An Interventionist Approach to Mediation Analysis

@article{Robins2020AnIA,
  title={An Interventionist Approach to Mediation Analysis},
  author={James M. Robins and Thomas S. Richardson and Ilya Shpitser},
  journal={Probabilistic and Causal Inference},
  year={2020}
}
Judea Pearl's insight that, when errors are assumed independent, the Pure (aka Natural) Direct Effect (PDE) is non-parametrically identified via the Mediation Formula was `path-breaking' in more than one sense! In the same paper Pearl described a thought-experiment as a way to motivate the PDE. Analysis of this experiment led Robins \& Richardson to a novel way of conceptualizing direct effects in terms of interventions on an expanded graph in which treatment is decomposed into multiple… 

Insights into the Cross-world Independence Assumption of Causal Mediation Analysis

TLDR
The relationship between assumptions for causal mediation analyses, causal models, and nonparametric identification of natural direct and indirect effects is discussed, and a practical example of an applied setting where the cross-world independence assumption is violated even without any post-treatment confounding is given.

Path-specific Effects Based on Information Accounts of Causality

TLDR
This paper proposes a new path intervention inspired by information accounts of causality, and develops the corresponding intervention diagrams and π-formula that could serve useful communications and theoretical focuses for mediation analysis.

Clarifying causal mediation analysis: Effect identification via three assumptions and five potential outcomes

Abstract Causal mediation analysis is complicated with multiple effect definitions that require different sets of assumptions for identification. This article provides a systematic explanation of

Clarifying causal mediation analysis for the applied researcher: Effect identification via three assumptions and five potential outcomes

TLDR
This presentation shows clearly why an assumption is required for one estimand and not another, and provides a succinct table from which an applied researcher could pick out the assumptions required for identifying the causal effects they target.

Multivariate Counterfactual Systems and Causal Graphical Models

TLDR
This chapter shows that d-separation} provides direct insight into an earlier causal model originally described in terms of potential outcomes and event trees, which leads to a simplification of the do-calculus that clarifies and separates the underlying concepts.

Discussion on “Causal mediation of semicompeting risks” by Yen‐Tsung Huang

We thank the editor for the opportunity to discuss the article by Huang on Causal mediation of semicompeting risks. Huang draws attention to the important task of defining a scientifically meaningful

A Causal Framework for Observational Studies of Discrimination*

In studies of discrimination, researchers often seek to estimate a causal effect of race or gender on outcomes. For example, in the criminal justice context, one might ask whether arrested

Conditional separable effects

TLDR
Alternative estimands are formulated, the conditional separable effects, that have a natural causal interpretation under assumptions that can be falsified in a randomized experiment and can be identified without relying on unfalsifiable assumptions.

On the Causal Interpretation of Randomized Interventional Indirect Effects

Identification of standard mediated effects such as the natural indirect effect relies on heavy causal assumptions. By circumventing such assumptions, so-called randomized interventional indirect

Discussion on “Causal mediation of semicompeting risks” by Yen‐Tsung Huang

TLDR
It is argued that statistical and causal assumptions in the causal mediation framework are not only difficult to interpret and justify, but are also likely violated in the hepatitis B motivating example and other survival/time to event settings as well.

References

SHOWING 1-10 OF 38 REFERENCES

CAUSAL INFERENCE WITH A GRAPHICAL HIERARCHY OF INTERVENTIONS.

TLDR
A unifying view of a large class of causal effects of interest, including novel effects not previously considered, in terms of a hierarchy of interventions is given, and it is shown that identification theory for this large class reduces to an identification theory of random variables under interventions from this hierarchy.

Counterfactual Graphical Models for Longitudinal Mediation Analysis With Unobserved Confounding

TLDR
This article shows the generality of a general counterfactual framework for reasoning about causality first described by Neyman and Rubin and linked to causal graphical models by Robins (1986) and Pearl (2006) by proving a novel result which allows mediation analysis to be applied to longitudinal settings with unobserved confounders.

Alternative Graphical Causal Models and the Identification of Direct E!ects

TLDR
This paper analyzes various measures of the ‘direct’ causal effect, focussing on the pure direct effect (PDE), and introduces the Minimal Counterfactual Model (MCM) which is referred to as ‘minimal’ because it imposes the minimal counterfactual independence assumptions.

A Potential Outcomes Calculus for Identifying Conditional Path-Specific Effects

TLDR
This paper uses po-calculus to give a complete identification algorithm for conditional path-specific effects with applications to problems in mediation analysis and algorithmic fairness.

Generalizing causal inferences from randomized trials: counterfactual and graphical identification

When engagement with a randomized trial is driven by factors that affect the outcome or when trial engagement directly affects the outcome independent of treatment, the average treatment effect among

Multivariate Counterfactual Systems and Causal Graphical Models

TLDR
This chapter shows that d-separation} provides direct insight into an earlier causal model originally described in terms of potential outcomes and event trees, which leads to a simplification of the do-calculus that clarifies and separates the underlying concepts.

Modeling Interference Via Symmetric Treatment Decomposition

TLDR
A new approach to decomposing the spillover effect into direct and indirect components that extends the DAG based treatment decomposition approach to mediation to causal chain graph models and has an identifying functional, which is called the symmetric mediation formula that generalizes the mediation formula in DAGs.

Separable Effects for Causal Inference in the Presence of Competing Events

TLDR
The new separable effects to study the causal effect of a treatment on an event of interest are proposed and do not require cross-world contrasts or hypothetical interventions to prevent death.

Time‐dependent mediators in survival analysis: Modeling direct and indirect effects with the additive hazards model

TLDR
This paper demonstrates that combining the g-formula with the additive hazards model and a sequential linear model for the mediator process results in simple and interpretable expressions for direct and indirect effects in terms of relative survival as well as cumulative hazards.

Conditional separable effects

TLDR
Alternative estimands are formulated, the conditional separable effects, that have a natural causal interpretation under assumptions that can be falsified in a randomized experiment and can be identified without relying on unfalsifiable assumptions.