Improving Small-Area Health Data: Developing an Outcome Screening Procedure to Support Small-Area Estimation Models

• Activity Number: 422
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Donald Malec*+ and Peter Meyer and Nathaniel Schenker and Pavlina Rumcheva
• Companies: National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics

Abstract

The demand for direct estimates of health outcomes both over small geographic areas and across small population domains often exceeds survey cost constraints. Lately, indirect, model-based, methods for small area and small domain estimation have helped address this issue, especially since their usefulness and limitations are being better understood. Although these indirect methods can help to avoid the need for increases in sample selection costs, they can be labor intensive, requiring careful modeling and selection of covariates on an outcome by outcome basis. The development of a relatively quick and relatively model free procedure to determine which outcomes vary substantially across areas, as a first step toward full-scale small area estimation, is outlined and current progress is documented.