Application of Respondent-Driven Sampling in Hard-to-Reach Populations
When |
May 06, 2013
from 12:00 PM to 01:00 PM |
---|---|
Where | 2302 School of Public Health Building |
Contact Name | Olivia Carter-Pokras |
Add event to calendar |
vCal iCal |
Abstract
It has been a challenge for researchers to construct scientifically sound samples of hard-to-reach populations (or hidden populations) as no sampling frame available for these populations. Because of this challenge, non-random sampling methods have been frequently used to sample hard-to-reach populations, such as snowball sampling, facility-based sampling, and time-location sampling. Limitations to these non-random methods have been well acknowledged. One relatively new method called respondent-driven sampling (RDS) has been designed to overcome the limitations inherent to the non-random sampling methods. It has been recognized as an innovative and powerful sampling method for sampling hidden populations. RDS combines chain-referral sampling with a mathematical model that weights the sample to compensate for the fact that the sample is collected in a non-random way. The RDS sample is expected to include a broad cross-section of the hidden population. Samples that satisfy this modest goal are considered representative. It has proven to be reliable and powerful in sampling hidden populations. In this seminar, the theory of RDS will be introduced, followed by an empirical example of its application in a social network study.