Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data

DOI10.1214/07-STS227zbMATH Open1246.62073arXiv0804.2958OpenAlexW3122781290WikidataQ31157218 ScholiaQ31157218MaRDI QIDQ449788FDOQ449788

Authors: Joseph D. Y. Kang, Joseph L. Schafer 

Publication date: 1 September 2012

Published in: Statistical Science (Search for Journal in Brave)

Abstract: When outcomes are missing for reasons beyond an investigator's control, there are two different ways to adjust a parameter estimate for covariates that may be related both to the outcome and to missingness. One approach is to model the relationships between the covariates and the outcome and use those relationships to predict the missing values. Another is to model the probabilities of missingness given the covariates and incorporate them into a weighted or stratified estimate. Doubly robust (DR) procedures apply both types of model simultaneously and produce a consistent estimate of the parameter if either of the two models has been correctly specified. In this article, we show that DR estimates can be constructed in many ways. We compare the performance of various DR and non-DR estimates of a population mean in a simulated example where both models are incorrect but neither is grossly misspecified. Methods that use inverse-probabilities as weights, whether they are DR or not, are sensitive to misspecification of the propensity model when some estimated propensities are small. Many DR methods perform better than simple inverse-probability weighting. None of the DR methods we tried, however, improved upon the performance of simple regression-based prediction of the missing values. This study does not represent every missing-data problem that will arise in practice. But it does demonstrate that, in at least some settings, two wrong models are not better than one.

Full work available at URL: https://arxiv.org/abs/0804.2958

Recommendations

• A comparison of doubly robust estimators of the mean with missing data
• Introduction to double robust methods for incomplete data
• Doubly Robust Estimation in Missing Data and Causal Inference Models
• Estimation with missing data: beyond double robustness
• A comparative study of doubly robust estimators of the mean with missing data

zbMATH Keywords

causal inferencemissing datapropensity scoremodel-assisted survey estimationweighted estimating equations

Mathematics Subject Classification ID

Point estimation (62F10) Foundational topics in statistics (62A99) Robustness and adaptive procedures (parametric inference) (62F35)

Cites Work

• Title not available (Why is that?)
• Model assisted survey sampling
• On the application of probability theory to agricultural experiments. Essay on principles. Section 9. Translated from the Polish and edited by D. M. Dąbrowska and T. P. Speed
• Title not available (Why is that?)
• Inference and missing data
• Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Bayesian Analysis of Binary and Polychotomous Response Data
• A Generalization of Sampling Without Replacement From a Finite Universe
• Observational studies.
• Doubly Robust Estimation in Missing Data and Causal Inference Models
• Unified methods for censored longitudinal data and causality
• The central role of the propensity score in observational studies for causal effects
• Statistics and Causal Inference
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Estimation of Regression Coefficients When Some Regressors Are Not Always Observed
• Semiparametric Regression for Repeated Outcomes with Nonignorable Nonresponse
• Analysis of Semiparametric Regression Models for Repeated Outcomes in the Presence of Missing Data
• Bayesian Inference in Econometric Models Using Monte Carlo Integration
• Bayesian inference for causal effects: The role of randomization
• Some results on generalized difference estimation and generalized regression estimation for finite populations
• Semiparametric Efficiency in Multivariate Regression Models with Missing Data
• Causal Inference Using Potential Outcomes
• On the Variances of Asymptotically Normal Estimators from Complex Surveys
• Title not available (Why is that?)
• Title not available (Why is that?)
• Inference for semiparametric models: Some questions and an answer. (With comments)
• Characterizing the Estimation of Parameters in Incomplete-Data Problems
• Semiparametric estimation of treatment effect in a pretest-posttest study with missing data (with comments and rejoinder)
• Survey Nonresponse Adjustments for Estimates of Means
• Transformation diagnostics for linear models
• The weighted residual technique for estimating the variance of the general regression estimator of the finite population total
• Applied Bayesian Modeling and Causal Inference from Incomplete‐Data Perspectives
• Title not available (Why is that?)

Cited In (only showing first 100 items - show all)

