LR5_3 Area-level small area estimation methods for domain statistics

Document date: 
Tuesday, 29 August, 2017

Paper reviewed:

  • Boonstra, H.J., van den Brakel, J., Buelens, B., Krieg, S. and M. Smeets (2008). Towards small area estimation at Statistics Netherlands. Metron 66 ( 1 ), pp. 21 - 49.

Design-based or direct survey estimates for small subpopulations often have large sampling variances due to small sample sizes. By linking the subpopulation quantities of interest through a model it is usually possible to obtain estimates with smaller mean squared errors. This is the subject of small area estimation (Rao and Molina, 2015). The subpopulations are not restricted to geographical areas, but can be anything as long as they are mutually exclusive, there are not too few of them, and the quantities of interest show some degree of similarity between the areas.

Document data sheet

Related content: