Surveying Rare or Hidden Populations Using a Probability-Based Household Panel
Vicki Pineau
NORC at the University of Chicago
J. Michael Dennis
NORC at the University of Chicago, Palo Alto, CA
Stuart Michaels
NORC at the University of Chicago
Sherry Emery
NORC at the University of Chicago
Nada Ganesh
NORC at the University of Chicago
Estimating economic, health, and social disparities among priority subpopulations (e.g., sexual and gender minorities) is increasingly regarded as essential for policy making and scientific inquiry but is problematic without resorting to non-probability sampling. Traditional probability-based sampling strategies are impractically expensive because of the large scale in-field screening required to find sufficient numbers of persons in so-called "rare" or "hidden" populations. As a result, rare or hidden populations are often studied using less rigorous methods such as "snow-ball" sampling and non-probability, opt-in web panels. Our paper is based on an NORC pilot study which tested a cost-effective alternative for surveys of rare or hidden populations. The tested approach combines probability sampling and Network and Respondent-Driven Sampling (RDS). Pilot study findings are presented with respect to assessing the feasibility of using a probability-based panel sample, with multiple rounds of nominations of non-AmeriSpeak panelists by AmeriSpeak panelists, to survey a larger sample of people who self-identify as lesbian, gay, bisexual, and transgender (LGBT) Americans.