Unbiased Estimation of the Average Treatment Effect in Cluster-Randomized Experiments

2015 ◽  
Vol 6 (1-2) ◽  
Author(s):  
Joel A. Middleton ◽  
Peter M. Aronow
Keyword(s):  
Treatment Effect ◽  
Average Treatment Effect ◽  
Unbiased Estimation ◽  
Random Allocation ◽  
Randomized Experiments ◽  
Average Treatment ◽  
Randomized Field Experiment ◽  
The Difference ◽  
Biased Estimators ◽  
Cluster Randomized

AbstractMany estimators of the average treatment effect, including the difference-in-means, may be biased when clusters of units are allocated to treatment. This bias remains even when the number of units within each cluster grows asymptotically large. In this paper, we propose simple, unbiased, location-invariant, and covariate-adjusted estimators of the average treatment effect in experiments with random allocation of clusters, along with associated variance estimators. We then analyze a cluster-randomized field experiment on voter mobilization in the US, demonstrating that the proposed estimators have precision that is comparable, if not superior, to that of existing, biased estimators of the average treatment effect.

OLS and 2SLS in Randomized and Conditionally Randomized Experiments

2018 ◽  
Vol 238 (3-4) ◽  
pp. 243-293 ◽  
Author(s):  
Jason Ansel ◽  
Han Hong ◽  
and Jessie Li
Keyword(s):  
Treatment Effect ◽  
Average Treatment Effect ◽  
Randomized Experiments ◽  
Adaptive Randomization ◽  
Working Paper ◽  
Average Treatment ◽  
Local Average Treatment Effect ◽  
Effect Parameter ◽  
Multiple Treatments ◽  
Local Average

Abstract We investigate estimation and inference of the (local) average treatment effect parameter when a binary instrumental variable is generated by a randomized or conditionally randomized experiment. Under i.i.d. sampling, we show that adding covariates and their interactions with the instrument will weakly improve estimation precision of the (local) average treatment effect, but the robust OLS (2SLS) standard errors will no longer be valid. We provide an analytic correction that is easy to implement and demonstrate through Monte Carlo simulations and an empirical application the interacted estimator’s efficiency gains over the unadjusted estimator and the uninteracted covariate adjusted estimator. We also generalize our results to covariate adaptive randomization where the treatment assignment is not i.i.d., thus extending the recent contributions of Bugni, F., I.A. Canay, A.M. Shaikh (2017a), Inference Under Covariate-Adaptive Randomization. Working Paper and Bugni, F., I.A. Canay, A.M. Shaikh (2017b), Inference Under Covariate-Adaptive Randomization with Multiple Treatments. Working Paper to allow for the case of non-compliance.

A Class of Unbiased Estimators of the Average Treatment Effect in Randomized Experiments

2013 ◽  
Vol 1 (1) ◽  
pp. 135-154 ◽  
Author(s):  
Peter M. Aronow ◽  
Joel A. Middleton
Keyword(s):  
Treatment Effect ◽  
Average Treatment Effect ◽  
Covariate Adjustment ◽  
Randomized Experiments ◽  
Treatment Assignment ◽  
Unbiased Estimators ◽  
Average Treatment

AbstractWe derive a class of design-based estimators for the average treatment effect that are unbiased whenever the treatment assignment process is known. We generalize these estimators to include unbiased covariate adjustment using any model for outcomes that the analyst chooses. We then provide expressions and conservative estimators for the variance of the proposed estimators.

scholarly journals High-dimensional regression adjustments in randomized experiments

2016 ◽  
Vol 113 (45) ◽  
pp. 12673-12678 ◽  
Author(s):  
Stefan Wager ◽  
Wenfei Du ◽  
Jonathan Taylor ◽  
Robert J. Tibshirani
Keyword(s):  
Treatment Effect ◽  
Average Treatment Effect ◽  
High Dimensional ◽  
Randomized Experiments ◽  
Finite Sample ◽  
Simple Method ◽  
Average Treatment ◽  
Population Average ◽  
High Dimensional Regression ◽  
Effect Estimation

We study the problem of treatment effect estimation in randomized experiments with high-dimensional covariate information and show that essentially any risk-consistent regression adjustment can be used to obtain efficient estimates of the average treatment effect. Our results considerably extend the range of settings where high-dimensional regression adjustments are guaranteed to provide valid inference about the population average treatment effect. We then propose cross-estimation, a simple method for obtaining finite-sample–unbiased treatment effect estimates that leverages high-dimensional regression adjustments. Our method can be used when the regression model is estimated using the lasso, the elastic net, subset selection, etc. Finally, we extend our analysis to allow for adaptive specification search via cross-validation and flexible nonparametric regression adjustments with machine-learning methods such as random forests or neural networks.

scholarly journals Time-to-event comparative effectiveness of NOACs vs VKAs in newly diagnosed non-valvular atrial fibrillation patients.

