Small area methods

from Wikipedia, the free encyclopedia

Small area methods (abbreviated to SAM) are essentially concerned with the development and improvement of statistical procedures for the estimation of small-scale data in which the regions show only very few or even missing (survey) values. The focus is on regionalization , i.e. the "downscaling" of aggregated data available for larger regions into finer geographic units.

backgrounds

Small area methods originate from the North American region and are also called small area estimates, small area statistics or small area methods. The advantages of SAM compared to conventional estimation models are that information from higher spatial levels flows into the model as well as information from similar regions. This leads to very precise information, even if the regions have only small or no case numbers. A prerequisite for the room levels is a hierarchically non-overlapping structure ('nested data').

It is assumed that certain relationships between variables are not only valid in one area, but are generally present in a similar way in the population. Thus, information from all or at least many regions can help stabilize the estimates in the individual regions. Prof. JNK Rao (2004) speaks of "borrow strength 'from related areas to find indirect estimates that increase the effective sample size and thus increase the precision". So-called auxiliary variables for the respective region provide important additional information, which can significantly improve the model. Possible help information is e.g. B. Characteristic values ​​from the past, values ​​from neighboring or superordinate subpopulations and / or values ​​of auxiliary variables that correlate strongly with the characteristic of interest. It is generally attempted to depict similarities and differences between different areas in a model and then to use this model for predicting population or other characteristic values.

literature

  • Münnich R .; Burgard, PJ; Vogt, M. (2013): Small Area Statistics: Methods and Applications. In: AStA Economic and Social Statistics Archive 6, No. 3/4, pp. 149–191.
  • Pfefferman, D. (2002): Small area estimation - New developments and directions, International Statistical Review, 70, 1, 125-143.
  • Rao, JNK (2004): Small Area Estimation with Applications to Agriculture, Journal of the Indian Society of Agricultural Statistics, Vol 57 (Special Volume), 159-170
  • M. Ghosh, JNK Rao. "Small area estimation: An appraisal", Statistical Science, vol 9, no.1 (1994), 55-76.

Web links

Individual evidence

  1. ^ Rao, JNK (2004): Small Area Estimation with Applications to Agriculture, Journal of the Indian Society of Agricultural Statistics, Vol 57 (Special Volume), 159-170