Explain Your Data: Metadata
Making your data understandable and thus improving its reproducibility, findability and reuseability is not only a service to the rest of the research community. It can also add to your scientific merit when your data are being cited. And additionally, a proper documentation is essential for you to understand your own research results months or years later.
Think about what you have produced and used, and write it down: definitions and methods, data formats and units, literature/sources. Add spatial and temporal references to your samples, use explicit variable and file names. In the information scientific context these information are understood as metadata. Metadata are data holding information about other data but not the data themselves. For further reading view the ANDS metadataguide.
Metadata are a way to describe your research data in a standardized way. Because they are the main access point to the data, they are also the key for data access and reuse.
„Standard metadata allows data to be processed, searched, preserved, recombined and reused across many different contexts.” (Ball, A., Chen, S., Greenberg, J., Jeffery, K., Koskela, R., and Perez, C. (2014). Building a disciplinary metadata standards directory. IDCC Practice Paper, 2013, p. 3).
Ideally, you use a standardised metadata terminology to describe your research data, which can be either generic (e.g. Dublin Core) or usage- and discipline-specific (e.g. Space-Time Coordinate Metadata for the Virtual Observatory). A generic research data standard such as Dublin Core, which can be applied to almost every kind of data, has less expressive power than a domain specific standard. In Dublin Core, for example, a distinct creator and a distinct contributor element exists, but within the contributor element it is not possible to make a distinction between editor, illustrator and translator. In other words: Dublin Core does not provide for granular distinctions within semantic elements.
Up-to-date lists of discipline-specific metadata standards can be found on the following sites:
- The Digital Curation Centre (DCC) maintains a Disciplinary Metadata Catalogue (a list of standards alongside with application profiles, use cases, and tools).
- The Metadata Standards Directory Working Group of the Research Data Alliance (RDA) initiated a community driven version of the above mentioned catalogue that provides more functionalities and possibilities for contributions.
The two directories cover the same sets of standards and have agreed to work together.
Click here to view an example of metadata.