|
Activity Number:
|
89
- SPEED: Survey Methods, Transportation Studies, SocioEconomics, and General Statistical Methods Part 2
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 28, 2019 : 5:05 PM to 5:50 PM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #307931
|
|
|
Title:
|
Statistical Data Integration and Inference via Multilevel Regression and Poststratification
|
|
Author(s):
|
Yajuan Si*
|
|
Companies:
|
University of Michigan
|
|
Keywords:
|
Multilevel Regression and Poststratification;
Data Integration;
Weighting;
Inference
|
|
Abstract:
|
Rapidly decreasing response rates and increasing measurement error rates motivate statistical agencies to put research priorities on combining multiple data sources, such as administrative records and surveys. We develop a unified framework under multilevel regression and poststratification (MRP) for data integration and inferences and handle the methodology issues such as the selection of auxiliary variables, unknown control information, and combination of probability and nonprobability-based surveys. In contrast with the superpopulation and weighting framework, MRP combines prediction and weighting as a hybrid approach. Leveraging the flexible Bayesian paradigm, MRP propagates all sources of uncertainty and stabilizes small area estimation while accounting for sample selection and response mechanisms into modeling. We use simulation studies to evaluate the frequentist properties and compare with alternative methods. The proposal is demonstrated and applied to real-life survey studies.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|
