Demystifying Double Robustness

By Joseph D. Y. Kang et al
Published on April 18, 2008
Read the original document by opening this link in a new tab.

Table of Contents

1. INTRODUCTION
1.1 Purpose
1.2 Description of the Problem
1.3 Assumptions
1.4 A Simulated Example
2. WEIGHTING, STRATIFICATION AND REGRESSION
2.1 Inverse-Propensity Weighting

Summary

This paper discusses the concept of doubly robust (DR) procedures designed to address selection bias in statistical analysis. It compares various estimation strategies for population mean in the presence of incomplete data, emphasizing the importance of modeling relationships between covariates and outcomes. The study demonstrates that DR estimates can offer consistent results even when one of the underlying models is misspecified. The example presented illustrates the challenges and potential benefits of using dual-modeling strategies in statistical estimation.
×
This is where the content will go.