Can We Avoid Problems with Movers? Some Analytical Issues with National Data Linked with State-Level Data

• Activity Number: 633
• Type: Topic Contributed
• Date/Time: Thursday, August 7, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Yulei He*+ and Eric A. Miller and Dean Judson and Jennifer Parker
• Companies: NCHS/CDC and NCHS and NCHS and NCHS
• Keywords: linked data

Abstract

Although data linkage provides an opportunity to conduct analyses that pull strength from each contributing source, additional complexity can arise due to the incoherence of data features. The CDC's National Center for Health Statistics (NCHS), in collaboration with the Florida Cancer Data System (FCDS) and University of Miami, conducted a pilot linkage between the 1986-2009 National Health Interview Survey (NHIS) and 1981-2010 FCDS. This linked data might be used to examine risk factors and characteristics of Floridians who were diagnosed with cancer, controlling for other demographic, socioeconomic, and health factors. Due to residential mobility in and out of Florida, however, there are analytic issues in constructing an analytic sample, calculating appropriate sampling weights, and accounting for the longitudinal feather of both datasets. We will describe these issues and discuss methods that could be used to address them. This specific example will provide valuable insight into the complexities of analyzing national data linked to state-level data.