The series of lectures on the estimation for domains and small areas will cover to some extent the following topics: (1) Estimator type, model choice and auxiliary information, (2) Design-based methods of estimation, (3) Horvitz-Thompson, GREG and model calibration estimators, (4) Model-based methods of estimation, (5) Synthetic and EBLUP estimators, (6) Composite estimators, and (7) Computation, Software, SAS Macro EBLUPGREG and Program Domest. The treatment of the materials is practically oriented. The necessary statistical theory will be introduced. Applications will be presented, including the estimation of domain totals and means and their variances. The estimation of selected poverty indicators for domains will be discussed. Properties of the estimators are compared and discussed based on Monte Carlo simulation results. An option will be offered for university students to gain some credit points with a completed practical homework.
|Title of host publication||Proceedings of the Baltic-Nordic-Ukrainian Summer School on Survey Statistics|
|Number of pages||1|
|Place of Publication||Kyiv|
|Publisher||TbiMC Scientific Publishers|
|Publication status||Published - 2009|
|MoE publication type||B3 Article in conference proceedings|
Fields of Science
- 112 Statistics and probability
Lehtonen, R. (2009). Estimation for domains and small areas with design-based and model-based methods. In Proceedings of the Baltic-Nordic-Ukrainian Summer School on Survey Statistics (pp. 7). Kyiv: TbiMC Scientific Publishers.