Activity Number:
|
41
|
Type:
|
Contributed
|
Date/Time:
|
Sunday, August 11, 2002 : 4:00 PM to 5:50 PM
|
Sponsor:
|
Section on Survey Research Methods*
|
Abstract - #300817 |
Title:
|
Combining Surveys to Estimate Cancer Risk Factors in Small Areas
|
Author(s):
|
Dawei Xie*+ and Trivellore Raghunathan and James Lepkowski
|
Affiliation(s):
|
University of Michigan and University of Michigan and University of Michigan
|
Address:
|
2151 Hubbard, Apt. 11, Ann Arbor, Michigan, 48105, US
|
Keywords:
|
small area estimate ; hierarchical Bayesian model ; Gibbs sampling ; Metropolis Hastings algorithm ; nonresponse ; noncoverage
|
Abstract:
|
Cancer surveillance research requires accurate estimates of risk factors at the small area level. The risk factors of interest include lifestyle characteristics and health care utilization. These risk factors are often obtained from surveys such as the National Health Interview Survey (NHIS) or the Behavioral Risk Factors Surveillance Survey (BRFSS). NHIS is a face-to-face survey with nearly complete coverage of the household population, large sample size, and high response rate, while BRFSS is a telephone survey with a larger sample size but lower response rate. Though BRFSS has sufficient sample size for the states, it may have poor coverage properties. A hierarchical Bayesian model is proposed to combine information from these two surveys to obtain small area estimates at the county level. The model incorporates potential differences due to noncoverage and nonresponse. The sample design features are also incorporated in the model. The posterior distributions of the parameters are obtained through MCMC method using Gibbs' sampling and Metropolis Hastings algorithm. Standard Bayesian model checking and sensitivity analysis are applied to the hierarchical Bayesian model.
|