Single World Intervention Graphs ( SWIGs ) : A Unification of the Counterfactual and Graphical Approaches to Causality

By Thomas S. Richardson. 2013

448 citations recorded by the source corpus.

Subfields: Graph Learning, Causal Inference & Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Graph Learning

In Causal Inference & Discovery

Citations in the corpus

Learning Optimal Fair Policies (2018, methodology): “1(c), due to the presence of the path Y ← M → A (Richardson and Robins 2013).”
Causal Reasoning for Algorithmic Fairness (2018, background): “Models that follow this approach and specify solely the univariate marginals of a counterfactual joint distribution are sometimes called single-world models [32].”
Fair Inference on Outcomes (2017, background): “Such models generalize independence models on directed acyclic graphs, also known as Bayesian networks (Pearl 1988), to also encode conditional independence statements on counterfactual random variables (Richardson and Robins 2013).”
Causal Models for Dynamical Systems (2020, background): “Many researchers from various disciplines engaged in this question and developed the fundamental assumptions that are often referred to as invariance, autonomy, or modularity [Wright 1921, Haavelmo 1944, Aldrich 1989, Hoover 1990, Pearl 2009, Richardson and Robins 2013, Imbens and Rubin 2015].”
Counterfactual Normalization: Proactively Addressing Dataset Shift and Improving Reliability Using Causal Mechanisms (2018, methodology): “Graphical Representations of Counterfactuals: The node-splitting operation we introduce in Section 3 is similar to the node-splitting operation in Single World Intervention Graphs (SWIGs) (Richardson and Robins, 2013).”
Causal Structure Learning (2017, methodology): “Without assuming anything on the joint noise distributions under diﬀerent interventions, a causal model can equivalently be formulated via structural equations, a graphical model, or potential outcomes (Richardson and Robins, 2013; Imbens, 2014).”
Long-term Causal Inference Under Persistent Confounding via Data Combination (2022, background): “See Figure 2a for a single world intervention graph illustration [Richardson and Robins, 2013] when intervening on the variable A.”
Quantum common causes and quantum causal models (2016, background): “a record of this value, and preparing Xi to have this value, corresponding to Pðki; Xi jXi Þ 1⁄4 δki;XIi δXOi ;Xi (passive observation of Xi); observing Xi , keeping a record of its value, and repreparing Xi to have a fixed value x, corresponding to Pðki; Xi jXiÞ 1⁄4 δki;XIi δXOi ;x (the sort of int”
Guarding Against Adversarial Domain Shifts with Counterfactual Regularization (2017, unknown): “For a formal justification of using a causal graph and potential outcome notation simultaneously see Richardson and Robins (2013).”
Causal Dantzig: Fast inference in linear structural equation models with hidden variables under additive interventions (2017, background): “Note that we did not make any cross-world assumptions (Richardson and Robins (2013)), that is, we made no assumptions on the joint distribution of Y(x), x ∈ range(X) or on the joint distribution of X(e), e ∈ E .”
On Partial Identification of the Pure Direct Effect (2015, background): “This issue has been discussed extensively (Robins and Richardson, 2010; Richardson and Robins, 2013), and in large part motivated the development of the single-world intervention graphs (SWIGs) of Richardson and Robins (2013).”
Causal mediation analysis for stochastic interventions (2019, unknown): “Related approaches have been proposed by Robins (1986), Spirtes et al. (2000), Dawid (2000) and Richardson and Robins (2013).”
Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates (2020, background): “structural causal models (SCM) (Pearl, 2009b) and potential outcomes (Rubin, 1974, 2005), which are complementary and theoretically connected (Pearl, 2009b; Richardson and Robins, 2013; Morgan and Winship, 2015).”
Causal diagrams for interference (2014, background): “Recently, Richardson and Robins (2013) introduced a new class of causal diagrams called single world intervention graphs (SWIGs).”