Testing Multiple Modes of Data Collection with Network Sampling with Memory

Duke University, Durham NC

Investigators

Linked publications & trials

Abstract

ABSTRACT: Many immigrants groups are either too small, resulting in a large number of screening interviews to recruit a sufficiently large sample, or are reluctant to respond to conventional surveys for fear of repatriation if they are undocumented, resulting in incomplete or biased samples. This proposed study extends work on an innovative approach, Network Sampling with Memory (NSM), to efficiently and cost-effectively sample from a rare population of immigrants in the U.S. In this proposed project, we will take three important steps designed to make NSM a viable alternative for researchers in the field. First, we will use NSM to collect data on a rare population, Chinese immigrants in a well-defined geographic area (the Raleigh- Durham, N.C. metro area), which will allow a comparison of the results from our network-based sample to population estimates from the American Community Survey. Second, we will experimentally field three different survey modes: telephone, face-to-face, and online modes of data collection to identify the factors that enable NSM to generate large-scale samples with minimum cost and maximum response rates. Third, we will disseminate the computer code necessary to make NSM available for public use, which will facilitate the extension of NSM to additional immigrant groups and other rare or hidden populations, on a larger geographic scale, and promote the use of NSM by other researchers. The network data collected as part of NSM will also allow the analysis of network-related social processes associated with labor market and health outcomes and network-based measures of immigrant social incorporation.