How to Publish Data?
You would like to publish some data? But where and how? On the different places and options and why Google Drive might just not be enough.
"As open as possible, as closed as necessary" is the motto for research data publication. We strongly encourage you to share as much of your data as you can, while we also acknowlegde that there is data you just cannot publish, or at least not yet. For any of these situations, there is a solution.
Open or closed access, the first place for storing your most important data sets is a repository.
Why? You might say: I have a Google Drive for my project, or GitHub, what do I need a repository for?
The Advantages of a Repository over GDrive and GitHub for Publishing:
Long-term Preservation
Repositories provide long-term preservation and archiving, ensuring that yor results remain accessible for future generation. Or only future you.
The point is, you do not need to worry about the long-term preservation on your GDrive, the professionals at the repository do that for you.
Persistence
Data sets on a good repository receive a persitent identifier such as DOIs. These identifiers stay the same and never change, so the link cannot break. This makes them a stable identifier for your files not only throughout your project, but throughout your career.
It also means, your data is findable for evaluation of your research, you will be able to provide your data to prove the integrity of your work.
(The same is true for ORCID, by the way.)
Increased Visibility
Repositories are out there to be seen. With a elaborate metadata management, a repository (especially the big ones like Zenodo) can give your research a lot more visibility than even a project website can.
If you publish your data with an open licence (such as CC-BY), this effect is increased with other researchers being able to see your work more easily.
Even if you only upload your data set with an embargo or with restriced access, the meta data can still be made findable, which makes your work more visible in the community.
Metrics
Repositories often provide metrics on the usage of your data like visits, downloads and citations. It is nice to be able to see the impact of your work – and to show it in your next grant application, isn't it?
Compliance with Funder Mandates
Some research funders like the EU have made it mandatory to publish openly on a repository. GitHub or your project website will not fulfill these funder requirements.
Legal Security
When publishing your data, you want to be sure that other people understand, what you want to permit them to do with your work and what not, right?
When you choose licences for your data sets, how do you make sure they are with your data sets when you share a link to your GDrive? In a repository, the licence is linked with your data set. So you know that everybody has seen the licence and that they can be held responsible for using it legally.
Costs
Up to 1 TB, storage in a repository usually is for free, especially in those run by a public institution (a university or CERN).
Credibility
Using an established repository makes you look more professional and gives your work more credibility than sharing a GDrive link would. That can be advantageous for your next article submission or grant applications with funders.
Comprehensible Versioning
You would like or need to make changes in your data sets and are afraid of the mess you might create with different versions? That can be avoided in repositories. Different repos handle this differently, but in a good repository, you will be able to upload subsequent versions of your original data set. The different versions will be shown under the same DOI.
If you do not want or cannot publish your data with an open licence, have you ever considered uploading the metadata only?
Not only is this a funder requirement under Horizon 2020 and Horizon Europe, it also makes your work more visible, too.
Another otion would be to publish your data with an embargo. After you have done your analyses, it can be open for others.
How to find a good repository?
You can find good repositories on re3data, a kind of white list for data repositories, where you can search for field specific repositories. (The equivalent for text repositories would be OpenDOAR).
If there is no field specific one, we strongly recommend Zenodo, Europe's biggest repository for all research fields and data types, hosted at CERN.
There is also the University of Konstanz' institutional data repository KonDATA.
If you have any questions on publishing data, please do not hesitate to contact your Research Data Managers.
We are happy to help.