Guide to Choosing a Data Repository

It can be important to publish your data when you’ve completed your research. Find out in this guide how to choose a repository to publish your data.

Publishing Your Data: When, Why and Where

When you have completed your research, the publisher, funder or your institution may require you to publish a final version of your data. This can be to verify the findings in your article or to enable reuse by other researchers that want to build on your findings. Reasons why you would want to publish your data include: 

1. meeting journal and funder requirements
2. promoting new discoveries and transparency and contributing to scholarly records
3. preserving data for the long term
4. increasing the research impact of your work. 

Publishing in a data repository is a common way to achieve this.

Choosing a Suitable Data Repository

When deciding what data repository to publish in key considerations include:

• security obligations
• publisher requirements
• community conventions
• institutional policy. 

Also consider: 

• the type of data you are publishing
• its format and file size
• its potential reuse and reuse conditions in the form of, say, a licence. 

Other considerations can include: 

• versioning of the data
• storage location of the data
• sensitivity of the data
• access conditions. 

The Digital Curation Centre provides a useful checklist for evaluating data repositories. The following workflow supports decision making when choosing a suitable data repository for publishing your data:

You should first try to select or identify a subject- or domain-specific repository. These repositories are the most findable and reusable to your peers and most relevant to your research. These will provide the best support when it comes to making your data FAIR. See a list of these repositories at re3data. Examples include ADA, Paradisec, Pangea, GenBank and the Crystallography Open Database.

If no subject repository is suitable, you should consider using institutional data repositories. These are reliable long-term homes for your data. Some institutions require you to publish a metadata record in the institutional repository, even if you’ve published in a subject-specific repository. See the repositories of Griffith University and Monash University, where you can submit a metadata record if you’ve published elsewhere.

If no institutional repository is suitable or available either, choose a generalist repository. They tend to have specific rules surrounding their use such as the use of a CC0 waiver to support maximum reuse. They have the ability to create DOIs and can handle various data types. Examples include Dryad and Figshare.

In addition to the type of repository you publish your data in, you might want to consider how trusted the repository is and use the TRUST principles to make your decision.

Next Steps

Your library and/or research office will be able to provide support when you’re deciding when, where and how to publish your research data. The ARDC can also provide guidance and support in relation to data publishing. For more information, contact us.

For additional support for choosing a data repository, visit Nature’s data repository guidance page.