FAIR Foundations
Conventional research often hides the protocols, datasets, and supporting materials produced along the way. FAIR workflows make those outputs Findable, Accessible, Interoperable, and Reusable — turning your whole research process into something others can discover, trust, and build on.
Why bother making your work FAIR?
There are three big payoffs, and they reinforce each other throughout a project:
Collaboration
Science thrives on collaboration. Sharing your work, process, and results — with enough context to make them FAIR — is the first step toward successful partnerships, even when you start out working alone.
Crediting
Every stage of research deserves recognition. Showcasing the work behind the paper, and attributing it to the people and organizations involved, lets you acknowledge contributions and demonstrate impact.
Compliance
Planning each stage with FAIR in mind helps you meet the policies of institutions, funders, and collaborators — streamlining data sharing, reporting, and preservation later.
The Four FAIR Principles
FAIR is about how a research output behaves once you share it. Each letter describes a property you design into the output and its metadata.
Findable
The output exists as a scholarly record that can be indexed and searched. The most comprehensive index of scholarly works is built from DOI metadata.
Accessible
There is a clear pathway to obtain the work — typically by depositing it in a repository or platform that registers a DOI for the completed output.
Interoperable
The work is described with standardized metadata and meaningfully linked to related works, so it can be incorporated into different settings and systems.
Reusable
Enough context is provided — licensing, provenance, and community standards in the metadata — that others can understand and responsibly reuse it.
Further reading
The FAIR Principles were formally published in Scientific Data (2016): Wilkinson, M.D., et al. The FAIR Guiding Principles for scientific data management and stewardship.
The Engine: PIDs & Metadata
Persistent identifiers (PIDs) are central to all four principles. A PID is a permanent, unique reference for a thing — an output, a person, an organization — together with structured metadata describing it.
The DOI and metadata created when you submit and share an output is what actually makes it FAIR: the metadata describes the work, and links it to related people, organizations, and funding through their PIDs. This shared infrastructure is what lets diverse scholarly works be discovered and cited globally.
Three PIDs you will use constantly
Outputs
Identifiers for most research outputs — datasets, code, protocols, preprints, papers, posters, DMPs.
People
Uniquely identify researchers and contributors, so credit follows the right person across every output.
Organizations
Identify the institutions that host or fund research, so affiliations and funding are unambiguous.
The one idea to hold onto
FAIRness is not a final publishing step — it is created output by output, throughout the project, by registering PIDs and describing each output with rich, linked metadata.
Check Your Understanding
Four questions. Each may have more than one correct answer — select all that apply.
Stage 1 — Project Planning
The best time to make research FAIR is before any data is collected. Project planning creates the reference points — project pages and data management plans — that every later output will link back to.
Why: Create points of reference through which all relevant outputs of the project can be linked and discovered, and establish a shared understanding of the FAIR practices the team will follow.
What you can share: a project page (e.g. the Open Science Framework) and a data management plan (e.g. DMP Tool, ARGOS).
How to keep it FAIR
Use tools and platforms integrated with open infrastructure to create and share outputs. Register DOIs for each output, and use PIDs to reference the related people, organizations, and resources in the metadata.
Project planning is the moment to list your resources — researchers, facilities, funding — and capture their PIDs so they can be credited consistently:
- Make sure every researcher has an ORCID iD and every organization a ROR iD.
- Register a DOI for the project and a DMP ID for the data management plan.
- Confirm the funding identifier with the project funder, and link it to the DMP and project.
- If the project extends earlier work or reuses datasets, reference and cite those in the metadata and DMP.
Stage 2 — Method Planning
Methodological decisions yield key reference outputs for the whole team — above all, a shared, citable research plan.
Why: Increase credibility and create a guardrail against bad practice, while boosting reproducibility and reuse through domain-specific metadata standards.
What you can share: a preregistration / study registration (e.g. the Open Science Framework) and metadata templates (e.g. CEDAR Workbench). For any method involving hypothesis testing, the best practice is to write a detailed research plan and share it in a registry so it is accessible and citable.
How to keep it FAIR
Everything from project planning still applies — and when you register a new output, reference the project and/or DMP in its DOI metadata. Acknowledge the method's creators, contributors, and supervisors through their ORCID iDs.
- Register a DOI for the study registration / experiment plan detailing data collection, treatment, and analysis.
- Register a DOI for any new tools, standards, or procedures that support the methods.
- Reference the umbrella project DOI in the study registration metadata.
- If you reuse existing protocols, or the registration describes analysis of existing datasets, cite them.
- If you create multiple registrations, link them with appropriate relation-type metadata.
The thread that connects everything
Each new output references the ones before it: the project DOI anchors the DMP; the DMP and project anchor the preregistration. By the time you collect data, there is already a linked web of context — interoperability in action.
Check Your Understanding
Four questions. Each may have more than one correct answer — select all that apply.
Stage 3 — Experiment & Analysis
Once data starts flowing, FAIR workflows means registering and linking the real products of research — datasets, protocols, code, and results.
Why: Follow up on the registered activities and mark any deviations, create robust records that supplement your outcomes, and credit everyone who contributed.
What you can share: experimental protocols (e.g. protocols.io), analysis code and software (e.g. Zenodo), and datasets (e.g. Dryad).
How to keep it FAIR
All earlier practices apply — and when you register a new output, reference the corresponding registration in its DOI metadata, and create references among the related data, code, and protocol.
- Register a DOI for each dataset produced in response to the preregistration, and when applicable, a DOI for the experimental protocol.
- Acknowledge contributors on the specific outputs they worked on, using ORCID iDs and appropriate contributor roles.
- Acknowledge the instrument or facility used, in the protocol metadata.
- Link the dataset, code, and results to their preregistration, to each other, and back to the DMP.
- If you adapt an existing protocol or reuse someone's software, cite the original in your derived output.
Stage 4 — Dissemination
Engaging the community starts well before the manuscript is published — and every shared output should carry its own identifier and credit.
Why: Expedite the spread of project outputs, invite community engagement and feedback, and lower barriers to access.
What you can share: preprints (e.g. PsyArXiv, bioRxiv), presentations and posters (e.g. Zenodo, Figshare), and papers (e.g. via ChronosHub).
How to keep it FAIR
Everything above still applies — and now you cite your shared outputs by their DOIs inside manuscripts and presentations, closing the loop between the work and the record of it.
- Make key communication materials (poster, slides, recording) publicly accessible with a DOI.
- Register a DOI for the preprint by depositing to a preprint server.
- Acknowledge authors, presenters, and contributors via ORCID iDs and contributor roles.
- Acknowledge the grants in the preprint and publication DOI metadata.
- Link the preprint to the project/registrations it reports on, and to the manuscript once submitted.
- Cite the key datasets and software in the preprint and the formal publication.
Putting it together
Across all four stages the move is the same: register a PID for each output, describe it with rich metadata, and link it to the people, organizations, funding, and outputs around it. Do that consistently and your research becomes FAIR by design.
Check Your Understanding
Four questions. Each may have more than one correct answer — select all that apply.