Causal Inference through a Witness Protection Program

One of the most fundamental problems in causal inference is the estimation of a causal effect when treatment and outcome are confounded. This is difficult in an observational study, because one has no direct evidence that all confounders have been adjusted for. We introduce a novel approach for estimating causal effects that exploits observational conditional independencies to suggest "weak" paths in an unknown causal graph. The widely used faithfulness condition of Spirtes et al. is relaxed to allow for varying degrees of "path cancellations" that imply conditional independencies but do not rule out the existence of confounding causal paths. The output is a posterior distribution over bounds on the average causal effect via a linear programming approach and Bayesian inference. We claim this approach should be used in regular practice as a complement to other tools in observational studies.

• Publication year

  2014

• Venue

  Journal of machine learning research

• Publication date

  2014-06-02

• Fields of study

  Mathematics, Computer Science, Economics

• Identifiers
  arXiv 1406.0531
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• An empirical study of one of the simplest causal prediction algorithms

  2015cited by this paper
• Addendum on the scoring of Gaussian directed acyclic graphical models

  2014cited by this paper
• Estimating Causal Effects by Bounding Confounding

  2014cited by this paper
• Data-driven covariate selection for nonparametric estimation of causal effects

  2013influential reference
• Copulae in Mathematical and Quantitative Finance

  2013cited by this paper
• PC algorithm for nonparanormal graphical models

  2013cited by this paper
• Copulas in Machine Learning

  2013cited by this paper
• Transparent Parametrizations of Models for Potential Outcomes

  2012cited by this paper
• Supplementary Material for “ Statistical test for consistent estimation of causal effects in linear non-Gaussian models ”

  2012influential reference
• Graphical methods for inequality constraints in marginalized DAGs

  2012cited by this paper
• Causal Bounds and Observable Constraints for Non-deterministic Models

  2012influential reference
• Likelihood analysis of the binary instrumental variable model

  2011cited by this paper
• A New Criterion for Confounder Selection

  2011cited by this paper
• Analysis of the Binary Instrumental Variable Model

  2010cited by this paper
• Myth, Confusion, and Science in Causal Analysis

  2009cited by this paper
• Bounds on Direct Effects in the Presence of Confounded Intermediate Variables

  2008cited by this paper
• Identification for Prediction and Decision

  2008influential reference
• Applied Econometrics with R

  2008cited by this paper
• Harnessing naturally randomized transcription to infer regulatory relationships among genes

  2007cited by this paper
• An Introduction to Copulas (Springer Series in Statistics)

  2006cited by this paper
• A Theoretical Study of Y Structures for Causal Discovery

  2006influential reference
• Extending the rank likelihood for semiparametric copula estimation

  2006cited by this paper
• Adjacency-Faithfulness and Conservative Causal Inference

  2006cited by this paper
• Causation: A Unified Framework for the Social Sciences

  2005cited by this paper
• Selection on Observed and Unobserved Variables: Assessing the Effectiveness of Catholic Schools

  2005cited by this paper
• Being Bayesian About Network Structure. A Bayesian Approach to Structure Discovery in Bayesian Networks

  2004cited by this paper
• Applying Discrete PCA in Data Analysis

  2004cited by this paper
• Monte Carlo sensitivity analysis and Bayesian analysis of smoking as an unmeasured confounder in a study of silica and lung cancer.

  2004cited by this paper
• Causal inference using influence diagrams: The problem of partial compliance (with Discussion)

  2003influential reference
• Observational Studies

  2003influential reference
• Uniform consistency in causal inference

  2003cited by this paper
• Covariance Adjustment in Randomized Experiments and Observational Studies

  2002influential reference
• Assessing the effect of an influenza vaccine in an encouragement design.

  2000cited by this paper
• An Introduction to Copulas

  1998cited by this paper
• Bounds on Treatment Effects from Studies with Imperfect Compliance

  1997influential reference
• A Simple Constraint-Based Algorithm for Efficiently Mining Observational Databases for Causal Relationships

  1997cited by this paper
• Strong completeness and faithfulness in Bayesian networks

  1995cited by this paper
• Causation, prediction, and search

  1993influential reference
• Effects of computer reminders for influenza vaccination on morbidity during influenza epidemics.

  1992cited by this paper
• Theory Refinement on Bayesian Networks

  1991cited by this paper
• The Sensitivity of an Empirical Model of Married Women's Hours of Work to Economic and Statistical Assumptions

  1987cited by this paper
• Smoking and lung cancer: recent evidence and a discussion of some questions.

  1959cited by this paper
• The Statistical Implications of a System of Simultaneous Equations

  1943cited by this paper
• Causality : Models , Reasoning , and Inference

  year unknowninfluential reference

• Bounding Causal Effects with Leaky Instruments

  2024cites this paper
• BudgetIV: Optimal Partial Identification of Causal Effects with Mostly Invalid Instruments

  2024cites this paper
• The Causal Marginal Polytope for Bounding Treatment Effects

  2022influential citation
• Partial Counterfactual Identification from Observational and Experimental Data

  2021cites this paper
• Generalizing experimental results by leveraging knowledge of mechanisms

  2020cites this paper
• A Class of Algorithms for General Instrumental Variable Models

  2020cites this paper
• Differentiable Causal Backdoor Discovery

  2020cites this paper
• Contrastive Algorithmic Fairness: Part 1 (Theory)

  2019cites this paper
• Contrastive Fairness in Machine Learning

  2019cites this paper
• Epidemiology, genetic epidemiology and Mendelian randomisation: more need than ever to attend to detail

  2019cites this paper
• Switching Regression Models and Causal Inference in the Presence of Discrete Latent Variables

  2018cites this paper
• Proof and Uncertainty in Causal Claims

  2018cites this paper
• Switching Regression Models and Causal Inference in the Presence of Latent Variables

  2018cites this paper
• Matching on Balanced Nonlinear Representations for Treatment Effects Estimation

  2017cites this paper
• Robust Causal Estimation in the Large-Sample Limit without Strict Faithfulness

  2017influential citation
• An algebraic characterisation of staged trees : their geometry and causal implications

  2017cites this paper
• Learning Instrumental Variables with Structural and Non-Gaussianity Assumptions

  2017cites this paper
• Counterfactual Fairness

  2017cites this paper
• Matching via Dimensionality Reduction for Estimation of Treatment Effects in Digital Marketing Campaigns

  2016cites this paper