Data citation

Data citation is important because it:

acknowledges the author's sources
makes identifying data easier
promotes the reproduction of research results
makes it easier to find data
allows the impact of data to be tracked
provides a structure that recognises and rewards data creators.

It is good practice to cite any existing datasets you use. If you use data from a repository that has been released under an open license, you are obliged to cite it (even under a CC0 license) and should do so with a full citation. It is also good practice to cite datasets published in a data journal with a full citation.

By citing the data paper, you also reward the author for sharing their data, as these citations can be tracked in the same way as for any scholarly paper. You should therefore include a reference to the data paper describing the data, followed by a reference to the data in the repository itself. In order for this to work, it is essential that the citations are in the references section of the article and include the DOI (or any other identifier the repository might use).

Our guidance on data access statements moved to a new section, Data access statements, and was updated to include more examples.

Although there might be differing requirements for data citation across disciplines, a general recommendation for a data citation is:

Creators, Publication year, Title. Publisher. Identifier

It may also be useful to include optional properties, version and resource type:

Creator (Publication year). Title. Version. Publisher. Resource type. Identifier

Similar to publications, datasets should be cited using persistent identifiers such as digital object identifiers (DOIs), which, unlike standard web links, allow permanent linkage to the digital object (dataset). DOIs are coupled with metadata, which can be modified over time to keep track of the locations and characteristics of the objects they identify. This makes it easier to keep track of the objects and information about them in the long term and to easily automate processes such as sharing this information with other electronic systems, e.g. databases, websites etc.

It is recommended that DOIs are displayed as linkable URLs in the format 'https://doi.org/10...'. Below are typical examples of data citations:

Bloggs, J; Brander, S. (2014); Chemical and mineral compositions of sediments. Version SA9992c. UK Data Archive [Computer file]. https://doi.org/10.5255/UKDA-SN-6614-2

Wahl, P. et al. Discovery of a strain-stabilized smectic electronic order in LiFeAs (dataset). University of St Andrews Research Portal. https://doi.org/10.17630/c47ff360-09d4-4620-9ebf-bbb8792fb808 (2018)

Journals might also have specific requirements regarding the citation of secondary data sources as well as datasets underpinning the publication. Authors should therefore also always check journals' citation guidelines.

Data citation

How to cite data