Choices of Frame Construction on the National Children's Study: Impacts on Address Quality and Survey Results

• Activity Number: 17
• Type: Topic Contributed
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Ned English*+ and Colm O'Muircheartaigh and Katie Dekker
• Companies: NORC at the The University of Chicago and NORC at the The University of Chicago and NORC at the The University of Chicago
• Keywords: address-based samples ; area probability ; National Children's Study ; listing ; frame construction

Abstract

The National Children's Study (NCS) is a large and complex longitudinal study, the initial stage of which is the enrolment of women of childbearing age. The NCS data collection depends on a sample of addresses within selected segments. Addresses could originate from a number of sources, depending on the nature of the environment: traditional listings, enhanced listings, or licensed address lists based on the USPS delivery-sequence file (DSF). The literature suggests that each method has different coverage properties in particular situations. At question are the characteristics of housing units, households, and ultimately women and children included using each method. Our analysis compares the characteristics of households that were present on the original DSF vs. those that were added through listing. We do so using questionnaire data from a subset of sites to understand the types of women, babies, and potential health outcomes whose inclusion can be influenced by frame type.