pure effects described by Robins and Greenland, and later called natural effects by Pearl, have been criticized because they require a cross-world independence assumption. In this paper, we use potential outcomes and sufficient causal sets to present a conceptual perspective of the cross-world independence assumption that explains why the clinical utility of natural effects is sometimes greater than that of controlled effects. Our perspective is consistent with recent work on mediation of natural effects, path specific effects and separable effects.



page 1

page 2

page 3

page 4


Insights into the "cross-world" independence assumption of causal mediation analysis

Causal mediation analysis is a useful tool for epidemiological research,...

Representing Independence Models with Elementary Triplets

In an independence model, the triplets that represent conditional indepe...

Causal Shapley Values: Exploiting Causal Knowledge to Explain Individual Predictions of Complex Models

Shapley values underlie one of the most popular model-agnostic methods w...

A Potential Outcomes Calculus for Identifying Conditional Path-Specific Effects

The do-calculus is a well-known deductive system for deriving connection...

A Uniformly Consistent Estimator of non-Gaussian Causal Effects Under the k-Triangle-Faithfulness Assumption

Kalisch and Bühlmann (2007) showed that for linear Gaussian models, unde...

An Interventionist Approach to Mediation Analysis

Judea Pearl's insight that, when errors are assumed independent, the Pur...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.