|Date:||Thursday, October 9, 2014|
Hard-to-reach populations are typically not covered by a sampling frame thereby making recruitment a difficult task. Consequently, conventional methods of sampling can lead to unreliable estimators of population quantities. Instead, adaptive (link-tracing) sampling can be used to study such populations as social links of individuals can be exploited to adaptively select units for the sample. An abundance of literature on estimation of population quantities exists for when the population size is known. However, strategies that can be used to study populations when the size is unknown have yet to be adequately explored. In this talk two adaptive sampling-based strategies for estimating the size and attributes of hard-to-reach populations are presented. The first strategy is based on a model-based approach to inference and the second strategy is based on a design-based approach to inference. Simulation results from applying the two strategies to an empirical data set based on a drug-using population at risk for HIV/AIDS is presented.
Statistics seminar: Miroslaw Pawlak: “Nearest Neighbor Estimates for Nonlinear Time Series Systems” — Thursday, November 23 at 2:45 p.m., 301 Biological Sciences.
PIMS lecture: Anthony Bonato — Thursday, November 30 at 4 p.m., Robert Schultz Theatre.
Interdisciplinary seminar: Lena Kourkoutis — Friday, December 1 at 4 p.m., Robert Schultz Theatre.