Text: Mari Kleemola

Finnish Social Science Data Archive is FAIR

Mari Kleemola

In the past year, there has been talk of the FAIR guiding principles for scientific data management in the field of open science. FAIR is an acronym formed from the adjectives Findable, Accessible, Interoperable and Re-usable. The FSD has promoted all of these principles for nearly 20 years and the oldest of our sister archives (such as the British UKDS) have done so for the past half a century.

Although these principles are not new to us, discussing about them has provided us a good incentive to look at the activities of the FSD from a different perspective. One impetus to doing so was a workshop organized by the OpenAIRE2020 project last November, where the FAIR principles were discussed in the context of various research support services. The FSD was one of the services examined.

Because we offer guidance and instructions on data management in addition to our data curation and dissemination activities, it is appropriate, and fair, to ask: "How FAIR is the FSD?"

My answer is that the FSD is very FAIR. Here is a summary why:

  • Data archived at the FSD are described in detail. Metadata is openly accessible and usable, although there may be restrictions to accessing the data. The datasets are always given persistent identifiers and they can be found through the FSD’s own Data Service Portal, Aila, and through national services, such as Finna and Etsin.
  • Metadata can be accessed by anyone on Aila and in the FSD’s OAI-PMH protocol. Registered users can download data on Aila. Aila makes use of HAKA identity federation, which allows the members of Finnish universities, polytechnics and research institutions to log in using the credentials provided by their institution.
  • Data descriptions at the FSD are created using the international DDI Codebook documentation format. The metadata contain references to other metadata, data, and publications. The data are available in SPSS format, which is commonly used by social scientists.
  • The metadata in DDI Codebook format contain a large amount of information about the contents and authors of the data, data collection methods, and ways to cite the data. Terms and conditions of using the data are clear and included in the metadata. The metadata are licensed under a Creative Commons license.

For the FSD, the most challenging FAIR principle is interoperability. The idea behind the principle of interoperability is that data should be machine accessible, processable and readable. This is not perfectly realized at the FSD, but we think interoperability is achieved as well as is possible and practical.

At the OpenAIRE2020 seminar last November, discussions arose regarding what is required of data, organizations or services for them to be FAIR. No one answer was found – and I don’t think finding one is necessary.

There are already detailed certificates and standards to assess, for instance, the activities of organizations. These include, among others, OAIS, Data Seal of Approval and ISO 16363. The FAIR principles, on the other hand, are catchily named general goals and they work fine as such.

The fact that the FSD’s practices are well-aligned with the FAIR principles did not come as a surprise to us, as we already have the DSA certificate. However, there is always room for improvement and the FAIR principles help us understand our strong points and areas that require improvement. I believe the FAIR principles also make it easy to provoke general discussion about sharing and managing research data and best practices related to them.

More information on the FAIR principles

Creative Commons -license