Nonparametric estimation for longitudinal data with informative missingness
Nonparametric estimation for longitudinal data with informative missingness
Longitudinal data analysis is of great interest in a wide array of disciplines across the medical, economic, and social sciences. In this chapter, the authors propose a new non-parametric estimating equation (NEE) approach to estimation based on longitudinal data subjected to informative missing mechanisms. The NEE approach provides a method for exploring informative non-response in the longitudinal setting, which is computationally easy and flexible in specification. Moreover, the plug-in observed NEE will be somewhat biased if the ‘score-term’ in the population EE is correlated with the response propensity, as in the case of informative non-response. The matter is considered, including possible venues for bias adjustment. The associated variance estimation is described. The authors illustrate and investigate the performance of the NEE approach using a simulation study.
491-512
Ahmad, Zahoor
1a4418f6-8855-4244-a89f-882eadff9721
Zhang, Li Chun
a5d48518-7f71-4ed9-bdcb-6585c2da3649
22 March 2021
Ahmad, Zahoor
1a4418f6-8855-4244-a89f-882eadff9721
Zhang, Li Chun
a5d48518-7f71-4ed9-bdcb-6585c2da3649
Ahmad, Zahoor and Zhang, Li Chun
(2021)
Nonparametric estimation for longitudinal data with informative missingness.
In,
Lynn, Peter
(ed.)
Advances in Longitudinal Survey Methodology.
(Wiley Series in Probability and Statistics)
Wiley, .
(doi:10.1002/9781119376965.ch20).
Record type:
Book Section
Abstract
Longitudinal data analysis is of great interest in a wide array of disciplines across the medical, economic, and social sciences. In this chapter, the authors propose a new non-parametric estimating equation (NEE) approach to estimation based on longitudinal data subjected to informative missing mechanisms. The NEE approach provides a method for exploring informative non-response in the longitudinal setting, which is computationally easy and flexible in specification. Moreover, the plug-in observed NEE will be somewhat biased if the ‘score-term’ in the population EE is correlated with the response propensity, as in the case of informative non-response. The matter is considered, including possible venues for bias adjustment. The associated variance estimation is described. The authors illustrate and investigate the performance of the NEE approach using a simulation study.
This record has no associated files available for download.
More information
Published date: 22 March 2021
Identifiers
Local EPrints ID: 477386
URI: http://eprints.soton.ac.uk/id/eprint/477386
PURE UUID: 483f8a35-3c2c-4a52-b287-34a935bc424e
Catalogue record
Date deposited: 05 Jun 2023 16:58
Last modified: 06 Jun 2024 01:51
Export record
Altmetrics
Contributors
Author:
Zahoor Ahmad
Editor:
Peter Lynn
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
