Making data FAIR

There has been a tremendous increase in the amount of scholar produced research data and hence, there is an emerging urge to benefit from the published research data at most in the digital era of science as stated by Wilkinson et al.. FAIR principles are guidelines for researchers to make their research data Findable, Accessible, Interoperable, and Reusable with an ultimate goal of making the data available for reusability both by humans and machines as initiated by Force 11.

FINDABLE: The first step to make research data FAIR is to be able to find the data and metadata, e.g. the (meta)data should be uniquely and persistently identifiable through persistent identifiers (PID) such as DOI. Moreover, it is of great importance that data are described with rich metadata.

ACCESSIBLE: It should be possible for humans and machines to gain access to your data, under specific conditions or restrictions where appropriate. The (meta)data should be easily retrievable through PIDs assigned by the repositories.

INTEROPERABLE: Interoperability is the ability of research data to be easily combined with other datasets, applications and workflows by humans as well as computer systems. This can be achieved; i) by using well-known and preferably open file formats (4TU.ResearchData, DANS-Easy) and software whenever it is possible, ii) by using relevant standards for metadata (4TU.ResearchData, DANS-Easy) and iii) by further using community agreed schemas, controlled vocabularies, keywords, thesauri or ontologies where possible.

REUSABLE: Data should be sufficiently well-described with metadata and provenance information (make clear how, why and by whom the data have been created and processed). There should be an accessible data usage license so others know what kinds of reuse are permitted.

Following FAIR principles brings great deal of benefits to the academic community as well as individual researchers, research organizations and funders:

• Achieving maximum potential from research data.
• Achieving maximum impact from research.
• Increasing the visibility and citations of research.
• Improving transparency in research and thus,  enabling reproducibility, replicability and reliability of research.
• Speeding up discoveries and revealing new insights into the research, thus facilitating new research questions to be answered. If all researchers make their data FAIR and publicly available it will save a lot of time and money. FAIR data accelerate research and make it more transparent.
• Staying aligned with international standards and approaches.
• Attracting new partnerships with researchers, business, policy and broader communities.
• Staying up to date with new innovative research approaches and tools.

In order to make research data FAIR, be aware of the following:

• FAIR does not necessarily mean that data need to be open. In the cases where the data cannot be made openly accessible, it should be still possible to make the metadata publicly available.
• Metadata is data describing other datasets, by means of formal labels, such as title, creator, year or publisher. Documentation describes data by means of broader categories, such as collection methodology, structure of and relations between data files, and terms of use. Metadata is crucial to find and manage your data, whereas documentation is needed to understand the context of your data and files. Both metadata and documentation are important for reuse of the data.
• FAIR data need a persistent identifier (PID). Trusted repositories, such as 4TU.ResearchData and DANS (list) will assign a PID when publishing datasets.
• Registering at ORCID  (your personal persistent author identifier) and using this identifier with all your (data) publications.
• Using open, non-proprietary or common file formats will increase accessibility and interoperability.
• Using consistency in your file names, data variables, scripts, scripts variables and throughout similar annotations.
• Attaching the programming scripts you used to analyze or gather your data.
• Providing your data with a clear license to govern the terms of its reuse. Commonly used licenses such as Creative Commons (CC)  can be linked to your data or software.
• Creating a README file (guidance / template) in order to enable that your data can be correctly interpreted and re-analyzed by others.

Your research data are a valuable asset to yourself and to other scientists, society, citizens and entrepreneurs. Therefore, the University of Twente, and a growing number of journals and funders require you to make your data FAIR and if possible open. Your data are ‘FAIR’ if they are easily Findable, Accessible, Interoperable and Re-usable. At the same time, you must comply with privacy legislation. Thus, making your data FAIR can sometimes be complex. In that case, you can ask UT’s FAIR Data Steward for help.