Inference on the Best Policies with Many Covariates

06/23/2022
by   Waverly Wei, et al.
0

Understanding the impact of the most effective policies or treatments on a response variable of interest is desirable in many empirical works in economics, statistics and other disciplines. Due to the widespread winner's curse phenomenon, conventional statistical inference assuming that the top policies are chosen independent of the random sample may lead to overly optimistic evaluations of the best policies. In recent years, given the increased availability of large datasets, such an issue can be further complicated when researchers include many covariates to estimate the policy or treatment effects in an attempt to control for potential confounders. In this manuscript, to simultaneously address the above-mentioned issues, we propose a resampling-based procedure that not only lifts the winner's curse in evaluating the best policies observed in a random sample, but also is robust to the presence of many covariates. The proposed inference procedure yields accurate point estimates and valid frequentist confidence intervals that achieve the exact nominal level as the sample size goes to infinity for multiple best policy effect sizes. We illustrate the finite-sample performance of our approach through Monte Carlo experiments and two empirical studies, evaluating the most effective policies in charitable giving and the most beneficial group of workers in the National Supported Work program.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/25/2019

Uniformly valid confidence intervals for conditional treatment effects in misspecified high-dimensional models

Eliminating the effect of confounding in observational studies typically...
research
11/29/2022

Double Robust Bayesian Inference on Average Treatment Effects

We study a double robust Bayesian inference procedure on the average tre...
research
11/21/2020

Robust statistical inference for the matched net benefit and the matched win ratio using prioritized composite endpoints

As alternatives to the time-to-first-event analysis of composite endpoin...
research
07/03/2023

Statistical Inference on Multi-armed Bandits with Delayed Feedback

Multi armed bandit (MAB) algorithms have been increasingly used to compl...
research
02/28/2020

Modelling Network Interference with Multi-valued Treatments: the Causal Effect of Immigration Policy on Crime Rates

Policy evaluation studies, which aim to assess the effect of an interven...
research
06/08/2021

Methodological considerations for estimating policy effects in the context of co-occurring policies

Objective. Understanding how best to estimate state-level policy effects...
research
08/17/2018

Fitting Probabilistic Index Models on Large Datasets

Recently, Thas et al. (2012) introduced a new statistical model for the ...

Please sign up or login with your details

Forgot password? Click here to reset