Research data publishing and discovery

Publishing research data and artefacts is increasingly required by funders and journal publishers, and supports the internationally recognised FAIR principles for research data.

You may also be expected to publish your research data as per the University's Research Data Management Policy.

Publishing research data directly benefits researchers by:

• Increasing citations for your research outputs
• Directing readers to associated publications and materials
• Facilitating research collaborations

Before publishing your data, consider:

• Describing research data with a README (a summary of human-readable metadata or information about your data), which helps others to understand your research and associated data.
• Creating machine-readable metadata using a discipline-specific standard (if appropriate). For more information, refer to Research metadata.
• Naming and organising files in a clear, concise, and consistent way to make it easier for others to understand the data.
• Saving a version of the data files in open, universal formats (e.g. saving a spreadsheet as a .csv file rather than an Excel spreadsheet, .xlsx).
• When working with data in spreadsheets (tabular data), use the following tidy data principles to make it easier for other people to ingest, manipulate, and visualise the data:
  • Ensure each column contains one variable, each row contains one observation, and each cell contains one value.
  • Move separate data tables and sheets into their own spreadsheet files.
  • Remove highlighting, borders, notes, and other formatting.

These strategies support others to find, understand, and work with your published data.

• The University-hosted Institutional Figshare is suitable in most cases.
• A disciplinary or community-recognised repository may be appropriate, such as the CERN Open Data Portal.
• You may prefer a general data repository such as Zenodo.

To find a suitable repository, search re3data (Registry of Research Data Repositories).

Licencing a work makes the terms of reuse clear to potential users. Data publishing repositories such as the University of Auckland's Institutional Figshare repository or Zenodo offer several licences to choose from.

These usually include licences from the Creative Commons family of licences. The Creative Commons licence chooser provides a step-by-step process to select a licence based on conditions you choose.

If you're unsure which licence to select, contact Te Kahu Tauira, Student and Scholarly Services or Copyright advisory services.

Researchers may have legal, funder, or institutional requirements which make it necessary to broadcast the existence of a dataset without actually sharing it.

If you’re working with sensitive data, it may be more appropriate to publish a descriptive record about the data instead of the data or artefact itself. A descriptive or metadata-only record indicates the existence of a particular dataset and provides information about it, without sharing the actual data.

Publishing a metadata-only record supports:

• Data and artefacts to be findable and accessible online, aligning with FAIR principles for research data, and
• Collective benefit, governance and the authority to control access and use of the data associated with CARE principles for Indigenous data or Māori data sovereignty.

A metadata-only record provides instructions for requesting access to access-controlled data. This process, known as mediated access, involves a review of the requester and their intended use before the data is released.

This process can be undertaken by an individual or a data access committee. To help researchers evaluate the dataset’s utility, the record may also include a real or synthetic sample file to demonstrate the quality, format and content before a formal application is submitted.

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

eResearch Engagement Lead
Email: Laura Armstrong