Pseudo-Populations

This book emphasizes that artificial or pseudo-populations play an important role in statistical surveys from finite universes in two manners: firstly, the concept of pseudo-populations may substantially improve users' understanding of various aspects in the sampling theory and survey methodology; an example of this scenario is the Horvitz-Thompson estimator. Secondly, statistical procedures exist in which pseudo-populations actually have to be generated. An example of such a scenario can be found in simulation studies in the field of survey sampling, where close-to-reality pseudo-populations are generated from known sample and population data to form the basis for the simulation process.

The chapters focus on estimation methods, sampling techniques, nonresponse, questioning designs and statistical disclosure control.

Andreas Quatember is Associate Professor at the Department for Applied Statistics (IFAS) at the Johannes Kepler University JKU in Linz (Austria). He studied Statistics and wrote his doctoral thesis on 'Quota Sampling'. His main research topics are sampling theory and survey methodology. Currently, he is working on the project 'Data quality in statistical surveys'.