FAIR principles for research data

The FAIR principles (Findable, Accessible, Interoperable and Reusable) are designed to make research data more discoverable by humans and machines, and to promote wider sharing and reuse. They can be applied to digital data and artefacts from any discipline.

Working towards making research data more FAIR:

• Increases the visibility and citations of your research
• Improves the reproducibility and reliability of your research
• Enables new innovative research approaches and tools
• Aligns with international standards and approaches
• Supports sharing with collaborators or with the wider world

This webinar provides an overview of the FAIR principles: their origins, Australian FAIR initiatives, what FAIR is (and what it is not), and the four findable principles that underpin the discoverability of data and resources, supporting institutional awareness and uptake of these principles to make your institutional data globally discoverable.

Read the transcript.

How to make data findable:

• Publish the data (or a metadata-only descriptive record) with as much research metadata as possible.
• Use a data publishing platform (e.g. the University's Institutional Figshare) to receive a Digital Object Identifier (DOI).
• Use this persistent link in your publications to help others find and cite your work.

How to make data accessible:

• Publish on a platform that allows direct internet downloads and supports API access. This enables both humans and machines (like search engines and databases) to easily harvest your data.
• If data cannot be public, publish a metadata-only record. Clearly document the specific legal or ethical conditions for access and include instructions on how others can request it. You can then control who can access the data and for what purpose through appropriate processes, including data-sharing agreements.

A for Accessible webinar

This ARDC webinar recording includes an overview of the accessible principles that underpin the access and reuse of data, as well as resources to support institutional awareness and uptake of these principles.

Read the transcript.

How to make data interoperable:

• Save data in non-proprietary formats (e.g. .csv instead of .xlsx, .json instead of .mat) to ensure it can be opened easily.
• Use standardised terms and controlled vocabularies specific to your field (e.g. MeSH for medicine, Darwin Core for biology) to describe your variables.
• Link your dataset (DOI) to your profile (ORCID), your funding (Grant ID), and related publications (DOIs).

I for Interoperable webinar

This ARDC webinar recording provides an overview of the three Interoperable principles, which use vocabularies for knowledge representation, standardisation and references other metadata, as well as the resources to support institutional awareness and uptake of interoperable principles.

Read the transcript.

How to ensure data is reusable:

• Use a machine-readable licence like Creative Commons (e.g. CC BY 4.0). This tells others exactly how they can use your data without needing to contact you for permission.
• Include provenance information by using a README file to document the "who, what, where, when, and how" of the data collection, including software versions, instrument settings, and any processing steps.
• Use the data structures and metadata standards common in your field (e.g. BIDS for neuroimaging or EML for ecology) to ensure the data remains meaningful to others.

R for Reuseable webinar

This webinar includes an overview of the four reusable principles, which bring together licensing, provenance and domain-relevant standards and resources to support institutional awareness and uptake of reusable principles.

Read the transcript.

The FAIR principles can complement the CARE and Māori Data Sovereignty principles by encouraging the consideration of both people and purpose. For more information, refer to CARE principles for Indigenous data and Māori data sovereignty. 

Practical steps for balancing these data principles

1. Publish a description or metadata-only record of research data in a data repository (e.g. Institutional Figshare) with citable DOIs for research. This enables others to discover and understand the applicability of the research data. Consider a meaningful name (e.g. avoid 'Thesis data'), appropriate metadata, and providing a sample of the data, actual or synthetic.
2. Establish and maintain a mediated access process (e.g. email request or form) linked to the published description. This process should follow agreed governance processes and ideally, take a 'people and purpose-oriented' approach to granting access.
3. Create a data sharing agreement to define and record who the data is shared with, for what purpose, under what conditions (e.g. method of transfer, security requirements) and by whose authority (governance). Dropbox for Researchers and the Institutional Figshare projects feature may be useful for storing agreements and sharing files with named individuals.
4. Produce a data availability statement within the publication (e.g. journal article, thesis) linking to the description, possibly referencing the mediated access process and alignment with FAIR, CARE and Māori data sovereignty principles.

• The FAIR principles (Force 11)
• The FAIR Guiding Principles for scientific data management and stewardship (Nature article)
• Cloudy, increasingly FAIR; revisiting the FAIR Data guiding principles for the European Open Science Cloud
• Jisc report: FAIR in practice
• SURF report: FAIR data advanced use cases
• FAIR data self-assessment tool
• The road to FAIR: FAIR principles for the Social Sciences and Humanities
• FAIRsharing: standards, policies and databases related to FAIR

Research Data Support Services
Email: researchdata@auckland.ac.nz

eResearch Engagement Lead
Email: Laura Armstrong