• A semiparametric method for evaluating causal effects in the presence of error‐prone covariates
• Determination of the optimal number of strata for propensity score subclassification
• Doubly robust policy evaluation and optimization
• Estimation of causal effects in observational studies with interference between units
• On model selection and model misspecification in causal inference
• Effect of breastfeeding on gastrointestinal infection in infants: a targeted maximum likelihood approach for clustered longitudinal data
• Estimation of causal effects with multiple treatments: a review and new ideas
• A method for increasing the robustness of multiple imputation
• Causal inference: a missing data perspective
• Comment: Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data
• Semiparametric estimation for average causal effects using propensity score-based spline
• Model selection and local geometry
• Second-order asymptotic theory for calibration estimators in sampling and missing-data problems
• Introduction to double robust methods for incomplete data
• Semiparametric theory for causal mediation analysis: efficiency bounds, multiple robustness and sensitivity analysis
• A general double robustness result for estimating average treatment effects
• Doubly robust kernel density estimation when group membership is missing at random
• Multiply robust estimation in regression analysis with missing data
• Adjustments and their consequences -- Collapsibility analysis using graphical models
• Covariate balancing propensity score by tailored loss functions
• Bayesian semiparametric modeling of response mechanism for nonignorable missing data
• Fractional imputation in survey sampling: a comparative review
• Improved doubly robust estimation in learning optimal individualized treatment rules
• A note on sliced inverse regression with missing predictors
• A further study of the multiply robust estimator in missing data analysis
• The performance of estimators based on the propensity score
• A comparison of doubly robust estimators of the mean with missing data
• Uniform improvement of empirical likelihood for missing response problem
• A comparative study of doubly robust estimators of the mean with missing data
• A Simple and Efficient Estimation of Average Treatment Effects in Models with Unmeasured Confounders
• Calibrated Bayes, for statistics in general, and missing data in particular
• Discussion of ``Calibrated Bayes, for statistics in general, and missing data in particular by R. Little
• Discussion of ``Impact of frequentist and Bayesian methods on survey sampling practice: a selective appraisal by J. N. K. Rao
• Rejoinder: Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data
• Robust location estimation with missing data
• Improved double-robust estimation in missing data and causal inference models
• A comparison study of nonparametric imputation methods
• The finite sample performance of semi- and non-parametric estimators for treatment effects and policy evaluation
• Causal inference: Critical developments, past and future
• Improved doubly robust estimation when data are monotonely coarsened, with application to longitudinal studies with dropout
• Causal inference in case of near‐violation of positivity: comparison of methods
• Doubly robust estimation with the R package \texttt{drgee}
• Model-assisted inference for treatment effects using regularized calibrated estimation with high-dimensional data
• Doubly robust estimation of attributable fractions
• Comparison of covariate balance weighting methods in estimating treatment effects
• Doubly robust difference-in-differences estimators
• Semiparametric regression and risk prediction with competing risks data under missing cause of failure
• Estimating Treatment Effects of Longitudinal Designs using Regression Models on Propensity Scores
• Mark-specific hazard ratio model with missing multivariate marks
• Doubly Robust Estimation in Missing Data and Causal Inference Models
• Doubly robust estimates for binary longitudinal data analysis with missing response and missing covariates
• Nonparametric likelihood and doubly robust estimating equations for marginal and nested structural models
• Doubly robust point and variance estimation in the presence of imputed survey data
• Identifying the consequences of dynamic treatment strategies: a decision-theoretic overview
• Properties of doubly robust estimators when nuisance functions are estimated nonparametrically
• A stable and more efficient doubly robust estimator
• Title not available (Why is that?)
• Robust inference on average treatment effects with possibly more covariates than observations
• Estimation with missing data: beyond double robustness
• Oracle, multiple robust and multipurpose calibration in a missing response problem
• Statistical data integration in survey sampling: a review
• Covariate balancing propensity score for a continuous treatment: application to the efficacy of political advertisements
• A nonparametric super-efficient estimator of the average treatment effect
• Covariate-balancing-propensity-score-based inference for linear models with missing responses
• Matching methods for causal inference: a review and a look forward
• Doubly robust inference with missing data in survey sampling
• Semiparametric double balancing score estimation for incomplete data with ignorable missingness
• Doubly robust inference for the distribution function in the presence of missing survey data
• Estimating multiple treatment effects using two-phase semiparametric regression estimators
• Approximate Bayesian inference for doubly robust estimation
• On varieties of doubly robust estimators under missingness not at random with a shadow variable
• Double robustness without weighting
• Efficient augmentation and relaxation learning for individualized treatment rules using observational data
• Comment: Stabilizing the doubly-robust estimators of the average treatment effect under positivity violations
• Causal inference with missingness in confounder
• Robust doubly protected estimators for quantiles with missing data
• An application of targeted maximum likelihood estimation to the meta-analysis of safety data
• Semiparametric estimation in regression with missing covariates using single-index models
• Matrix Completion under Low-Rank Missing Mechanism
• Doubly robust estimator for indirectly standardized mortality ratios
• Estimation and inference for the counterfactual distribution and quantile functions in continuous treatment models
• Robust estimation of generalized estimating equations with finite mixture correlation matrices and missing covariates at random for longitudinal data
• Performance Evaluation Based on the Robust Mahalanobis Distance and Multilevel Modeling Using Two New Strategies
• Robust inference using hierarchical likelihood approach for heavy-tailed longitudinal outcomes with missing data: an alternative to inverse probability weighted generalized estimating equations
• Using missing types to improve partial identification with application to a study of HIV prevalence in Malawi
• Confounding adjustment methods for multi-level treatment comparisons under lack of positivity and unknown model specification
• Robust inference when combining inverse-probability weighting and multiple imputation to address missing data with application to an electronic health records-based study of bariatric surgery
• Robust Causal Inference with Continuous Instruments Using the Local Instrumental Variable Curve
• Demystifying a class of multiply robust estimators
• Efficient estimation of quantiles in missing data models
• Covariate balancing functional propensity score for functional treatments in cross-sectional observational studies
• Impact of sufficient dimension reduction in nonparametric estimation of causal effect
• Doubly robust semiparametric inference using regularized calibrated estimation with high-dimensional data
• Valid inference for treatment effect parameters under irregular identification and many extreme propensity scores
• Doubly robust augmented-estimating-equations estimation with nonignorable nonresponse data
• Doubly Robust Inference With Nonprobability Survey Samples
• Multiply robust estimation in nonparametric regression with missing data
• Mean estimation with data missing at random for functional covariables
• Coarsened Propensity Scores and Hybrid Estimators for Missing Data and Causal Inference
• Robust Bayesian inference for big data: combining sensor-based records with traditional survey data

Uses Software

• BayesDA

This page was built for publication: Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q449788)