Bayesian Small-Area Unit-Level Modeling: A Discussion of Viable Approaches

• Activity Number: 100
• Type: Roundtables
• Date/Time: Monday, August 10, 2015 : 7:00 AM to 8:15 AM
• Sponsor: Survey Research Methods Section
• Author(s): Donald Malec*
• Companies: National Center for Health Statistics

Abstract

The goal of full Bayesian methods for modeling survey outcomes at the unit level is to provide more precise estimates with an accurate assessment of uncertainty. This should be accomplished without resorting to asymptotic assumptions based on small sample sizes. The way to achieve this is by using all available, relevant data and carefully model the uncertainty within both the population model and due to sample selection. The difficulties of this approach, in part, are due to the need to model many details not directly relevant to the outcome of interest, lack of data for model checking, and missing information. The purpose of this discussion is to bring together the experiences of researchers who are working in this area to share approaches---both the successes and the failures.