Data-Sharing and publication
The decision about what data for a project should be archived and for how long depends on the academic value of the data as well as on legal, regulatory and financial factors.
As a minimum, however, all the data on which a publication is based must be stored and the corresponding metadata must be published online.
The Digital Curation Centre (DCC) and forschungsdaten.info list five steps for deciding what data to keep.
Finding a repository
Wherever possible, data should be deposited in disciplinary repositories. These are designed to meet the needs of the particular field, are aware of specific data formats and often also offer specific disciplinary metadata.
On its website SNF provides a checklist (pt. 5.1) that you can use to check whether your chosen repository complies with the FAIR principles.
The best starting point for finding a suitable repository is the Registry of Research Data Repositories (re3data.org). The Open Access Directory and PLoS also provide an extensive list of data repositories.
The following repositories adhere to the FAIR principles and are approved by SNF. They are open to researchers in all disciplines. The list is not exhaustive.
BORIS satisfies the requirements for a repository as specified by SNF. It is therefore possible to store supplementary data for publications for which there is as yet no disciplinary repository in BORIS. However, BORIS is not suitable for large raw datasets. We recommend you store these in a general repository.
An institutional data repository (BORIS Research Data) is currently under construction.
Choose a license
Before being published, data should be provided with a license. You could use Creative Commons licenses version 4.0 to do so. You find more information about Creative Commons licenses here.
Research data generated and collected during a project can often be useful beyond its original purpose. It is therefore worthwhile making the data obtained publicly accessible. For this purpose it is important to ensure that your data is assigned persistent identifiers, good metadata is generated and sufficient documentation is provided to enable the data to be reused.
There are currently three ways of publishing research data.
Publication in a repository
Research data can be published in a disciplinary or a general repository. If possible, it is preferable to publish data in a disciplinary repository rather than in a generic one. Further information about selecting a suitable repository can be found in Finding a repository.
Publication in a data journal
Data papers published in data journals are documents that facilitate the dissemination and reuse of published data. These publications contain all information about data collection, methods, licenses and access rights along with information about potential reuse opportunities. The data itself is usually deposited in a repository.
The website of the Humboldt University of Berlin has a list of data journals.
Publication as a supplement to an article
Data can also be published as additional information for an article in a periodical. This is usually the data on which the publication is based which enables the findings to be understood. The data may either be deposited directly on the periodical's platform or in an external data repository.
When citing data it is advisable to use either the standards applicable to the research field in question or the form suggested by the repository in which the dataset was deposited. If there are no particular standards or recommendations, Datacite recommends providing the following details as a minimum:
- Year of publication (of the dataset)
- Edition or version (optional)
- Publisher (for data this is usually the archive in which the data is stored)
- Resource type (optional)
- Persistent identifier (as a permanent linkable URL)