Simultaneous estimation of the mean of a binary variable from a large number of small areas
Original version
Journal of Official Statistics, 19, No. 3 (2003), pp. 253-263Abstract
We develop a frequentist method of simultaneous small area estimation under hierarchical models. The simultaneous estimator is the best ensemble predictor under the model. It is preferable to the area-specific best predictor when the distribution of the small area parameters is of interest as well. We provide details of application to binary data. We illustrate the proposed methodology on register employment and unemployment data, and validate the results by the true population values.
Description
The original publication is available at http://www.jos.nu/Contents/jos_online.asp Artikkelen er også utgitt i Statistisk sentralbyrås serie Særtrykk/Reprints nr 251.