Estimation under mode effects and proxy surveys, accounting for non-ignorable nonresponse
Estimation under mode effects and proxy surveys, accounting for non-ignorable nonresponse
We propose a new, model-based methodology to address two major problems in survey sampling: The first problem is known as mode effects, under which responses of sampled units possibly depend on the mode of response, whether by internet, telephone, personal interview, etc. The second problem is of proxy surveys, whereby sampled units respond not only about themselves but also for other sampled. For example, in many familiar household surveys, one member of the household provides information for all other members, possibly with measurement effects. Ignoring the existence of mode effects and/or possible measurement effects in proxy surveys could result in possible bias in point estimators and subsequent inference. Our approach accounts also for nonignorable nonresponse. We illustrate the proposed methodology by use of simulation experiments and real sample data, with known true population values.
EM algorithm; measurement effects; NMAR nonresponse; probability and nonprobability sampling, selection effects.
253-266
Pfeffermann, Danny
c7fe07a0-9715-42ce-b90b-1d4f2c2c6ffc
Preminger, Arie
19b1dcab-c791-41ff-bcdc-8153d887483e
August 2021
Pfeffermann, Danny
c7fe07a0-9715-42ce-b90b-1d4f2c2c6ffc
Preminger, Arie
19b1dcab-c791-41ff-bcdc-8153d887483e
Pfeffermann, Danny and Preminger, Arie
(2021)
Estimation under mode effects and proxy surveys, accounting for non-ignorable nonresponse.
Sankhya A, 38 (1), .
(doi:10.1007/s13171-020-00229-w).
Abstract
We propose a new, model-based methodology to address two major problems in survey sampling: The first problem is known as mode effects, under which responses of sampled units possibly depend on the mode of response, whether by internet, telephone, personal interview, etc. The second problem is of proxy surveys, whereby sampled units respond not only about themselves but also for other sampled. For example, in many familiar household surveys, one member of the household provides information for all other members, possibly with measurement effects. Ignoring the existence of mode effects and/or possible measurement effects in proxy surveys could result in possible bias in point estimators and subsequent inference. Our approach accounts also for nonignorable nonresponse. We illustrate the proposed methodology by use of simulation experiments and real sample data, with known true population values.
Text
PAPER_SANKHYA A (002)
- Accepted Manuscript
More information
Accepted/In Press date: 1 October 2020
e-pub ahead of print date: 29 June 2021
Published date: August 2021
Additional Information:
Funding Information:
This paper has benefited from valuable comments and suggestions by members of Israel's Public Advisory Council for Statistics and by participants in the 2019 Ottawa Group meeting in Rio de Janeiro. We are particularly grateful to Yoel Finkel, Larisa Fleishman, Can Tongur, and Jan de Haan for their detailed thoughtful comments. The opinions expressed in this paper are solely of the authors and do not necessarily reflect the views of institutions other than the ICBS, with which the authors are affiliated.
Publisher Copyright:
© 2021 International Institute of Forecasters
Keywords:
EM algorithm; measurement effects; NMAR nonresponse; probability and nonprobability sampling, selection effects.
Identifiers
Local EPrints ID: 444916
URI: http://eprints.soton.ac.uk/id/eprint/444916
PURE UUID: c4ccb158-2832-4593-8ab5-3a7402fc7684
Catalogue record
Date deposited: 11 Nov 2020 17:31
Last modified: 17 Mar 2024 06:03
Export record
Altmetrics
Contributors
Author:
Arie Preminger
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