After the completion of your research, you are obliged to keep the raw research data for at least ten years and to make them available to other researchers upon request. You can do this yourself or you can elect to deposit your data in a data archive or repository.
On this page:
The Netherlands Code of Conduct for Scientific Practice requires that the raw research data are stored for a period of at least ten years.
Choosing a repository
If your research funder or the journal in which you publish does not prescribe a specific data repository, you are free to choose where to deposit your research data. Apart from the costs, the repository may charge you for preserving your research data, there are four other points to consider:
- the quality of the repository
- the options offered by the repository: persistent identifier, restrictions of access, licences
- requirements of the repository: file formats, metadata
- findability of the data
The quality of data archives may be assessed in various ways: certification according to the criteria of the Data Seal of Approval, the Trusted Repository Audit Checklist (TRAC) or the Trusted Digital Repository Checklist (TDR, ISO 16363).
If you wish to deposit your research data securely in a repository but do not (yet) wish to share them with others, it is important to know what possibilities there are for creating restricted access. Is it possible to deposit the data, but publish only the description of the dataset? Is it possible to place an embargo on the publication of the data? Is it possible to give access to the datasets only on request? There are four likely scenarios, but not all data archives support all four:
- description and dataset are both public
- description is public, dataset becomes public after an embargo
- description is public, dataset is shared with specific users only (upon request)
- description is public, dataset is not public
If you wish to archive and publish your research data, you must find out what exactly is allowed by the repository: what type of licence will there be for your research data? Will the repository give your dataset a persistent identifier, so that others can simply and correctly cite your dataset? Does the repository provide statistics, so that you can see how often your dataset is consulted?
Repositories can make requirements as to file formats and metadata. Various repositories indicate which file formats offer the best guarantee for usability and access. In any case, a repository will require metadata so as to enable efficient retrieval via the search function on the repository's website.
The choice of repository may influence the findability of your data. A large international repository is likely to be consulted sooner by your fellow researchers than a small local one, but on the other hand, in a large repository your dataset is only one among many. By depositing your data in a repository for your own discipline you are likely to reach mainly researchers from your own discipline, but not those from other disciplines.
Types of repositories
A majority of repositories are discipline specific. In a data-repository register you can find relevant repositories by discipline.
General repositories are not limited to data from a specific discipline or a specific research institute. National repositories are also general repositories, such as the UK Data Service (UKDS) and the Australian Data Services (ANDS)
A variant of general repositories are institutional repositories: repositories which accept data from all disciplines, provided they originate from the institute - for example a university - which manages the repository.
The Library UvA/AUAS is currently developing a research data repository.