The Role of Metadata in Census Data Dissemination

1. The Role of Metadata in Census Data Dissemination United Nations Regional Seminar on Census Data Dissemination and Spatial Analysis Bangkok, Thailand 5-8 October 2010 United Nations Statistics Division

2. Outline of presentation • What is metadata? • Basic roles of metadata • Types of metadata • Users of metadata • Metadata in census data dissemination • Metadata standards • Conclusion

3. What is metadata? • It is descriptive and structured information or documentation about data which: • informs users about the content, quality and condition of data; • guides on proper usage or interpretation of data; • describes the structure of datasets, explains, locates, or makes it easier to retrieve, use, or manage data; • provides information on the processes of production. • “Data that define and describe other data” (ISO definition) • “Information about information” (Dion, 2006)

4. Basic Roles of Metadata • Assist in retrieving and processing data • Support correct use of data • Provide transparency in data • Enhance interpretability or confidence of users in the information • Improve archiving, preservation, institutional memory over time

5. Types of metadata • Structural metadata • provide information about the structure of the dataset • act as identifiers and descriptors of the data, making it possible to properly identify, retrieve and browse the data • Reference metadata • allow a thorough understanding and interpretation of the corresponding statistical data • Describe the concepts, definitions, methodology and quality of data; production and dissemination process, data access conditions, release policy, confidentiality, etc

6. Users of metadata Metadata support the knowledge of potential user of statistical information. The major users include: • Users of statistical data • Producers of statistical data • Researchers on the development of statistical systems

7. Metadata in Census Data Dissemination All tabulations of census data should include: • Census questions • Reason why the questions are asked • Conceptual definitions • Geographic hierarchies used • Changes since the previous census, regarding content, operational methods or geographic boundaries • Quality indicators (ex. coverage rates and item non-response) • Methodological note on the rules and methods of suppression If a long-form sample is used... • Sampling variability of the results (C&P, 2008)

8. Metadata standards • Currently, metadata suffer from a lack of　common standards and guidelines • Standardization: • ensures consistency and comparability of content • avoids duplication and diversity of definitions • ensures reduction in cost of data development

9. Metadata standards (cont’d) • Two international metadata standards are becoming well established • SDMX • DDI • a number of international agencies have endorsed SDMX; supported by the UN Statistical Commission • IHSN’s Microdata Management Toolkit uses the DDI metadata standard

10. Conclusion • Metadata are at the heart of the information management. • They should be an integral part of statistical dissemination strategies. • Standards for metadata management are important • Regional and international collaboration between NSOs is an important consideration

11. Points for discussion - What are the success factors of metadata? Strategies? Infrastrucure? Resources? - How to ensure the security of metadata? - How “open” or “transparent” the metadata can be? - How to improve standerdization of metadata?

12. Thank you !