Minimizing Target Population Burden and Confusion by Coordinated Sampling Using Permanent Random Numbers: An Example from Katrina/Rita Temporary Housing Unit Residents
*Darryl V. Creel, RTI International
Keywords: coordinated sampling, minimizing respondent burden, Katrina/Rita Temporary Housing Units
There are multiple data collection efforts related to displaced persons who resided in temporary housing units as a result of hurricanes Katrina and Rita. RTI International has two studies related to this population, the Children’s Health after the Storms and the Katrina/Rita Pilot Registry projects. With RTI having two studies using the same population, there were a couple of concerns: how to minimize confusing about the two RTI studies and how to minimize the possibility that the same sampling unit would be selected to participate in both studies. To address both of these concerns, RTI used coordinated sampling using permanent random numbers (Ohlsson 1995). Consequently, the concerns were in fact advantages. By having both studies, it allowed RTI to systematically minimize the possibility that the two samples would overlap and minimize confusion for which RTI study the unit was sampled. In the future, this type of coordination could be extended to populations expected to be sampled for numerous studies either within a single government agency or by several government agencies.