The archives used should adhere to the FAIR principles to ensure that the data is findable, accessible, interoperable, and reusable. By implementing the FAIR principles in research data management, you contribute to ensuring that data eligible for sharing is well-prepared for archiving.
Useful Resources:
Can My Research Data Be Archived?
At the conclusion of a research process, it must be assessed whether, and to what extent, research data can be shared. The principle that "research data should be as open as possible, as closed as necessary" should form the basis for this assessment. Several factors may limit how openly research data can be archived and shared, including legal, security, and ethical considerations.
- What can be shared openly?
- What can be shared with restricted access?
- What can be shared at a later date?
- What cannot be shared?
When academic, financial, resource-related, or other considerations require selecting specific data for archiving, the responsibility for this selection lies with the researcher, provided no specific requirements are set by the funder. Where established selection norms exist within the discipline, these should be adhered to. The evaluation should focus on identifying data necessary to ensure the verifiability of research results, as well as data with long-term value for the public, the research community, and the researcher.
It is also important to consider what documentation about the project is necessary to make available so that others can understand your research data.
What Should You Consider When Choosing an Archive?
There are national and international, general, and more discipline-specific archives that meet standards for archiving research data. Here are some tips on how to select an appropriate archive for your data.
It is important to follow current recommendations for archiving research data, and repositories should align with the FAIR principles to ensure that data is Findable, Accessible, Interoperable, and Reusable. The archive should support effective retrieval, follow recommended metadata standards, and use persistent identifiers.
Some archives provide curation, meaning that metadata is reviewed and quality-checked by specialists in the relevant discipline. The curation process typically involves a dialogue with the researcher to improve metadata quality before the dataset is archived.
Be mindful of the copyright associated with your data. Make sure to review the terms of use for the repository and be aware of any costs related to archiving.
In some cases, funders may require the use of a specific repository.
If you want to learn more about selecting a trusted repository for archiving your research data, we recommend consulting OpenAIRE's guide.
Discipline-specific Archives
To find a discipline-specific archive, it may be helpful to talk to colleagues at your department, as well as see what is used by colleagues internationally. Alternatively, you can use one of the following search tools:
- re3data.org – an overview of data archives.
- FAIRsharing Databases – an overview of databases and data archives.
General Archives
- DataverseNO – A national archive for general research data from all disciplines. Before uploading, it is important to read the DataverseNO archiving guide. Read "How to use DataverseNO for NIH researchers" [only for NIH employees].
- Sikt – A national archive managed by Sikt (the Norwegian Agency for Shared Services in Education and Research). Its focus is on preserving data from research in the social sciences and humanities, as well as some from medical and health research. Data are curated by Sikt.
- NIRD – A national archive for general research data. NIRD does not offer curation services for archived data.
- Zenodo – An EU research data archive that allows you to archive data and other documents from all subject areas. Zenodo does not curate archived data.
Archiving Sensitive Data or Data Containing Personal Information
Not all types of research data can be made available in open data repositories. Some data must remain restricted due to necessary exceptions and limitations. This includes data containing personal information or data considered sensitive for other reasons. Restrictions on data sharing may stem from legal, security, commercial, or ethical considerations. When restrictions apply, the rationale should be clearly documented in the Data Management Plan.
Ensuring the long-term preservation of research data is often essential, even when the data cannot be openly archived. When feasible, this can be done by implementing measures to access controls to the dataset. In many cases, data can be archived under restricted access, allowing only qualified individuals to access it under predefined conditions.
Lawful Basis for Processing Personal Data
Archiving, sharing, and anonymizing data containing personal information constitute the processing of personal data, which requires a legal basis. To archive and reuse research data containing personal information, it is essential to plan for this and establish a lawful basis in advance.
A guide providing concrete advice on what to consider when sharing research data containing personal information is now available. The guide provides a step-by-step introduction to ensuring that the sharing of research data complies with laws, agreements, licenses, and ethical guidelines.
There must be a data processing agreement in place with the repository that receives research data containing personal information on behalf of the institution. Read about NIH as data controller and about Privacy in research at NIH.
Available Research Data Repositories for Datasets Containing Personal Information
- Sikt archives all types of digital research data about people and society, including data containing personal information, and offers various forms of access control. Sikt does not archive anonymized qualitative datasets (e.g., transcribed interview data).
- FEGA Norway is a repository for human -omics data intended for reuse in research, with access control.
- CLARINO is a repository for language data, offering access control and licenses.
Anonymisation of Personal Data
Anonymized research data is information that cannot, in any way, be linked to individuals, either directly or indirectly. As long as information in the data can be connected to individuals, it will be considered personal data.
Before retaining the information in an anonymized form, you must ensure that there is a lawful basis for processing.
Useful Resources:
What Can Be Shared If Research Data Cannot Be Openly Archived?
Documentation of research data and metadata can be valuable contributions to an open research data repository, even if the dataset is archived with restricted access.
Metadata is essential for making archived data searchable and discoverable.
Licensing Research Data
When archiving research data, the data must be licensed. A license specifies what users are permitted to do with a dataset and regulates any conditions for its use. Standard licenses, such as Creative Commons licenses, are recommended because they are internationally recognized and easy to understand and apply for both humans and machines.
In general, a license that is as open as possible should be chosen to promote the reuse of data. By selecting an appropriate license, you ensure that research data can be shared and used effectively while adhering to legal and ethical standards.
When archiving open source code, it is recommended to use dedicated software licenses, as these are specifically designed to address how code can be modified and further developed.