Skip to main content
Researchdata.se

The FAIR principles

The FAIR principles describe some of the most fundamental guidelines for good data management and open access to reusable research data. FAIR is an acronym for Findable, Accessible, Interoperable, and Reusable.  

Data repositories that adhere to secure and reliable data management principles (e.g., CoreTrustSealOpens in a new tab) are designed in such a way that you automatically comply with many of the FAIR principles by entering the required information into the web form used when registering data in the data repositoryOpens in a new tab.

When sharing data in a repository, you are encouraged to provide comprehensive documentation to facilitate data reuse. You can also assign a licence to your data, clearly specifying the conditions and restrictions for reuse.

Illustrations of the FAIR principles. A magnifying glass symbolizes Findable, a hand clicking represents Accessible, puzzle pieces illustrate Interoperable, and two circular arrows indicate Reusable.

FAIR comprises a total of 15 principles distributed under the four categories: Findable, Accessible, Interoperable, and Reusable.

Read more about the FAIR principles:

Findable

As a first step to comply with the FAIR principles, researchers must be able to find the data, meaning that they must be discoverable through various data indexing systems. Research data are considered findable when:

  • Data and metadata are assigned a unique and persistent identifier;
  • Data are described with rich metadata;
  • Metadata include the persistent identifier to the data they describe;
  • (Meta)data can be found via a searchable web resource.

Repositories have a minimum requirement for mandatory metadata to ensure that data are at least minimally findable. By providing additional metadata beyond the required fields, you can make your data even more findable.

Accessible

In addition to being searchable and findable, data must also be accessible. Research data are considered accessible when:

  • (Meta)data can be retrieved, read, and accessed by their identifier using a standardized communication protocol (such as http or ftp);
  • The communication protocol is open, free, and universally implementable;
  • The protocol allows for creation of different user roles and procedures for user authentication and authorization, and for access control to the data;
  • Metadata remain accessible even when the data are no longer available.

Making a dataset accessible does not necessarily mean that it is openly shared. If data contain, for example, sensitive personal information, a confidentiality assessment must be conducted before others can access the material. However, metadata are not considered sensitive, so even if the dataset cannot be openly shared, you can use metadata to indicate that the data material exists and the conditions for accessing and reusing it.

Interoperable

For research data to be interoperable, the following requirements must be met:

  • (Meta)data are presented with semantic descriptions that are standardized, documented, and accessible;
  • The vocabularies, terminologies, and ontologies used are well-established, controlled, and described in a manner that meets the FAIR criteria;
  • Relationships between different data and metadata are described so that it is possible to understand how they are connected.

The primary responsibility for ensuring that data are interoperable lies with the organization responsible for making the data accessible. However, as a researcher you can contribute by using standards for specifying elements such as dates, time periods, and geographical coordinates, selecting commonly used scientific vocabularies to describe categories, and encoding variables according to recognized standards. Where possible, you should also store data in widely used file formats that are supported by different operating systems and software, or use programs that can export data in such file formats when the project concludes.

Reusable

Finally, data must be reusable. A prerequisite for reuse is that data are described with rich metadata, metadata can be read by both humans and machines, and there is clear contextual information regarding, for example, the scientific purpose of the dataset, the circumstances under which the data were collected or created, and the equipment and software used. The description of data and metadata should make it possible for users to understand the origin and content of the material and determine whether it is suitable for the intended reuse. It must also be clearly stated which conditions, such as licences, apply to the use of data and metadata.

Research data become reusable when:

  • (Meta)data are described with extensive contextual information, allowing users to understand the data and metadata and assess their suitability for reuse;
  • (Meta)data have clear and accessible data usage licences;
  • The provenance, or origin and history, of the (meta)data is described in detail;
  • (Meta)data are structured and documented according to applicable, domain-specific standards and established file formats.