Activity Number:
|
48
- Bayesian Adaptive Survey Designs
|
Type:
|
Invited
|
Date/Time:
|
Sunday, July 30, 2017 : 4:00 PM to 5:50 PM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #322174
|
View Presentation
|
Title:
|
Test of Adaptive Survey Design from a Bayesian Perspective in a Longitudinal Survey
|
Author(s):
|
Peter Lundquist* and Anton Johansson and Gabriele Durrant
|
Companies:
|
Statistics Sweden and Statistics Sweden and University of Southampton
|
Keywords:
|
data collection ;
auxiliary variables ;
survey quality indicators
|
Abstract:
|
It has been demonstrated that it is possible to reduce unproductive calls in CATI-surveys by using information from previous waves in longitudinal surveys. In the Swedish Labour Force Survey have data from both registers and paradata been used to build nonresponse models. The models are used to formulate strategies to support the survey agency in the data collection work with the ambition to reduce costs but at the same time maintain the survey quality. In this paper we will investigate if a Bayesian analysis could be an option in monitoring the data collection. Different strategies based on the previous nonresponse models are tested.
|
Authors who are presenting talks have a * after their name.