Online Program Home
My Program

Abstract Details

Activity Number: 364
Type: Contributed
Date/Time: Tuesday, August 2, 2016 : 10:30 AM to 12:20 PM
Sponsor: Survey Research Methods Section
Abstract #321081 View Presentation
Title: Exploration of Methods for Blending Unconventional Samples with Traditional Probability Samples
Author(s): Jonathan Gellar* and Hanzhi Zhou and Michael Sinclair
Companies: Mathematica Policy Research and Mathematica Policy Research and Mathematica Policy Research
Keywords: Non-probability Sample ; Survey Sampling ; Calibration ; Model-based Estimation ; Composite Estimation

With increasing survey costs, reduced participation rates, and a greater need for timely and domain specific estimates to inform treatment and policy decisions, interest has grown in the use of data obtained from unconventional sampling to supplement or replace traditional survey programs. Given these unconventional data sources may result in unknown levels of bias in their estimates, we explore the use of composite model-based estimation methods to blend data from an unconventional sample with parallel data from a traditional probability-based sampling strategy. We present a number of approaches for blending data from two sources. We then study their utility in a large simulation study based on public use data from the National Health Interview Survey (NHIS). We show that use of the proposed methods enables one to determine the fitness of use of an unconventional data source and to leverage such data appropriately to fulfill program objectives.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association