2021 ◽  
Author(s):  
Blanca Gallego Luxan ◽  
Jie Zhu
Keyword(s):  
Atrial Fibrillation ◽  
Major Bleeding ◽  
Treatment Effect ◽  
Average Treatment Effect ◽  
Ischaemic Event ◽  
Newly Diagnosed ◽  
Time To Event ◽  
Oral Anticoagulant Treatment ◽  
Average Treatment ◽  
The Difference

Objective: To investigate the difference in the time-to-event probabilities of ischaemic events, major bleeding and death of NOAC vs VKAs in newly diagnosed non-valvular atrial fibrillation patients. Design: Retrospective observational cohort study. Setting: UK's Clinical Practice Research Data linked to the Hospital Episode Statistics inpatient and outpatient data, mortality data and the Patient Level Index of Multiple Deprivation. Participants: Patients over 18 years of age, with an initial diagnosis of atrial fibrillation between 1st-Mar-2011 and 31-July-2017, without a record for a valve condition, prosthesis or procedure previous to initial diagnosis, and without a record of oral anticoagulant treatment in the previous year. Intervention: Oral anticoagulant treatment with either vitamin K antagonists (VKAs) or the newer target-specific oral anticoagulants (NOACs). Main Outcome Measures: Ischaemic event, major bleeding event and death from 15 days from initial prescription up to two years follow-up. Statistical Analysis: Treatment effect was defined as the difference in time-to-event probability between NOAC and VKA treatment groups. Treatment and outcomes were modelled using an ensemble of parametric and non-parametric models, and the average and conditional average treatment effects were estimated using one-step Targeted Maximum Likelihood Estimation (TMLE). Heterogeneity of treatment effect was examined using variable importance methods in Bayesian Additive Regression Trees (BART). Results: The average treatment effect of NOAC vs VKA was consistently close to zero across all times, with a temporal average of $0.00[95\%0.00,0.00]$ for ischaemic event, $0.00\%[95\%-0.01,0.01]$ for major bleeding and $0.00[95\%-0.01,0.01]$ for death. Only history of major bleeding was found to influence the distribution of treatment effect for major bleeding, but its impact on the associated conditional average treatment effect was not significant. Conclusions: This study found no statistically significant difference between NOAC and VKA users up to two years of medication use for the prevention of ischaemic events, major bleeding or death.

scholarly journals Regression-adjusted average treatment effect estimates in stratified randomized experiments

Biometrika ◽  
2020 ◽  
Vol 107 (4) ◽  
pp. 935-948
Author(s):  
Hanzhong Liu ◽  
Yuehan Yang
Keyword(s):  
Treatment Effect ◽  
Asymptotic Variance ◽  
Average Treatment Effect ◽  
Randomized Experiments ◽  
Asymptotically Normal ◽  
Variance Estimators ◽  
Average Treatment ◽  
Effect Estimation ◽  
Regression Adjustment ◽  
And Control

Summary Linear regression is often used in the analysis of randomized experiments to improve treatment effect estimation by adjusting for imbalances of covariates in the treatment and control groups. This article proposes a randomization-based inference framework for regression adjustment in stratified randomized experiments. We re-establish, under mild conditions, the finite-population central limit theorem for a stratified experiment, and we prove that both the stratified difference-in-means estimator and the regression-adjusted average treatment effect estimator are consistent and asymptotically normal; the asymptotic variance of the latter is no greater and typically less than that of the former. We also provide conservative variance estimators that can be used to construct large-sample confidence intervals for the average treatment effect.

Clarifying selection bias in cluster randomized trials

2021 ◽  
pp. 174077452110568
Author(s):  
Fan Li ◽  
Zizhong Tian ◽  
Jennifer Bobb ◽  
Georgia Papadogeorgou ◽  
Fan Li
Keyword(s):  
Selection Bias ◽  
Treatment Effect ◽  
Treatment Effects ◽  
Randomized Trials ◽  
Additional Data ◽  
Average Treatment Effect ◽  
Causal Effects ◽  
Cluster Randomized Trials ◽  
Average Treatment ◽  
Cluster Randomized

Background In cluster randomized trials, patients are typically recruited after clusters are randomized, and the recruiters and patients may not be blinded to the assignment. This often leads to differential recruitment and consequently systematic differences in baseline characteristics of the recruited patients between intervention and control arms, inducing post-randomization selection bias. We aim to rigorously define causal estimands in the presence of selection bias. We elucidate the conditions under which standard covariate adjustment methods can validly estimate these estimands. We further discuss the additional data and assumptions necessary for estimating causal effects when such conditions are not met. Methods Adopting the principal stratification framework in causal inference, we clarify there are two average treatment effect (ATE) estimands in cluster randomized trials: one for the overall population and one for the recruited population. We derive analytical formula of the two estimands in terms of principal-stratum-specific causal effects. Furthermore, using simulation studies, we assess the empirical performance of the multivariable regression adjustment method under different data generating processes leading to selection bias. Results When treatment effects are heterogeneous across principal strata, the average treatment effect on the overall population generally differs from the average treatment effect on the recruited population. A naïve intention-to-treat analysis of the recruited sample leads to biased estimates of both average treatment effects. In the presence of post-randomization selection and without additional data on the non-recruited subjects, the average treatment effect on the recruited population is estimable only when the treatment effects are homogeneous between principal strata, and the average treatment effect on the overall population is generally not estimable. The extent to which covariate adjustment can remove selection bias depends on the degree of effect heterogeneity across principal strata. Conclusion There is a need and opportunity to improve the analysis of cluster randomized trials that are subject to post-randomization selection bias. For studies prone to selection bias, it is important to explicitly specify the target population that the causal estimands are defined on and adopt design and estimation strategies accordingly. To draw valid inferences about treatment effects, investigators should (1) assess the possibility of heterogeneous treatment effects, and (2) consider collecting data on covariates that are predictive of the recruitment process, and on the non-recruited population from external sources such as electronic health records.

Sign in / Sign up

Export Citation Format

Share Document