COLLABORATIVE RESEARCH: The Data Documentation Initiative (DDI) Version 1.1

Regents Of The University Of Michigan - Ann Arbor, Ann Arbor MI

Investigators

Abstract

This collaborative project will extend, enhance, and promote the Data Documentation Initiative (DDI) specification. The DDI is intended to be a Publicly Available Standard for the content, presentation, transport, and preservation of metadata about data sets in the social and behavioral sciences. Written in XML, the lingua franca of the World Wide Web, Version 1.0 of the DDI specification was published in March 2000. Among the changes and extensions planned for Version 1.1 of the specification are more robust mechanisms for describing aggregate/tabular data and complex file types, such as hierarchical and relational files. In addition, the project will conduct training at meetings of professional associations to familiarize researchers and others with the project and to encourage them to adopt this method for documenting social science datasets. The project also will focus on refining the instructions for employing the standard and providing a set of best practices to demonstrate optimal usage. Finally, project participants will explore a new approach to documentation. This approach likely will have at its core a diagram that defines the relationships among elements. It also will have attributes of a social science codebook. This project has the potential to improve the conduct of social science research in several ways. First, the DDI model provides a comprehensive set of codebook elements designed to incorporate all of the information a potential data analyst must have to conduct analysis accurately and efficiently. The proposed refinements will render the standard applicable to all types of social science data. Training and outreach to the research community will help to ensure that the content standard is widely applied, and best practice activities will promote uniform element usage. Second, the specification is structured in such a way that a single core document can be adapted to produce several other documents, including data definition statements for the major statistical software packages, bibliographic records, study abstracts, etc. The documentation author need only make changes in one place. Third, the project promotes data interoperability by creating a Web-based nonproprietary documentation standard that permits seamless exchange of files among data users and data providers. Finally, the shared specification provides the necessary infrastructure for porting data into online analysis systems, thus putting more social science data on the desktops of a wider community of users.