In some surveys, for example in developing countries, reliable estimates of summary statistics are needed for small areas with only very small samples or no samples available. The challenge is how to produce reliable estimates and measures of errors for these areas.
Small area estimation (SAE) has become one of the most researched and practised topics in sample surveys, attracting some of the leading theoretical and applied statisticians from all areas of statistics. The small sample sizes do not allow the use of direct estimators that only use the data for the response variable for the particular areas under consideration, and basing the inference on the classical randomisation distribution over repeated sampling is not suitable for predicting the small areas quantities in non-sampled areas. As a result, statistical models that borrow strength across areas and over time have to be used instead.
| Related Projects | Status | Type |
|---|---|---|
| Small area estimation under a two-part random effects model with application to estimation of literacy in developing countries | Active | Other |
| New bootstrap bia corrections with application to estimation of prediction MSE in small area estimation with binary data | Active | Other |