Doubly Robust Estimation of Causal Effects
Open accessThe DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index.
Abstract
Doubly robust estimation combines a form of outcome regression with a model for the exposure (i.e., the propensity score) to estimate the causal effect of an exposure on an outcome. When used individually to estimate a causal effect, both outcome regression and propensity score methods are unbiased only if the statistical model is correctly specified. The doubly robust estimator combines these 2 approaches such that only 1 of the 2 models need be correctly specified to obtain an unbiased effect estimator. In this introduction to doubly robust estimators, the authors present a conceptual overview of doubly robust estimation, a simple worked example, results from a simulation study examining performance of estimated and bootstrapped standard errors, and a discussion of the potential advantages and limitations of this method. The supplementary material for this paper, which is posted on the Journal's Web site (http://aje.oupjournals.org/), includes a demonstration of the doubly robust property (Web Appendix 1) and a description of a SAS macro (SAS Institute, Inc., Cary, North Carolina) for doubly robust estimation, available for download at http://www.unc.edu/~mfunk/dr/.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“regression models were re-estimated within these matched samples and included all covariates when predicting outcomes (doubly robust estimation; Funk et al., 2011).”
From School uniforms and student behavior: is there a link? (2022) · cites this paper as Funk et al., 2011
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The independent variables of the MLR included not only exposure (i.e., receiving SSRI/SNRI vs. psychotherapy) but the unbalanced covariates (ASMD > 0.1) after sIPTW (i.e., doubly robust approach) [30].”
From Association between first-line antidepressant use and risk of dementia in older adults: a retrospective cohort study (2023) · cites this paper as [30]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
An Introduction to the Bootstrap
Efron, B., & Tibshirani, R. J. (1994) · doi:10.1201/9780429246593 · Cited by 40,500
The central role of the propensity score in observational studies for causal effects
ROSENBAUM, P. R., & RUBIN, D. B. (1983) · Biometrika · doi:10.1093/biomet/70.1.41 · Cited by 31,241 · cited 1×
“The estimated parameters from this model can be used in conjunction with each individual’s actual covariate values to calculate the predicted probability of statin initiation conditional on those covariates, the PS, for each person in the cohort (11).”
MethodAn Introduction to the Bootstrap
Efron, B., & Tibshirani, R. J. (1993) · doi:10.1007/978-1-4899-4541-9 · Cited by 21,230
Marginal Structural Models and Causal Inference in Epidemiology
Robins, J. M., Hernán, M. Á., & Brumback, B. (2000) · Epidemiology · doi:10.1097/00001648-200009000-00011 · Cited by 5,581
Two-sided confidence intervals for the single proportion: comparison of seven methods
Newcombe, R. G. (1998) · Statistics in Medicine · doi:10.1002/(sici)1097-0258(19980430)17:8<857::aid-sim777>3.0.co;2-e · Cited by 5,107 · cited 1×
“Two-sided 95% confidence intervals on the estimated confidence interval coverage were calculated using the Wilson score method without continuity correction (23).”
MethodEstimation of Regression Coefficients When Some Regressors are not Always Observed
Robins, J. M., Rotnitzky, A., & Zhao, L. P. (1994) · Journal of the American Statistical Association · doi:10.1080/01621459.1994.10476818 · Cited by 2,823
Constructing Inverse Probability Weights for Marginal Structural Models
Cole, S. R., & Hernan, M. A. (2008) · American Journal of Epidemiology · doi:10.1093/aje/kwn164 · Cited by 2,796 · cited 3×
“estimators, it is known that weights for individuals with unusual combinations of characteristics and exposures can lead to unstable estimates with relatively large standard errors (19).”
BackgroundRobins, J. (1986) · Mathematical Modelling · doi:10.1016/0270-0255(86)90088-6 · Cited by 2,451
Doubly robust estimation in missing data and causal inference models
Bang, H., & Robins, J. M. (2005) · Biometrics · doi:10.1111/j.1541-0420.2005.00377.x · Cited by 1,998 · cited 1×
“Prior simulations have confirmed that the doubly robust estimator is unbiased when a confounder is omitted from 1 (but not both) of the component models (3, 20).”
BackgroundLunceford, J. K., & Davidian, M. (2004) · Statistics in Medicine · doi:10.1002/sim.1903 · Cited by 1,674 · cited 6×
“Thus, there is a trade-off to consider between potentially reducing bias at the expense of precision (20).”
BackgroundMethodInfluentialSelf-citationBootstrapping: A Nonparametric Approach to Statistical Inference.
