enEnglish
CROS

Overall Design (Theme)

Summary

Design refers to the design of a new survey, to re-design of a survey, and to continuous improvements in a repeated survey. Two core activities in design is to choose methods – e.g. for sampling and estimation, data collection mode(s), contact strategies, and editing – and to allocate resources to the sub-processes in the statistics production. Adjustments of allocations may dominate the work with improvements, whereas the choices frequently are more prominent for new and renewed survey designs. The aim of the design is, in principle, to find some optimum, e.g., maximum quality for a given cost. However, quality is multifaceted and depends on both uses and users, so the task to find an optimal solution has to be further developed and specified. In practice the "optimisation" rather means striving for something good and appropriate based on requests and costs rather than solving complex optimisation problems.

Mostly much of the practical design work is devoted to the accuracy of the statistics. There are further important quality components, e.g., timeliness and coherence. The "optimisation" may include one or more of these components in the search of a solution, often with trade-offs. An alternative – which may be more frequent – is to treat some components, such as timeliness, as constraints. There are further aspects, which may act both as restrictions and support. For instance, standards and common tools have a strong influence on the design. This means, for example, that the sampling and estimation methods may be chosen with regard to the practices and the IT-tools of the statistical office.

The present module gives an overall description of design and provides some general examples. There are a few handbook modules devoted to design, and there are sections within modules about specific design aspects in those topics, for instance editing and estimation. There is a topic on repeated surveys, for which more knowledge and possibilities are available when striving towards optimisation.

 

To read the entire document, please access the pdf file (link under "Related Documents" on the right-hand-side of this page).

 

Your feedback is appreciated. Please send your remarks, suggestions for improvement, etc. to memobust@cbs.nl.