National Health Interview Survey: research for the 1995-2004 redesign.

Objectives: This report presents a major portion of the research carried out for the 1995-2004 redesign of the National Health Interview Survey. The primary focus of the 1995-2004 NHIS redesign research was to explore sample design options to improve the reliability of NHIS statistics for racial, ethnic, economic, and geographic domains. Another objective was to assess the integration of the sample designs of the NCHS surveys of health care providers and NHIS.

Methods: A number of research tasks were carried out by Westat, Inc., to explore the feasibility and cost of various sample design options for the 1995-2004 National Health Interview Survey redesign. This report provides a detailed description of the research that was carried out.

Results: The research results presented in this report were the primary basis for assessing the technical soundness and costs for a number of design alternatives. The first option, called the alpha option, was developed under the assumption of a 50-percent data collection budget increase. The second option, called the beta option, was developed under the assumption of no change in the data collection budget. The beta option was the design implemented in 1995. Other important research items described in this report include oversampling methods for minority populations and dual-frame sampling methods. The research results in this report also led to additional research, primarily in the area of survey integration. A more detailed description of the 1995-2004 NHIS design will appear in a forthcoming NCHS Series 2 report titled Design and Estimation for the 1995-2004 National Health Interview Survey.