Simulation of Alternative Single-Frame and Dual-Frame Sample Designs for the NIS

• Activity Number: 186
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Kirk Wolter and Nadarajasundaram Ganesh*+ and Wei Zeng and Kennon Copeland and Stacie Greby and Meena Khare
• Companies: NORC at the University of Chicago and NORC at the University of Chicago and NORC at the University of Chicago and NORC and Centers for Disease Control and Prevention and NCHS/CDC
• Keywords: dual-frame sample design ; National Immunization Survey ; RDD survey ; single-frame sample design

Abstract

The National Immunization Survey (NIS) is a nationwide random digit dial (RDD) survey monitoring the vaccination coverage among children of ages 19-35 months. Prior to 2011, NIS used a traditional landline telephone sample design. Beginning in 2011, official vaccination coverage was calculated using a dual-frame landline and cell-phone sample design. Recent estimates from the National Center for Health Statistics indicate that the proportion of children living in landline-only households was 4.5 percent, creating the possibility for a viable single-frame cell-phone sample design for the NIS. To assess optimal NIS sample designs in terms of data quality and cost, this paper compares alternative single-frame and dual-frame sample designs via a simulation study. Specifically, we compare vaccination coverage rates, key data collection indicators, and design effects for the following sample designs: (a) current dual-frame sample design, (b) landline telephone sample supplemented by a sample of the cell-phone-only population, (c) cell-phone sample supplemented by a sample of the landline-only population, and (d) single-frame, cell-phone sample design.