A Search for Robust Model-Based Small-Area Estimation Methods for the National Ambulatory Medical Care Survey (NAMCS)

• Activity Number: 414
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Vladislav Beresovsky*+ and Donald Malec
• Companies: National Center for Health Statistics and National Center for Health Statistics
• Keywords: direct estimates ; model-assisted estimates ; model-based estimates ; hierarchical logistic model ; small area estimation ; ambulatory care

Abstract

NAMCS has been recently redesigned to allow for reliable direct estimates for larger states and remaining smaller states grouped within Census divisions. When covariates correlated with the outcome variable are available for every unit in the population it is possible to gain efficiency of estimates in small areas by using model-based or model-assisted estimators. On the other hand these methods are sensitive in varying degree to possible misspecification of model assumptions with respect to the superpopulation model. In this simulation study we assess robustness of model-based estimators to possible misspecifications of the estimating models and compare their efficiency to direct and model-assisted estimators.