Design and Operational Changes for REACH US

• Activity Number: 151
• Type: Invited
• Date/Time: Monday, August 2, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Rachel Harter*+
• Companies: NORC
• Keywords: address-based sampling ; household survey ; multi-mode ; subpopulations

Abstract

Increasing use of cell telephones, and corresponding decrease in households with landline phones, has led to coverage concerns in RDD surveys. For Racial and Ethnic Approaches to Community Health (REACH) U.S., a set of CDC-sponsored community surveys, NORC interviews minority households, for which the landline coverage issue may be more acute. Recently the REACH U.S. Survey converted to an address-based sampling approach with multi-mode data collection. Applications of existing methodologies were challenging due to limited availability of paradata to guide design and operational planning. In this paper we discuss the address-based design, alternatives considered, and related assumptions regarding response rates, eligibility rates, and other factors affecting sample size calculations and costs. We also discuss design and operational adaptations made as we gained experience.