Activity Number:
|
301
- Weighting Adjustments
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Survey Research Methods Section
|
Abstract #324533
|
|
Title:
|
Evaluation of Nonresponse Weighting Adjustment Methods for a National Survey of People Living with HIV
|
Author(s):
|
Adam Lee and Ronaldo Iachan and Richard Lee Harding and Davia Spado* and Jason Craw and Christopher Johnson and Luke Shouse
|
Companies:
|
ICF and ICF International and ICF and ICF and CDC and CDC and CDC
|
Keywords:
|
Non-response ;
Propensity Model ;
Weighting ;
Weighting class adjustments
|
Abstract:
|
In 2015, the Medical Monitoring Project (MMP) was redesigned to sample individuals diagnosed with HIV in order to provide national and local estimates of various characteristics of persons living with HIV. Weight adjustments for use with this new stratified multistage sample design needed to be developed. We compared weight adjustments generated from a weighting class model with those from a propensity weighting model. Two components of nonresponse were accounted for: non-contact and nonresponse among those contacted. Weight adjustments methods were applied separately to each to allow for differing predictors by nonresponse component. The propensity method has greater flexibility to incorporate continuous variables and more than two variables into the nonresponse adjustment. However, this flexibility led to only minor improvements in model fit. We examine how the magnitude and variance of key estimates differ between methods at each stage of adjustment. Adjusted weights from both methods resulted in similar weighted estimates with small differences in variances.
|
Authors who are presenting talks have a * after their name.