Population surveys are often implemented using a complex sampling design which can include features such as stratification, clustering, multi-stage sampling, and weighting. The statistical analysis of such a survey may yield biased population estimates and incorrect variance estimates if the design is not properly taken into consideration in the analysis.
The purpose of this workshop is to offer participants an overview of complex survey design and how to account for design features when performing statistical analysis. Examples will be presented and practical aspects of performing complex survey analysis such as available software will be discussed.