Efficient Adjustment Sets for Population Average Causal Treatment Effect Estimation in Graphical Models

The method of covariate adjustment is often used for estimation of total treatment effects from observational studies. Restricting attention to causal linear models, a recent article (Henckel et al., 2019) derived two novel graphical criteria: one to compare the asymptotic variance of linear regression treatment effect estimators that control for certain distinct adjustment sets and another to identify the optimal adjustment set that yields the least squares estimator with the smallest asymptotic variance. In this paper we show that the same graphical criteria can be used in non-parametric causal graphical models when treatment effects are estimated using non-parametrically adjusted estimators of the interventional means. We also provide a new graphical criterion for determining the optimal adjustment set among the minimal adjustment sets and another novel graphical criterion for comparing time dependent adjustment sets. We show that uniformly optimal time dependent adjustment sets do not always exist. For point interventions, we provide a sound and complete graphical criterion for determining when a non-parametric optimally adjusted estimator of an interventional mean, or of a contrast of interventional means, is semiparametric efficient under the non-parametric causal graphical model. In addition, when the criterion is not met, we provide a sound algorithm that checks for possible simplifications of the efficient influence function of the parameter. Finally, we find an interesting connection between identification and efficient covariate adjustment estimation. Specifically, we show that if there exists an identifying formula for an interventional mean that depends only on treatment, outcome and mediators, then the non-parametric optimally adjusted estimator can never be globally efficient under the causal graphical model.

[1]  I. Shpitser,et al.  A New Criterion for Confounder Selection , 2011, Biometrics.

[2]  J. Hahn.  On the Role of the Propensity Score in Efficient Semiparametric Estimation of Average Treatment Effects , 1998 .

[3]  N. Breslow.  Design and analysis of case-control studies. , 1982, Annual review of public health.

[4]  James M. Robins,et al.  Probabilistic evaluation of sequential plans from causal models with hidden variables , 1995, UAI.

[5]  J. Robins,et al.  Recovery of Information and Adjustment for Dependent Censoring Using Surrogate Markers , 1992 .

[6]  David Maxwell Chickering,et al.  Learning Equivalence Classes of Bayesian Network Structures , 1996, UAI.

[7]  Emilija Perkovi'c,et al.  Identifying causal effects in maximally oriented partially directed acyclic graphs , 2019, UAI.

[8]  G. Imbens,et al.  Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score , 2000 .

[9]  A. Tsiatis.  Semiparametric Theory and Missing Data , 2006 .

[10]  W. G. Cochran.  The effectiveness of adjustment by subclassification in removing bias in observational studies. , 1968, Biometrics.

[11]  Manabu Kuroki,et al.  Selection of Identifiability Criteria for Total Effects by using Path Diagrams , 2004, UAI.

[12]  R Scheines,et al.  The TETRAD Project: Constraint Based Aids to Causal Model Specification. , 1998, Multivariate behavioral research.

[13]  J. Robins.  A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect , 1986 .

[14]  J. Pearl.  Causality: Models, Reasoning and Inference , 2000 .

[15]  James M. Robins,et al.  A unifying approach for doubly-robust $\ell_1$ regularized estimation of causal contrasts , 2019, 1904.03737.

[16]  D. Madigan,et al.  A characterization of Markov equivalence classes for acyclic digraphs , 1997 .

[17]  M. Maathuis,et al.  On efficient adjustment in causal graphs , 2020, 2002.06825.

[18]  M. Maathuis,et al.  Graphical criteria for efficient total effect estimation via adjustment in causal linear models , 2019, 1907.02435.

[19]  Jin Tian,et al.  On the Testable Implications of Causal Models with Hidden Variables , 2002, UAI.

[20]  Matias D. Cattaneo,et al.  Econometric Methods for Program Evaluation , 2018, Annual Review of Economics.

[21]  Michael I. Jordan.  Graphical Models , 2003 .

[22]  G. Imbens,et al.  Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score , 2002 .

