Abstract

This presentation deals with the estimation of domain and small area parameters, a standard problem in official statistics, e.g. for resource distribution, monitoring and evaluation.
In this talk we (a) highlight the importance of valid inferential tools for doing comparative statistics in this context, (b) show that classical standard methods like the well-known Bonferroni correction do generally not work, (c) introduce new methods for such simultaneous or joint analysis, and (d) discuss robustness issues that can be tackled by appropriate bootstrap procedures.