Propensity score adjustment for nonignorable nonresponse

Abstract

Propensity-score adjustment is a popular technique for handling unit nonresponse in sample surveys. If the response probability depends on the study variable that is subject to missingness, estimating the response probability often relies on additional distributional assumptions about the study variable. Instead of making fully parametric assumptions about the population distribution of the study variable and the response mechanism, we propose a new approach of maximum likelihood estimation that is based on the distributional assumptions of the observed part of the sample. Because the model for the observed part of the sample can be verified from the data, the proposed method is less sensitive to failure of the assumed model of the outcomes. Results from two limited simulation studies are presented to compare the performance of the proposed method with the existing methods. The proposed method is applied to the exit poll data for the nineteenth legislative election in Korea.