Abstract
This paper approaches the problem of small area estimation in the framework of spatially correlated data. We propãose a class of estimators allowing the integration of sample information of a spatial nature. Those estimators are based on linear models with spatially correlated small area effects where the neighbourhood structure is a function of the distance between small areas. Within a Monte Carlo simulation study we analyze the merits of the proposed estimators in comparison to several traditional estimators. We conclude that the proposed estimators can compete in precision with competitive estimators, while allowing significant reductions in bias. Their merits are particularly conspicuous when analyzing their conditional properties.
Original language | English |
---|---|
Pages (from-to) | 155-180 |
Number of pages | 26 |
Journal | REVSTAT: Statistical Journal |
Volume | 9 |
Issue number | 2 |
Publication status | Published - 1 Jun 2011 |
Keywords
- Combined estimator
- Empirical best linear unbiased prediction
- Small area estimation
- Spatial models
- Unit level models