|
Abstract:
|
Over the last three decades, Small Area Estimation (SAE) became one of the main topics of research and application in survey sampling. The great interest in this topic is in response to a continuing demand for reliable small area statistics, even for areas or domains with very small samples or no samples. Interest in SAE methods has further enhanced due to the recent trend in many countries to base their censuses on administrative records, supplemented by small area estimates that serve for testing, correcting and supplementing the administrative data. In the present talk I shall overview some of the main developments in SAE in recent years, distinguishing between Bayesian and frequentist approaches, applied to cross-sectional and times series models. The overview will be preceded by a brief discussion of earlier developments, which will serve as a background for the new developments.
|
