Regression‐assisted inference for the average treatment effect in paired experiments

Summary In paired randomized experiments, individuals in a given matched pair may differ on prognostically important covariates despite the best efforts of practitioners. We examine the use of regression adjustment to correct for persistent covariate imbalances after randomization, and present two regression‐assisted estimators for the sample average treatment effect in paired experiments. Using the potential outcomes framework, we prove that these estimators are consistent for the sample average treatment effect under mild regularity conditions even if the regression model is improperly specified, and describe how asymptotically conservative confidence intervals can be constructed. We demonstrate that the variances of the regression‐assisted estimators are no larger than that of the standard difference‐in‐means estimator asymptotically, and illustrate the proposed methods by simulation. The analysis does not require a superpopulation model, a constant treatment effect, or the truth of the regression model, and hence provides inference for the sample average treatment effect with the potential to increase power without unrealistic assumptions.

• Publication year

  2016

• Venue

  Biometrika

• Publication date

  2016-12-15

• Fields of study

  Mathematics, Economics

• Identifiers
  DOI 10.1093/biomet/asy034arXiv 1612.05179
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• On mitigating the analytical limitations of finely stratified experiments

  2017cited by this paper
• Decomposing Treatment Effect Variation

  2016cited by this paper
• Lasso adjustments of treatment effect estimates in randomized experiments

  2015cited by this paper
• Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction

  2015cited by this paper
• SHARP BOUNDS ON THE VARIANCE IN RANDOMIZED EXPERIMENTS

  2014cited by this paper
• A paradox from randomization-based causal inference

  2014influential reference
• A Class of Unbiased Estimators of the Average Treatment Effect in Randomized Experiments

  2013influential reference
• Agnostic notes on regression adjustments to experimental data: Reexamining Freedman's critique

  2012influential reference
• Improving Efficiency of Inferences in Randomized Clinical Trials Using Auxiliary Covariates

  2008cited by this paper
• On regression adjustments to experimental data

  2008cited by this paper
• Comment: Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data.

  2008cited by this paper
• Covariate adjustment for two‐sample treatment comparisons in randomized clinical trials: A principled yet flexible approach

  2008cited by this paper
• Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data

  2007cited by this paper
• Optimal multivariate matching before randomization.

  2004influential reference
• Covariance Adjustment in Randomized Experiments and Observational Studies

  2002influential reference
• Using Heteroscedasticity Consistent Standard Errors in the Linear Regression Model

  2000cited by this paper
• [On the Application of Probability Theory to Agricultural Experiments. Essay on Principles. Section 9.] Comment: Neyman (1923) and Causal Inference in Experiments and Observational Studies

  1990cited by this paper
• On the Application of Probability Theory to Agricultural Experiments. Essay on Principles. Section 9

  1990cited by this paper
• The Jackknife: The Bootstrap and Other Resampling Plans.

  1983cited by this paper
• Estimating causal effects of treatments in randomized and nonrandomized studies.

  1974cited by this paper
• Theory of rank tests

  1969cited by this paper
• Estimates of Location Based on Rank Tests

  1963cited by this paper
• (www.interscience.wiley.com) DOI: 10.1002/sim.3337 Variance identification and efficiency analysis in randomized experiments under the matched-pair design

  year unknowncited by this paper

• A New Design-Based Variance Estimator for Finely Stratified Experiments

  2025influential citation
• Design-based causal inference in bipartite experiments

  2025cites this paper
• Design-based theory for causal inference

  2025cites this paper
• Design-based finite-sample analysis for regression adjustment

  2025influential citation
• Neumann-series corrections for regression adjustment in randomized experiments

  2025influential citation
• One-shot variable-ratio matching with fine balance

  2025cites this paper
• Covariate Adjustment Cannot Hurt: Treatment Effect Estimation under Interference with Low-Order Outcome Interactions

  2025cites this paper
• Manipulating an Instrumental Variable in an Observational Study of Premature Babies: Design, Bounds, and Inference

  2025cites this paper
• What randomization can and cannot guarantee

  2025cites this paper
• Adjustments with many regressors under covariate-adaptive randomizations

  2025cites this paper
• Risk attitudes towards climate variation and food crop diversification, and their implications for food security among farm households in Central Region, Ghana

  2025cites this paper
• Manipulating a Continuous Instrumental Variable in an Observational Study of Premature Babies: Algorithm, Partial Identification Bounds, and Inference under Randomization and Biased Randomization Assumptions

  2024cites this paper
• Sharp Variance Estimator and Causal Bootstrap in Stratified Randomized Experiments

