Towards a Best Practice of Modeling Unit of Measure and Related Statistical Metadata

Data and metadata exchange between organizations requires a common language for describing structure and content of statistical data and metadata. The SDMX consortium develops content oriented guidelines (COG) recommending harmonized cross-domain concepts and terminology to increase the efficiency of (meta-) data exchange. A recent challenge is a recommended code list for the unit of measure. Based on examples from SDMX sponsor organizations this paper analyses the diversity of "unit of measure" as used in practice, including potential breakdowns and interdependencies of the respective meta-information that go beyond the current SDMX COG recommendations and possible value domains for the identified components.