Aims and Scope
Modern statistical applications cover topics like poverty measurement and census estimates. Both areas are of immense policitical implication and, hence, need highly accurate data as a thorough basis for possible decision processes. However, classical statistical methods do not yield adequate data since data have to be delivered also in small regions or subclasses of the whole universe. This urges the need for modern methods, based on sophisticated computational statistics methods, the so-called small area methods.
Small Area Estimation (SAE) deals with estimating a parameter of a subset (called small area) of the population from which the sample has been drawn. A direct estimate of a small-area parameter is a design-unbiased estimate calculated just with the sample data coming from the corresponding small area. Thus, the lack of sample data from the target small area affects seriously the accuracy of the direct estimate, and this fact has given rise to the development of new tools for deriving more precise estimates. SAE is a branch of the Statistical Science providing methodology that borrows strength from auxiliary data, from other domains, or from time or spatial correlations.
Modern applications using small area methods suffer from a more and more demanding need of computers, software, and efficient algorithms. The computational burden, e.g. in census applications, is more and more an important task for modern research in small area statistics. This includes efficient and fast algorithms as well as computer efficient programming. An additional important problem to be solved is the evaluation of small area statistics outcomes. This will have to be done by computer simulation studies – facing the difficulty of having two very different strategies, model-based and design-based studies, which have to be thoroughly evaluated under practical and theoretical aspects.
During the last 20 years there has been a great important activity to research and develop new SAE methodologies. The basic monograph of Rao (2003), or the reviews of Ghosh and Rao (1994), Rao (1999), Pfeffermann and Buck (2002) and Jiang and Lahiri (2006) have been establishing a solid theoretical body. European Statistical Institutes have started to apply SAE methods to enlarge their published data. Several European projects (e.g. EURAREA, AMELI and SAMPLE) have been funded by the European Commission to develop software applicable to SAE. A series of European Statistical Conferences (SAE 2005, 07 and 09) is systematically being organized. Because of all these facts, in the present days SAE is an active field of R&D that concentrates the interest of researchers and public statisticans.
This ECAS-Course tries to create a bridge between theoretical body and the practical application with future prospects on recent developments in Small Area Statistics.