Estimating Population Size with Link-Tracing Sampling

• Activity Number: 590
• Type: Topic Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Kyle Vincent*+
• Companies: Bank of Canada
• Keywords: Adaptive sampling ; Link-tracing sampling ; Mark-recapture ; Rao-Blackwellization

Abstract

There has been a growing interest in the study of hard-to-reach populations like those comprised of injection drug-users and the homeless. New link-tracing sampling designs and inferential methods for measuring attributes of such networked populations have therefore received an abundance of attention recently. In this talk I will present a novel strategy for estimating the size of networked populations when samples are selected via a link-tracing design. Preliminary estimates of the population size are obtained with the use of common mark-recapture estimation methods and are based on the conventionally selected members of the samples. A Rao-Blackwellization strategy is then implemented by incorporating the adaptively selected units into the analysis and will in turn achieve an improved estimator. Results from an empirical study of a networked population at risk for HIV/AIDS will be presented to demonstrate the efficiency of the new strategy.