Rec. 16: Broad application of FAIR

[sjDCC]

FAIR should be applied broadly to all objects (including metadata, identifiers, software and DMPs) that are essential to the practice of research, and should inform metrics relating directly to these objects.

• Policies must assert that the FAIR principles should be applied to research data, to metadata, to code, to DMPs and to other relevant digital objects. Stakeholders: Policymakers.

• The FAIR data principles and this Action Plan must be tailored for specific contexts and the precise application nuanced, while respecting the objective of maximising data accessibility and reuse. Stakeholders: Research communities; Data services; Policymakers.

• Guidelines for the implementation of FAIR in relation to research data, to metadata, to code, DMPs and other relevant digital objects should be developed and followed. Stakeholders: Data services; Data stewards; Research communities; Funders.

• Examples and case studies of implementation should be collated so that other organisations can learn from good practice. Stakeholders: Global coordination fora; Research communities.

[ChiaraGabella]

SIB position : SIB Group Leaders are already striving to produce resources that are as user friendly and accessible as possible and that naturally implies striving for following (qualitatively) the principles that FAIR espouses. However, as exposed under Rec. 11, there is certain resistance in formally adopting those metrics as a regular practice because this requires an excessive effort and cost (in time, knowledge and developer resources). We believe that the community needs simple and accessible guidelines.

[WietskeD]

The University of Groningen and the UMCG: We would propose to include the institutions as stakeholders. Institutions play a very important role when it comes to implementing FAIR. The institution is the place where most of the research is performed and these institutions should know what will work and what is an effective way for a broad application. They can help shape the guidelines for implementation. Moreover, institutions will have a big role in implementing these guidelines in their own organisations.

[holubp]

BBMRI-ERIC Position: It is important to understand that research data may come from non-research sources: this is typical for medical research where substantial part of data comes from health care (see also comment to Rec. 14 on incentives). It is important that FAIR [and FAIR-Health] principles are equally applied at source in such cases.

[katerbow]

DFG position: The implementation of the FAIR-principles is part of a large endeavour that has the potential to alter scientific working culture. Ideally, it would be a great success to apply the principles to any aspect of science as this recommendation suggests. Due to the complexity of the task and the long time needed DFG advises not to seek a full application at once and instead to prioritise the different fields of science – without losing sight of the full application. That would help the process of a systematic application and building up trust and acceptance in the scientific communities.

[ScienceEurope]

Science Europe is currently working on the alignment of funders’ policies (core requirements for DMPs and criteria for Trusted Repositories). These policies take all aspects of FAIR data into account and even go beyond those. These policies will be finalised and published by the end of 2018.

[pkdoorn]

Hmm, FAIR data, metadata and software, OK. FAIR DMPs? In principle yes, although for the near future I would be happy that DMPs are required at all… FAIR identifiers? I don’t know. This implies, among other things, that PIDs need to have a PID… Take care not to fall into a circular trap here.

[mromanie]

ESO position Again, in principle broad application of FAIR makes sense, for as long as it doesn't over-constrain the individual projects.

[npch]

SSI position:

We strongly endorse the application of FAIR to other research outputs as was the original intention in the FAIR Guiding Principles.

For this to become adopted, we must avoid language that treats other outputs as subsidiary to (or a subset of) data, but instead recognises them as complementary outputs - similar in many ways, but different in others. We must also be careful of using “data” where unnecessary e.g. FAIR principles rather than FAIR Data principles, FAIR Data Action Plan vs FAIR Action Plan.

Clearly, we particularly advocate for the application of FAIR to software, though we caution that the implementation must take in to account the specific differences between software and data, in particular around inspectability, licensing, and differing interpretations of interoperability and reusability.

[gtoneill]

Fully support the broad application of FAIR principles also to non-data research outputs. It is not clear if the FAIR principles would then need to be tailored to the various ouputs and thus involve different subsets of FAIR principles alongside the FAIR Data principles. An interesting question is what the future relationship will be between FAIR and Open? There is, for example, a clear difference between FAIR Data versus Open Data. Will broad application of FAIR lead to a distinction between FAIR versus Open Science? Should research and all resarch outputs be minimally FAIR but preferably open?

[etothczifra]

DARIAH-ERIC position: In the Arts and Humanities context, the inclusion of GLAM institutions (galleries, libraries, archives and museums) would be inevitable. These institutions are not only data providers in the context of research as ownership of cultural heritage data is inherently shared between the GLAM institutions and the people and cultures that give rise to the objects in question. Therefore harmonization efforts between FAIR and open policies and mandates on the one hand and the legal frameworks, IPR and GDPR in which the use conditions of cultural heritage sources are embedded on the other is a key priority.