Benchmark Assessment of Respondent Driven Sampling Data for Foreign Born Korean Americans

• Activity Number: 227 - Non-Traditional Approaches for Sampling Rare Populations
• Type: Topic Contributed
• Date/Time: Monday, July 31, 2017 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Sooin Lee* and Sunghee Lee and Michael R. Elliott and Tuba Suzer Gurtekin
• Companies: University of Michigan and University of Michigan and University of Michigan and University of Michigan
• Keywords: Respondent Driven Sampling ; RDS ; sampling rare populations

Abstract

Demands for data collection of small and rare population subgroups are on the rise. However, traditional sampling methods may not provide practical solutions for such demands. The present study investigates a new sampling method, known as respondent driven sampling (RDS), applied to an immigrant group that is rare in numbers. The best way to evaluate RDS is empirically comparing estimates from RDS data to those from traditional sample data or population data. Specifically, this study uses an RDS dataset that is collected for non-US-born Koreans in the U.S. and compares its estimates to benchmarks from the American Community Survey (ACS) in order to assess error properties in the estimates. In doing so, we propose new inference approaches for univariate and multivariate statistics that are subject to fewer assumptions than existing approaches, while being more flexible to reflect realities of RDS recruitment processes. This study warrants further research about innovative estimation methods.