520 – Contributed Oral Poster Presentations: Survey Research Methods Section
Designing Flexibility for State Samples into the Redesigned 2016 National Health Interview Survey (NHIS)
Van Parsons
National Center for Health Statistics
The NHIS is a multi-purpose face-to-face health survey conducted annually by the National Center for Health Statistics (NCHS). Historically, the NHIS has been designed to produce accurate statistics at the national level. The traditional design has been based upon a one-time static sample of coarsely defined geographical units, typically a probability sample of counties and block clusters, to define the areas for annual household sampling conducted over a 10 year NHIS design cycle. Recently, there has been a greater interest in structuring geographical flexibility into the NHIS. In particular, with recent attention given to state-level health measures,the NCHS wants the NHIS to have the ability to expand and contract the sample in targeted geographical areas in order to meet both estimation needs and budgetary conditions. Any flexible design structure must be cost-effective and implementable in a timely fashion with respect to field operations, and be amenable to creating the design features needed for data analysis for both annual and multiple years of data. In this work a proposed systematic sampling method which meets many of the goals of a flexible design is discussed.