AA, Mooney, C. Z., & Duval, R. D. (1994) · Journal of the American Statistical Association · doi:10.2307/2290969 · Cited by 1,351 · cited 1×
“) We simulated 1,000 cohorts of size n (where n 1⁄4 100, 500, 1,000, or 2,000), and, within each simulated cohort, we bootstrapped 1,000 complete resamples with replacement (21, 22).”
MethodEstimating causal effects from epidemiological data
Hernán, M. A., & Robins, J. M. (2006) · Journal of Epidemiology and Community Health · doi:10.1136/jech.2004.029496 · Cited by 1,028 · cited 1×
“Inverse probability weights are calculated as the inverse of the conditional probability that an individual received the exposure he or she actually received, that is, 1/PS for the exposed and 1/ (1 PS) for the unexposed (12, 13).”
BackgroundUnified Methods for Censored Longitudinal Data and Causality
van der Laan, M. J., & Robins, J. M. (2003) · Springer Series in Statistics · doi:10.1007/978-0-387-21700-0 · Cited by 959
Confounding and Collapsibility in Causal Inference
Greenland, S., Pearl, J., & Robins, J. M. (1999) · Statistical Science · doi:10.1214/ss/1009211805 · Cited by 933 · cited 1×
“If the distributions of confounders are the same within each exposure group, then there is no longer an association between the confounders and exposure, making the exposed and unexposed exchangeable (15).”
BackgroundMooney, C., & Duval, R. (1993) · doi:10.4135/9781412983532 · Cited by 893
Semiparametric Theory and Missing Data
(2006) · Springer Series in Statistics · doi:10.1007/0-387-37345-4 · Cited by 827
Targeted Maximum Likelihood Learning
van der Laan, M. J., & Rubin, D. (2006) · The International Journal of Biostatistics · doi:10.2202/1557-4679.1043 · Cited by 823
Kang, J. D. Y., & Schafer, J. L. (2007) · Statistical Science · doi:10.1214/07-sts227 · Cited by 797 · cited 1×
“Moreover, when both models are misspecified, the resulting effect estimate may be more biased than that of a single, misspecified maximum likelihood model (26).”
BackgroundIdentifiability, Exchangeability, and Epidemiological Confounding
GREENLAND, S., & ROBINS, J. M. (1986) · International Journal of Epidemiology · doi:10.1093/ije/15.3.413 · Cited by 686 · cited 1×
“The fundamental assumptions required for the effect estimates to have a causal interpretation include exchangeability (16), positivity (17), consistency (18), and no interference (19).”
BackgroundKang, J., & Schafer, J. L. (2007) · Statistical Science · Cited by 508
Westreich, D., Lessler, J., & Funk, M. J. (2010) · Journal of Clinical Epidemiology · doi:10.1016/j.jclinepi.2009.11.020 · Cited by 502 · cited 1×
“(Alternatively one could use, for example, a probit model or a machine learning approach (9, 10)).”
MethodSelf-citationSturmer, T., Rothman, K. J., Avorn, J., & Glynn, R. J. (2010) · American Journal of Epidemiology · doi:10.1093/aje/kwq198 · Cited by 446
Self-citationSemiparametric Theory and Missing Data
Fokianos, K. (2007) · Technometrics · doi:10.1198/tech.2007.s488 · Cited by 418
The Consistency Statement in Causal Inference: A Definition or an Assumption?
Cole, S. R., & Frangakis, C. E. (2009) · Epidemiology · doi:10.1097/ede.0b013e31818ef366 · Cited by 413
Invited Commentary: Positivity in Practice
Westreich, D., & Cole, S. R. (2010) · American Journal of Epidemiology · doi:10.1093/aje/kwp436 · Cited by 412 · cited 1×
“The fundamental assumptions required for the effect estimates to have a causal interpretation include exchangeability (16), positivity (17), consistency (18), and no interference (19).”
Self-citationRobins, J. M., Sued, M., Lei-Gomez, Q., Rotnitzky, A. G., Ciencias, F., & Naturales, E. (2010) · Cited by 346
Robins, J., Sued, M., Lei-Gomez, Q., & Rotnitzky, A. (2007) · Statistical Science · doi:10.1214/07-sts227d · Cited by 313
Robins, J. (1987) · Journal of Chronic Diseases · doi:10.1016/s0021-9681(87)80018-8 · Cited by 305 · cited 2×
“2011;173(7):761–767 D ow naded rom http/academ ic.”
