Robust inference using inverse probability weighting

DOI10.1080/01621459.2019.1660173zbMATH Open1458.62076arXiv1810.11397OpenAlexW2972432428WikidataQ127239696 ScholiaQ127239696MaRDI QIDQ5146038FDOQ5146038

Authors: Xinwei Ma, Jingshen Wang 

Publication date: 22 January 2021

Published in: Journal of the American Statistical Association (Search for Journal in Brave)

Abstract: Inverse Probability Weighting (IPW) is widely used in empirical work in economics and other disciplines. As Gaussian approximations perform poorly in the presence of "small denominators," trimming is routinely employed as a regularization strategy. However, ad hoc trimming of the observations renders usual inference procedures invalid for the target estimand, even in large samples. In this paper, we first show that the IPW estimator can have different (Gaussian or non-Gaussian) asymptotic distributions, depending on how "close to zero" the probability weights are and on how large the trimming threshold is. As a remedy, we propose an inference procedure that is robust not only to small probability weights entering the IPW estimator but also to a wide range of trimming threshold choices, by adapting to these different asymptotic distributions. This robustness is achieved by employing resampling techniques and by correcting a non-negligible trimming bias. We also propose an easy-to-implement method for choosing the trimming threshold by minimizing an empirical analogue of the asymptotic mean squared error. In addition, we show that our inference procedure remains valid with the use of a data-driven trimming threshold. We illustrate our method by revisiting a dataset from the National Supported Work program.

Full work available at URL: https://arxiv.org/abs/1810.11397

Recommendations

• Nonparametric Inverse-Probability-Weighted Estimators Based on the Highly Adaptive Lasso
• Inverse probability weighted estimation for general missing data problems
• Nonparametric inference for an inverse-probability-weighted estimator with doubly truncated data
• Inverse Probability Weighting with Missing Predictors of Treatment Assignment or Missingness
• Inverse probability weighting with error-prone covariates

zbMATH Keywords

trimminginverse probability weightingbias correctionrobust inferenceheavy tail

Mathematics Subject Classification ID

Applications of statistics to economics (62P20) Causal inference from observational studies (62D20) Robustness and adaptive procedures (parametric inference) (62F35)

Cites Work

• Semiparametric instrumental variable estimation of treatment response models.
• Title not available (Why is that?)
• Causal inference for statistics, social, and biomedical sciences. An introduction
• Semiparametric Difference-in-Differences Estimators
• Dealing with limited overlap in estimation of average treatment effects
• Estimation of Regression Coefficients When Some Regressors Are Not Always Observed
• Approximate residual balancing: debiased inference of average treatment effects in high dimensions
• Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score
• Does matching overcome LaLonde's critique of nonexperimental estimators?
• Efficient semiparametric estimation of multi-valued treatment effects under ignorability
• Inverse probability weighted estimation for general missing data problems
• Robust inference on average treatment effects with possibly more covariates than observations
• Large sample confidence regions based on subsamples under minimal assumptions
• Irregular identification, support conditions, and inverse weight estimation
• On the Effect of Bias Estimation on Coverage Accuracy in Nonparametric Inference
• Subsampling inference for the mean in the heavy-tailed case.
• Bootstrap of the mean in the infinite variance case
• Limit distributions of self-normalized sums
• Title not available (Why is that?)
• On the bootstrap of the sample mean in the infinite variance case
• Additions and correction to ``The bootstrap of the mean with arbitrary bootstrap sample size
• Kernel-based semiparametric estimators: small bandwidth asymptotics and bootstrap consistency
• Two-step estimation and inference with possibly many included covariates
• Stable Limit Theory for the Variance Targeting Estimator

Cited In (21)

• Hypothesis Testing for Matched Pairs with Missing Data by Maximum Mean Discrepancy: An Application to Continuous Glucose Monitoring
• Estimation and inference for moments of ratios with robustness against large trimming bias
• Instability of inverse probability weighting methods and a remedy for nonignorable missing data
• Two-stage communication-efficient distributed sparse M-estimation with missing data
• Valid inference for treatment effect parameters under irregular identification and many extreme propensity scores
• Nonparametric weighted average quantile derivative
• Subgroup analysis and adaptive experiments crave for debiasing
• Nonparametric Inverse-Probability-Weighted Estimators Based on the Highly Adaptive Lasso
• Nonparametric identification and estimation of heterogeneous causal effects under conditional independence
• Causal inference of general treatment effects using neural networks with a diverging number of confounders
• Multiple Testing and the Distributional Effects of Accountability Incentives in Education
• Causal inference in the absence of positivity: the role of overlap weights
• On uniform inference in nonlinear models with endogeneity
• Nonparametric tests for treatment effect heterogeneity in observational studies
• Mahalanobis balancing: a multivariate perspective on approximate covariate balancing
• A compound representation of the multiple treatment propensity score with applications to marginal structural modeling
• Multiply robust estimation of the average treatment effect with missing outcomes
• Information projection approach to smoothed propensity score weighting for handling selection bias under missing at random
• Comment: Performance of double-robust estimators when ``inverse probability weights are highly variable
• Identification and estimation of spillover effects in randomized experiments
• Model misspecification and bias for inverse probability weighting estimators of average causal effects

This page was built for publication: Robust inference using inverse probability weighting

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5146038)