Limiting the Risk of Data Disclosure Using Replication Techniques in Variance Estimation

• Activity Number: 66
• Type: Topic Contributed
• Date/Time: Monday, August 12, 2002 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods*
• Author(s): Sylvia Dohrmann*+ and Leyla Mohadjer and Jill Montaquila and Thanh Le and Lester Curtin
• Affiliation(s): Westat, Inc. and Westat, Inc. and Westat, Inc. and Westat, Inc. and National Center for Health Statistics
• Keywords: Disclosure limitation ; replication techniques ; jackknife replicates

Abstract

The National Health and Nutrition Examination Surveys (NHANES) are one of the series of health-related programs conducted by the National Center for Health Statistics (NCHS). A unique feature of these surveys is the collection of health data by means of medical examinations carried out for a nationally representative sample of the U.S. population. Beginning in 1999, NHANES is being implemented as a continuous, annual survey. NHANES 1999-2000 was fielded in 27 locations comprising 26 PSUs that were primarily single counties. The small number of single-county PSUs in the sample, together with geographic data and other characteristics available in the data files, led to concerns about disclosure risks in the release of the NHANES 1999-2000 data file. As a result, alternative approaches for variance estimation (to mask the PSUs) are under consideration using the replication technique. This paper will present the methods used to create three alternative sets of jackknife replicates, the comparison of the standard errors of 70 descriptive statistics, and the results of three logistic regression models under the original and alternative replication designs.