Preserving and publishing data (FAIR)

FAIR principles are widely adopted guidelines to make data discoverable and reusable by humans and machines: Findable, Accessible, Interoperable, Reusable, as first formally published by Wilkinson et al. (2016). Applying FAIR makes your datasets findable, accessible under clear conditions, interoperable and reusable, which increases visibility and citations, improves transparency and reproducibility, and reduces duplication and costs. It also keeps you aligned with community standards and fulfils funder requirements while respecting proportionate openness—as open as possible, as closed as necessary.

• FAIR ≠ open: if data can’t be open, publish the metadata and set appropriate access conditions.
• Use a trusted repository to preserve (archive) and/or publish: e.g. 4TU.ResearchData or DANS Data Stations. A persistent identifier (PID)—usually a DOI—is assigned by the repository on publication.
• Provide rich metadata and a short README (guidance / template) so others can understand and cite the dataset.
• Choose a clear licence (e.g. a Creative Commons licence) to state reuse conditions.
• Favour open or widely used formats (4TU.ResearchData, DANS) and software whenever it is possible and keep consistent file/folder names.
• Link your ORCID, include scripts/code, and use community standards (metadata schemas, controlled vocabularies, ontologies), where relevant.
• For research software/code, see the five recommendations at fair-software.nl (repository, licence, registry, citation, checklist), and the Research Software Management section at the UT Service Portal.

For general advice on FAIR at UT, contact the UT FAIR data steward. For discipline-specific questions, consult your faculty data steward(s).

Data preservation (archiving) prevents loss and protects authenticity, improving research quality through verification and reuse, and fostering new collaborations.

• For preserving data at the UT and sharing within your research group or with specified external collaborators, use Areda.
• For publishing data, use 4TU.ResearchData, DANS or another trusted data repository.

The scope of the UT RDM Policy and the research group’s responsibility for data preservation (archiving) does not include research data collected or generated by bachelor’s or master’s students. However, study programmes or research groups may have their own policies and guidelines (e.g., on selection, retention periods, supervision, or where to store). Where appropriate, they may use Areda to preserve student datasets that are static (no longer changing). Please consult your programme or research group for the applicable guidance.

In general, selecting the research data to preserve (archive) is the responsibility of the researcher, the (former) project leader, the supervisor in case of BSc/MSc students, or the head of the research group. What must be preserved should follow the data policy of the research group or higher organisational entity and any agreements with involved third parties. (For retention, see UT RDM Policy Section 3.6: at least ten years.)

Before preserving or publishing, check the applicable agreements (e.g., project or consortium agreements) and follow their conditions alongside the UT RDM Policy.

Trusted repositories assign a persistent identifier (PID)—usually a DOI—to support long-term access and citation. Watch this video about Persistent identifiers and data citation explained (Research Data Netherlands).

When publishing in a data repository, for proper reuse it is important to add metadata and a README file with documentation (guidance / template). You can use the same you added in Areda. Before publishing, please check specific project or research group policies. See furthermore step by step publishing guidelines for 4TU.ResearchData and for DANS.

Once your dataset is published, cross-link it with your articles. Add the article DOI in the dataset metadata (where supported), and cite the dataset DOI in your article (Data Availability Statement and reference list).

Both 4TU.ResearchData and DANS Data Stations support adding references to related publications in the dataset record.

For new submissions, consider mentioning the dataset citation in your cover letter so reviewers can verify the data.

Areda is the University of Twente archive for the long-term storage of static data, which is collected, generated or used in UT research projects.

Areda offers research groups:

• Certified storage for long-term preservation (at least 10 years) of static datasets.
• A ‘bucket’ where data archive files (zip or tar) can be uploaded by group members.
• Technically suitable storage for all kind of static data, including special categories of personal data.
• The possibility to add metadata and documentation making use of the UT research information system (Pure).

Please, keep in mind that:

• Areda should not be used for datasets that still might change (dynamic data), or need to be used regularly.
• For the moment there is a maximum upload of 1 TB.
• Access to datasets is at research group level: all group members have access to the datasets in the group’s bucket (folder). Groups and members are based on the HR system.
• In case of uploading personal or other confidential data, access has to be managed by using encryption. Areda offers an encryption instruction. Encryption keys need to be securely managed by at least one permanent staff member of the research group besides the principal investigator, preferably the chair of the research group.
• Areda should not be used for archiving digital informed consent forms or pseudonymization keys. These must be stored encrypted and separately from the anonymized (or pseudonymized) data, for instance in JOIN (for ET, BMS, and ITC, please contact the data steward for the procedure), or the p-drive.
• Although in the group’s bucket a folder structure (by creating a path) is possible, it is advised to use it as a single archive (database) of zip-files containing a structured set of data files of a research project.
• For making overviews and searching of datasets, use the UT Research Information Portal.

