Semiparametric response model with nonignorable nonresponse

TL;DR

This paper introduces a semiparametric response model for handling nonignorable nonresponse in missing data analysis, proposing efficient estimators and validating their performance through simulations and real data application.

Contribution

It develops a semiparametric response model that relaxes parametric assumptions and proposes two efficient estimators with proven asymptotic properties.

Findings

Proposed estimators outperform existing methods in simulations

Method effectively handles nonignorable nonresponse

Application demonstrates practical utility on survey data

Abstract

How to deal with nonignorable response is often a challenging problem encountered in statistical analysis with missing data. Parametric model assumption for the response mechanism is often made and there is no way to validate the model assumption with missing data. We consider a semiparametric response model that relaxes the parametric model assumption in the response mechanism. Two types of efficient estimators, profile maximum likelihood estimator and profile calibration estimator, are proposed and their asymptotic properties are investigated. Two extensive simulation studies are used to compare with some existing methods. We present an application of our method using Korean Labor and Income Panel Survey data.