Citations & Identifiers

Metadata

It is now understood that data generated in the course of research is as valuable to the ongoing academic discourse as papers and monographs. In many disciplines, the journal article alone is no longer sufficient to assert research results. Many funders, journal publishers, and institutions now require that the data underpinning the journal article be deposited or published, as a "replicating dataset", a dataset that be used again to replicate results asserted in the journal article. Data must therefore be made easily discoverable with proper citations holding persistent identifiers.

Citing data should be done in similar manner to citing papers or monographs, and should hold these elements:

  • author,
  • title,
  • year of publication,
  • publisher (or the name of the archive where the data is stored),
  • edition or version, and
  • access information

When you would like to cite a dataset available on HKU DataHub, you may select your preferred citation style from a drop-down list on the specific dataset record, and directly copy it. A guide with illustrations on how to cite a dataset on DataHub is available here.

Identifiers

Monographs have ISBNs, journals have ISSNs, journal articles have DOIs, and people have ORCIDs, for similar reasons. They disambiguate similar items one from another, and make discovery and identity unique and persistent to that one item or person. In like manner, datasets can be assigned DOIs for unique identity and disambiguous discovery. Additionally, statistics and metrics on usage across the internet can be compiled for datasets citable with a DOI, while usage on those without DOIs is much less discoverable.

When you deposit data into the HKU DataHub, it will automatically register a DOI for your data. You will be able to locate the DOI on your dataset record immediately upon publication. This DOI can then be used in monographs and journals articles to cite this data. It can also be included in research completion reports to funders and institutions, to show that a replicating dataset has been duly deposited.

You may wish to use data repositories external to HKU, which are also able to store your data, provide access, and mint a DOI for it. Some common generalist repositories are listed below. If you do so, you should also deposit metadata in the HKU DataHub, to comply with HKU requirements.