Archiving of datasets demands some preparation:

• selecting and organizing the data files,
• writing data documentation,
• creating an archive file (zip or tar), and
• encryption in case of personal or other confidential data.

For more information, please consult the guide Archiving datasets in Areda.

Archiving datasets in Areda means upload the archive file you created during the preparation to the bucket of your research group. Next, add metadata about the dataset, such as title, creator, etc. and a copy of the README file in the UT research information system (Pure). There you can also link the dataset to one or more of your publications.

Here is a visual of the archiving intake process:

To help you with the intake process, please consult the guide Archiving datasets in Areda.

Metadata will be reviewed by the data steward of the faculty. After this, the archive file will be transferred to the bucket of the research group.

Apart from archiving in Areda, it is recommended to publish research data and share it with others outside the UT. Sharing or publishing datasets means that you upload them to a trusted data repository, preferably 4TU.ResearchData or DANS.

Datasets are first archived in the research group bucket before moved and permanently archived at Surf Data Archive (SDA). All research group members have access to the bucket of their own group. Access to files can only be restricted by means of encryption.

It is advisable to add terms of use in the documentation of the dataset in the README file (guidance / template).

After data sets are moved to SDA they can be retrieved by means of an access request. To make a request, please fill in the form. The department chair will examine your request and decide if access can be granted. You will be notified of the outcome.

The General Data Protection Regulation (GDPR) requires that personal (any information which are related to an identified or identifiable natural person) data are not kept longer than necessary to achieve the purposes for which they are processed. If longer preservation is needed, anonymization is preferred. More information about handling personal data in research can be found here.

In any case, be sure that you registered the processing of personal data in compliance with the GDPR. When you indicated in your GDPR registration the need of preserving personal data, you should act in accordance with this information about preservation and protection of the data.

When archiving datasets with personal data, please be aware that file encryption is needed. Instruction about file encryption is offered during the Areda archiving process.

You can archive datasets in the UT facility Areda. Apart from that you can publish the dataset in a trusted repository, preferably 4TU.ResearchData or DANS.

You can archive all types of datasets, both as supportive material to a publication (PhD-theses, journal articles, etc.) and as stand-alone items.

Datasets may be accompanied by related materials, such as

• specific viewing and analysis tools (models, algorithms, scripts, analysis or simulation software, schemas)
• laboratory or field notebooks, diaries
• questionnaires, transcripts, codebooks
• standard operating procedures and protocols
• informed consent forms

Yes, as long as archiving fulfils the following requirements:

• it is aimed at securing data authenticity, verification and/or reuse
• the data are static, so not subject to changes anymore.

See: What data should I archive and publish?

The M- or P-drive are not suitable for long-term persistent storage of data because they cannot guarantee that the datasets will remain immutable, which is a prerequisite for the authenticity of the data.

Trusted data repositories are certified because they abide to 16 requirements, such as

• the explicit mission to provide access to and preserve data in its domain,
• having a continuity plan to ensure ongoing access to and preservation of its holdings and
• having adequate funding and sufficient numbers of qualified staff managed through a clear system of governance to effectively carry out the mission.

Whether datasets are FAIR highly depends on the way they are described by means of metadata and documentation, more than merely the location where they are archived or published. Look at Making data FAIR for more information.

The table below presents what to archive and publish depending of the purpose of preservation.

The General Data Protection Regulation (GDPR) requires that personal data (any information related to an identified or identifiable natural person) are not kept longer than necessary to achieve the purposes for which they are processed. If longer preservation is needed, anonymization is preferred.

When datasets contain personal data, please be aware that file encryption is needed when archiving. Instruction about file encryption is offered during the Areda archiving process.

If the data subjects agreed in the participation of your research project, it means that they implicitly agreed in archiving the research data. Therefore, they cannot demand that their data should be erased before the end of the retention period.

In case data subjects want to exercise their right, they can contact the Data Protection Officer (DPO).

Use general, non-proprietary formats to prevent loss of access to files and to enhance the chance of future interpretability of the data. Preferred file formats are e.g. PDF, Plain text, TIFF, FLAC, CSV or XML (see also extended list of formats from DANS or from 4TU.ResearchData).