Abstract
A number of basic concepts and ideas concerning metadata in statistical information systems are presented and discussed. Some experiences from earlier developments in the field of statistical metainformation systems are accounted for. An outline of an architecture for future developments of statistical metainformation systems is given. The architecture is based on the principle that the structure of an information system should reflect the structure of the object system that the information system processes information about. Thus a statistical metainformation system should have a structure reflecting the structure of a statistical survey. Two main sources of survey metadata are identified: survey design decisions and the survey process itself. Survey metadata should be captured as close to the sources and as automatically as possible. In practice this means that survey metadata should be captured and organized as an automatical side-effect of other tasks during the planning, operation, and evaluation of the survey.
Get full access to this article
View all access options for this article.