Considerations in Weighting Survey Data
Sponsor(s): Biostatistics and Bioinformatics
nalysis of survey data often involves weights. Stratified samples may require weighting in order to generate estimates representative of the larger population from which the strata were drawn, and even when all the sample members are selected with equal probability, weighting is often required to compensate for such factors as survey nonresponse. This seminar presents the essential concepts and principles of weighting using illustrations from commonly used survey designs such as simple and stratified random sampling and provides a brief overview of how the concepts discussed can be applied to more complex survey designs and sampling situations.
Type: PANEL/SEMINAR/COLLOQUIUM