MethodEvaluating uses of data mining techniques in propensity score estimation: a simulation study
Setoguchi, S., Schneeweiss, S., Brookhart, M. A., Glynn, R. J., & Cook, E. F. (2008) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.1555 · Cited by 271 · cited 1×
“(Alternatively one could use, for example, a probit model or a machine learning approach (9, 10)).”
MethodSelf-citationTargeted Maximum Likelihood Based Causal Inference: Part II
van der Laan, M. J. (2010) · The International Journal of Biostatistics · doi:10.2202/1557-4679.1241 · Cited by 239
Stürmer, T., Rothman, K. J., & Glynn, R. J. (2006) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.1231 · Cited by 120 · cited 1×
“Weighting by this quantity creates a pseudopopulation in which the distributions of confounders among the exposed and unexposed are the same as the overall distribution of those confounders in the original total population (14).”
BackgroundSelf-citationAdjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models: Comment
Freedman, D. A. (1999) · Journal of the American Statistical Association · doi:10.2307/2669924 · Cited by 20
Causal effects in longitudinal studies: Definition and maximum likelihood estimation
Neugebauer, R. S., & van der Laan, M. J. (2006) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2006.06.013 · Cited by 8 · cited 1×
“) This approach is more formally known as estimation by maximum likelihood of the g-computation formula (6, 7) and is the equivalent of maximum likelihood estimation of the parameters of a marginal structural model (8).”
Method
Research works printed but not linked: 1
- Jonsson Funk, 2008, Doubly robust estimation under realistic conditions of model misspecification [abstract], Pharmacoepidemiol Drug Saf, 17, S241
Cited by
Long-term cardiovascular outcomes of COVID-19
Xie, Y., Xu, E., Bowe, B., & Al-Aly, Z. (2022) · Nature Medicine · doi:10.1038/s41591-022-01689-3 · Cited by 2,029
Muller, C. J., & MacLehose, R. F. (2014) · International Journal of Epidemiology · doi:10.1093/ije/dyu029 · Cited by 775
Estimating individual treatment effect: generalization bounds and algorithms
Shalit, U., Johansson, F., & Sontag, D. A. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1606.03976 · Cited by 127
Propensity Score Methods for Confounding Control in Nonexperimental Research
Brookhart, M. A., Wyss, R., Layton, J. B., & Stürmer, T. (2013) · Circulation: Cardiovascular Quality and Outcomes · doi:10.1161/circoutcomes.113.000359 · Cited by 588
“One principled approach to combining regression and PS weighting is through the use of the augmented IPTW estimator.(23,24) This estimator depends on both a PS and multivariable outcome model and results in an estimate of the ATE.”
MethodSelf-citationIs financial inclusion good for bank stability? International evidence
Ahamed, M. M., & Mallick, S. K. (2019) · Journal of Economic Behavior & Organization · doi:10.1016/j.jebo.2017.07.027 · Cited by 577
Metadata
- Authors
- Funk, M. J., Westreich, D., Wiesen, C., Stürmer, T., Brookhart, M. A., & Davidian, M.
- Affiliations
- University of North Carolina at Chapel Hill (US)
- Venue
- American Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2011-03-08
- Volume
- 173
- Issue
- 7
- Pages
- 761-767
- ISSN
- 0002-9262 (ISSN-L)
- 1476-6256
- PMC
- PMC3070495
- OpenAlex
- W2144387798
- Cited by
- 1,158
- References
- 33
- Citations by year
- 2026: 136 · 2025: 163 · 2024: 163 · 2023: 144 · 2022: 126 · 2021: 106 · 2020: 78 · 2019: 72 · 2018: 52 · 2017: 40 · 2016: 33 · 2015: 15 · 2014: 8 · 2013: 16 · 2012: 3
- FWCI
- 7.83
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- American Journal of Epidemiology (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- Carolina Digital Repository (University of North Carolina at Chapel Hill) (submittedVersion)
- UNC Libraries
- Funders
- North Carolina State University
- Agency for Healthcare Research and Quality (3 U18 HS010397, K02HS017950, U18 HS 010397)
- University of North Carolina at Chapel Hill
- National Institute on Aging (K25AG027400, RO1 AG023178, AG023178, R01AG023178)
- National Institute of Allergy and Infectious Diseases (5T32AI07001, T32AI007001)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Epidemiology
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Keywords
- doubly robust estimation, causal effect estimation, propensity score, outcome regression, model misspecification, estimation bias, SAS macro
- MeSH terms
- Computer Simulation, Epidemiologic Methods, Humans, Monte Carlo Method, Regression Analysis, Models, Statistical, Causality, Confounding Factors, Epidemiologic, Confidence Intervals, Likelihood Functions, Propensity Score
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.