Online Program Home
My Program

Abstract Details

Activity Number: 508
Type: Contributed
Date/Time: Wednesday, August 3, 2016 : 8:30 AM to 10:20 AM
Sponsor: Section on Nonparametric Statistics
Abstract #319891
Title: Semiparametric Estimation of Longitudinal Data with Nonignorable Attrition Using Refreshment Samples
Author(s): Jianfei Zheng* and Lan Xue
Companies: Oregon State University and Oregon State University
Keywords: Attrition ; Longitudinal Data ; Kernel Density ; Refreshment Sample ; Non-parametric ; Wave Data

Attrition is one of the major methodological problems in longitudinal studies. It not only reduces the sample size, but also can result in biased estimation and inference. It is crucial to correctly understand the attrition mechanism and appropriately incorporate it into the estimation and inference procedures. Traditional methods, such as the complete case analysis and imputation methods, are designed to deal with missing data under unverifiable assumptions of MCAR and MAR. The purpose of this paper is to identify and estimate attrition parameters under the nonignorable missing assumption utilizing the refreshment sample. In particular, we propose a semiparametric method to estimate the attrition parameters by comparing the marginal density estimator using the model (Hirano et al. 1998) with the one using only the observed information from the refreshment sample. For a bivariate Gaussian model, we also considered a parametric approach by applying the adaptive Gaussian quadrature to overcome the difficulty in the integration. Our simulation has shown that both methods are able to estimate all three attrition parameters with relatively better MSE.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association