Nonsampling Error in a Survey of Department of Defense Health Care Beneficiaries

• Activity Number: 73
• Type: Contributed
• Date/Time: Monday, August 12, 2002 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods*
• Author(s): Nancy Clusen*+ and Eric Schone and Dan Kasprzyk and Thomas Williams
• Affiliation(s): Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc. and Mathematica Policy Research, Inc. and TRICARE Management Activity
• Keywords: Nonsampling error ; nonresponse ; mode of data collection

Abstract

The Health Care Survey of Department of Defense (DoD) Beneficiaries (HCSDB) is the primary tool used by DoD to monitor satisfaction with, access to, and use of care in the military health system. We review three sources of nonsampling error in the HCSDB: errors associated with the sampling frame, nonresponse (including unit and item nonresponse), and measurement issues associated with the mode of data collection. The HCSDB, conducted in 2000, included an experiment in which nonrespondents were contacted by telephone in an effort to understand, identify, and quantify the bias due to nonresponse. The results of this experiment and other research provide some insight into nonsampling error in the survey. This paper describes the errors, their cause, and their impact on survey estimates.