Calibration in Missing Data Analysis

• Activity Number: 167
• Type: Topic Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Peisong Han*
• Companies: University of Waterloo
• Keywords: Augmented inverse probability weighting (AIPW) ; Double robustness ; Empirical likelihood ; Calibration ; Missing at random (MAR)

Abstract

Calibration is a technique developed in sampling survey literature. Its application in missing data analysis has attracted considerable research interests recently. We will discuss how calibration, combined with the empirical likelihood method, can lead to many desirable properties when analyzing incomplete data. Especially, the robustness against model misspecification can be significantly improved, resulting in the so-called multiply robust estimators. These estimators are consistent if any one of the postulated parallel parametric models is correctly specified.