[23]  Peter Bühlmann,et al.  Characterization and Greedy Learning of Interventional Markov Equivalence Classes of Directed Acyclic Graphs (Abstract) , 2011, UAI.

[24]  M. Gail.  The Effect of Pooling Across Strata in Perfectly Balanced Studies , 1988 .

[25]  James M. Robins,et al.  On the Validity of Covariate Adjustment for Estimating Causal Effects , 2010, UAI.

[26]  Takahiro Hayashi,et al.  On Estimating Causal Effects based on Supplemental Variables , 2014, AISTATS.

[27]  A. B. Kahn,et al.  Topological sorting of large networks , 1962, CACM.

[28]  James M. Robins,et al.  Unified Methods for Censored Longitudinal Data and Causality , 2003 .

[29]  H Becher,et al.  Improved odds ratio estimation by post hoc stratification of case-control data. , 1997, Statistics in medicine.

[30]  David R. Cox,et al.  On the consequences of overstratification , 2008 .

[31]  J. Robins,et al.  Estimation of Regression Coefficients When Some Regressors are not Always Observed , 1994 .

[32]  T. Richardson,et al.  Markovian acyclic directed mixed graphs for discrete data , 2013, 1301.6624.

[33]  James M. Robins,et al.  INTRODUCTION TO NESTED MARKOV MODELS , 2014 .

[34]  Manabu Kuroki,et al.  Covariate selection for estimating the causal effect of control plans by using causal diagrams , 2003 .

[35]  W. Newey,et al.  Semiparametric Efficiency Bounds , 1990 .

[36]  J. Robins.  Addendum to “a new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect” , 1987 .

[37]  Aad van der Vaart,et al.  Higher Order Tangent Spaces and Influence Functions , 2014, 1502.00812.

[38]  Maciej Liskiewicz,et al.  Finding Minimal d-separators in Linear Time and Applications , 2019, UAI.

[39]  Preetam Nandy,et al.  Structure Learning of Linear Gaussian Structural Equation Models with Weak Edges , 2017, UAI.

[40]  Joaquin Quiñonero Candela,et al.  Counterfactual reasoning and learning systems: the example of computational advertising , 2013, J. Mach. Learn. Res..

[41]  Yuhao Wang,et al.  Permutation-based Causal Inference Algorithms with Interventions , 2017, NIPS.

[42]  Christopher Meek,et al.  Causal inference and causal explanation with background knowledge , 1995, UAI.

[43]  J. Robins,et al.  Double/Debiased Machine Learning for Treatment and Structural Parameters , 2017 .

[44]  P. Shrout,et al.  Causality and Psychopathology: Finding the Determinants of Disorders and their Cures , 2010 .

[45]  J. Robins,et al.  Semiparametric Efficiency in Multivariate Regression Models with Missing Data , 1995 .

[46]  N. Jewell,et al.  Some surprising results about covariate adjustment in logistic regression models , 1991 .

[47]  Dan Geiger,et al.  Identifying independence in bayesian networks , 1990, Networks.

[48]  W. Haenszel,et al.  Statistical aspects of the analysis of data from retrospective studies of disease. , 1959, Journal of the National Cancer Institute.

[49]  Aapo Hyvärinen,et al.  Causal discovery of linear acyclic models with arbitrary distributions , 2008, UAI.

[50]  Robin J. Evans,et al.  Graphs for Margins of Bayesian Networks , 2014, 1408.1809.

[51]  T. Richardson.  Single World Intervention Graphs ( SWIGs ) : A Unification of the Counterfactual and Graphical Approaches to Causality , 2013 .

[52]  Judea Pearl,et al.  Causal networks: semantics and expressiveness , 2013, UAI.

[53]  Tom Burr,et al.  Causation, Prediction, and Search , 2003, Technometrics.

[54]  Judea Pearl,et al.  Complete Identification Methods for the Causal Hierarchy , 2008, J. Mach. Learn. Res..