FAIR data: what are the pros and cons?

In recent years, there has been a lot of focus on making research data "FAIR". FAIR is the abbreviation for Findable, Accessible, Interoperable and Reusable. What exactly do these terms mean?

• Findable means that other researchers should be able to find data, for example through online sources. As such, the use of a comprehensive set of metadata is crucial.
• Accessible means that others should be able to access the data. Specific conditions or restrictions may be attached to that access, for example when confidential data is involved. Again, metadata plays an important role here.
• Interoperable means that data is compatible with other data so it can be combined and exchanged. One way to do this is to use open formats and software as far as possible.
• Reusable means that research data is ready for future processing, so that new research can efficiently build upon results already acquired.

FAIR data is therefore data that satisfies a framework of conditions. Data can be FAIR to a greater or lesser extent, so it is not a black-and-white matter.

Myths about FAIR

In short, FAIR ensures that research data can be used and reused to the maximum extent possible. This not only improves the efficiency and impact of research, but also has advantages for the researcher themselves: those who make their data FAIR will be cited more often and will better meet the requirements for research funding. A win-win, then!

While the benefits speak for themselves, there are still many myths circulating about FAIR data. Our colleagues at the Danish research and education network DeiC have responded well by collecting a number of these misconceptions, such as:

• "My research data is confidential, so I can't make it FAIR."
• "Why should I be bothered about FAIR data? The most important thing for my scientific career is citations of scientific publications and impact factors."
• "FAIR data is only useful for researchers doing the exact sciences."

Wondering why these statements are wrong? Then download the communication materials on the DeiC website. Each of the cards contains one myth on the front and the correct information on the back. The cards have a Creative Commons license, so you can easily reuse them at your organisation. The website also contains a great deal of interesting documentation (in English) about FAIR data for beginners.

Collaborating across research domains

By making data FAIR, researchers can build upon each other's results and make connections across different scientific disciplines. For example, data from the climate sciences (e.g. global warming statistics) can be linked to data from the social sciences (e.g. migration movements).

FAIR data ≠ open data

It is often assumed that FAIR data is open data anyway. This is not the case: FAIR data can be "closed" with no problems and protected by patents or privacy laws, for example. So even confidential data can be made perfectly FAIR. In short, the principle of "as open as possible, as closed as necessary" applies.

How does Belnet support researchers?

As the Belgian mandated organisation for the European Open Science Cloud (EOSC), Belnet aims to promote open science and the use of FAIR data in the Belgian R&E community. From our mission as a national research and education network (NREN), we have all the assets in-house for supporting the Belgian R&E community in making their research even more efficient, reliable and relevant. We do this by developing specific services and tools for researchers and data stewards, and by facilitating collaboration / knowledge exchange within our community.

Since last year, we have been managing two services specifically for researchers: DMPonline.be and Orfeo. Thanks to DMPonline.be, research institutions can easily write and manage data management plans (DMP). Such a DMP is an essential part of the process of managing the FAIR data life cycle.

In turn, Orfeo is the open science platform for the Federal Scientific Institutions (FWI). This Belnet-managed and hosted database contains some 8,000 publications, providing a wealth of renewable and reusable scientific data.

Belnet is also looking to the future: for example, we are currently surveying our community about interest in Belnet's possible development of a 'PID' (Persistent Identifier) system. A PID is a unique digital identifier given to a dataset so that it can be identified at all times, even with changes in location on the Internet. Thus, assigning a PID contributes greatly to the findability of data – one of the FAIR principles.

The feedback from our educational and research institutions will give us a greater insight into our community's needs and practical expectations.

Collaborating on services for the research community

Do you have a practical project on open science or FAIR data? Then do be aware that Belnet is ideally placed to support you as a (technical) project partner. Through co-creation, we can work with your organisation to put FAIRness into practice and develop new solutions that promote collaboration between different research institutions.

One important thing here is that Belnet represents the broader research community, and thus can also bring parties together so that project solutions will have broad support from the start.

Even after the project phase is over, Belnet can provide you with the necessary support for further hosting the end result if required, and providing the necessary continuity.