Methods and Results for Small-Area Estimation Using Smoking Data from the 2008 National Health Interview Survey

• Activity Number: 129
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Neung Soo Ha*+ and Van Parsons and Partha Lahiri
• Companies: Centers for Disease Control and Prevention and Centers for Disease Control and Prevention and University of Maryland
• Keywords: National Health Interview Survey ; Small area estimation ; Fay-Herriot method ; Variance estimation ; Hierarchical modeling

Abstract

The National Health Interview Survey (NHIS), conducted by the National Center for Health Statistics (NCHS), is designed to provide reliable design-based estimates for national and four major geographical regions of the United States for a wide range of health related variables. However, state or sub-state level estimates are likely to be unreliable since they are based on small sample sizes. In this paper, we compare the efficiency of different area level models in estimating smoking prevalence for the fifty U.S. states and the District of Columbia using the 2008 NHIS survey data in conjunction with a a number of potentially related auxiliary variables obtained from the Area Resource File (ARF), a large county level database compiled from different U.S. federal agencies. A major portion of this study is devoted to the investigation of various methods to estimate the sampling variances needed to implement an area level model. In our data analysis, the hierarchical Bayes method based on the random sampling variance model appears to be have an edge over other area level models considered in the paper.