|
Activity Number:
|
82
- Contributed Poster Presentations: Government Statistics Section
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Government Statistics Section
|
|
Abstract #312954
|
|
|
Title:
|
The Pattern of Calibrated Estimates for Demographic and Health Variables in Probability-Based Web Surveys, a Case Study
|
|
Author(s):
|
Bill Cai* and Katherine Irimata and Hee-Choon Shin
|
|
Companies:
|
National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics
|
|
Keywords:
|
RANDS;
NHIS;
NRMSE;
propensity model
|
|
Abstract:
|
The Division of Research and Methodology of the National Center for Health Statistics (NCHS) has been conducting a series of web surveys, referred to as the Research and Development Survey (RANDS). RANDS is based on recruited probability panels and focuses on collecting health-related information. In this study, the National Health Interview Survey (NHIS), which provides “gold standard” estimates for federal and private surveys, served as the benchmark for calibrating estimates from RANDS. The calibration method is based on a propensity score weighting approach which combines information from both NHIS and RANDS. The analytic samples include RANDS round 1 and data from the fourth quarter of 2015 NHIS. Normalized root-mean-square error (NRMSE) is used as a metric to measure the difference between RANDS (and calibrated RANDS) estimates and estimates from NHIS. This study will identify and describe patterns in the calibrated estimates by examining NRMSEs for different demographic and health variables as well as alternative propensity score models.
|
Authors who are presenting talks have a * after their name.