  2024cites this paper
• Integer Programming for Generalized Causal Bootstrap Designs

  2024cites this paper
• Some theoretical foundations for the design and analysis of randomized experiments

  2024cites this paper
• Covariate adjustment in experiments with matched pairs

  2024cites this paper
• Covariate Adjustment in Experiments with Matched Pairs

  2023cites this paper
• Regression adjustment in completely randomized experiments with many covariates

  2023cites this paper
• Causal inference in network experiments: Regression-based analysis and design-based properties

  2023cites this paper
• Adjustment with Many Regressors Under Covariate-Adaptive Randomizations

  2023cites this paper
• Covariate adjustment in stratified experiments

  2023cites this paper
• Efficient Covariate Adjustment in Stratified Experiments

  2023cites this paper
• On randomization inference after inexact Mahalanobis matching

  2022cites this paper
• Optimality of Matched-Pair Designs in Randomized Controlled Trials

  2022cites this paper
• E M ] 9 F eb 2 02 2 Regression Adjustments under Covariate-Adaptive Randomizations with Imperfect Compliance ∗

  2022cites this paper
• Inference for matched tuples and fully blocked factorial designs

  2022cites this paper
• Design-based theory for Lasso adjustment in randomized block experiments with a general blocking scheme

  2022cites this paper
• Covariate-adaptive randomization inference in matched designs

  2022cites this paper
• Improving Estimation Efficiency via Regression-Adjustment in Covariate-Adaptive Randomizations with Imperfect Compliance

  2022cites this paper
• Berry–Esseen bounds for design-based causal inference with possibly diverging treatment levels and varying group sizes

  2022cites this paper
• Design-Based Theory for Lasso Adjustment in Randomized Block Experiments and Rerandomized Experiments

  2021cites this paper
• No star is good news: A unified look at rerandomization based on p-values from covariate balance tests

  2021cites this paper
• Efficiency of Regression (Un)-Adjusted Rosenbaum’s Rank-based Estimator in Randomized Experiments

  2021influential citation
• Regression-Adjusted Estimation of Quantile Treatment Effects under Covariate-Adaptive Randomizations

  2021cites this paper
• Reconciling design-based and model-based causal inferences for split-plot experiments

  2021cites this paper
• Model‐assisted analyses of cluster‐randomized experiments

  2021cites this paper
• Interactive identification of individuals with positive treatment effect while controlling false discoveries

  2021cites this paper
• Randomization Tests to Assess Covariate Balance When Designing and Analyzing Matched Datasets

  2021influential citation
• Blocking, rerandomization, and regression adjustment in randomized experiments with high-dimensional covariates

  2021cites this paper
• Model-Assisted Complier Average Treatment Effect Estimates in Randomized Experiments with Noncompliance

  2021influential citation
• Gaussian prepivoting for finite population causal inference

  2020influential citation
• Causal Inference in Randomized Experiments

  2020cites this paper
• Bridging preference‐based instrumental variable studies and cluster‐randomized encouragement experiments: Study design, noncompliance, and average cluster effect ratio

  2020cites this paper
• and adjustment

  2020cites this paper
• Sampling‐Based versus Design‐Based Uncertainty in Regression Analysis

  2020cites this paper
• The Generalized Oaxaca-Blinder Estimator

  2020cites this paper
• Design-Based Covariate Adjustments in Paired Experiments

  2020influential citation
• Covariate-adjusted Fisher randomization tests for the average treatment effect

  2020cites this paper
• The P-LOOP Estimator: Covariate Adjustment for Paired Experiments.

  2019cites this paper
• Regression-adjusted average treatment effect estimates in stratified randomized experiments

  2019influential citation
• Artificial neural network for predicting values of residuary resistance per unit weight of displacement

  2019cites this paper
• Regression-adjusted average treatment effect estimates in stratified and sequentially randomized experiments

  2019influential citation
• Rerandomization and regression adjustment

  2019cites this paper
• Inference in Experiments With Matched Pairs

  2019cites this paper
• Regression adjustment in randomized experiments with a diverging number of covariates

  2018cites this paper
• Regression adjustment in completely randomized experiments with a diverging number of covariates

  2018cites this paper
• Is My Matched Dataset As-If Randomized, More, Or Less? Unifying the Design and Analysis of Observational Studies

  2018cites this paper
• Randomization Tests for Weak Null Hypotheses in Randomized Experiments

  2018cites this paper
• On mitigating the analytical limitations of finely stratified experiments

  2017cites this paper
• On randomization-based causal inference for matched-pair factorial designs

  2017cites this paper
• Decomposing Treatment Effect Variation

  2016cites this paper