Small area estimation is a research area in official and survey statistics of great practical relevance for National Statistical Institutes and related organisations. Despite rapid developments in methodology and software, researchers and users would benefit from having practical guidelines that assist the process of small area estimation. In this paper we propose a general framework for the production of small area statistics that is based on three broadly defined stages namely, Specification, Analysis/Adaptation and Evaluation. The corner stone of the proposed framework is the principle of parsimony. Emphasis is given on the interaction between a user and a methodologist for specifying the target geography and parameters in light of the available data. Model-free and model-dependent methods are described with focus on model selection and testing, model diagnostics and adaptations e.g. use of data transformations. The use of uncertainty measures and model and design-based simulations for method evaluation are also at the centre of the paper. We illustrate each stage of the process both theoretically and by using real data for estimating a simple and complex (non-linear) indicators.
