The SNSF expects that the so-called FAIR principles are adhered to when publishing data.  These guarantee that the data will be Findable, Accessible, Interoperable and Reusable.

For researchers, this means, among other things, the following:

• The SNSF recommends the use of non-commercial repositories that comply to the FAIR principles. For the social sciences and humanities, FORSbase and DaSCH can be recommended, for example.
• In order to guarantee findability and reusability, research data must be enriched with comprehensive, systematic and standardised (as far as possible) metadata.
• Making the data accessible does not mean that the data must be completely open and freely accessible. Rather, it means that information is provided about HOW the data can be accessed. If research data cannot be published due to ethical or legal reasons, then the metadata should still be accessible.
• When storing the data, suitable file formats should be selected that have widespread use in the scientific community, are compatible with different systems and ideally should be free of charge and open source. File formats that require specialised algorithms or uncommon programs should be avoided.
• The research data should be provided with a persistent identifier, so that they can be permanently and uniquely identified, e.g. through a DOI (Digital Object Identifier) or an ARK (Archival Resource Key). Many repositories automatically generate a persistent identifier.  
• Data should be provided with a licence such as, for example, a Creative Commons licence which regulates its reuse.

The University Library will support you in finding the most suitable repository for your data and managing your data according to the FAIR principles.

When starting a project, it is strongly recommended that you select a logical and consistent data organisation that will enable you and others to easily find, access and use your data, thereby avoiding duplication of work, and ensuring that your data can be backed up. The following tips can help you to develop your own organisation system:

• Use a hierarchical and clearly arranged file structure.
• Separate active and completed work and delete all unused temporary files.
• Establish a naming convention for file names.
• Use file formats that are widespread and preferably open source.

Describe your files using good documentation and metadata.

Documentation implies creating good information for later use and application. The aim is to make information or documents findable and reproducible. Structured information about an object is called metadata.

If research data are to be published or archived in a repository, metadata that reflect the contents of a document are essential, thereby making it easier to find the document. A description of the contents can be in the form of subject headings and abstracts. For indexing, the use of a standardised vocabulary is recommended. An overview of freely-accessible vocabularies can be found here: http://bartoc.org/.

Ideally, documentation and metadata should already be recorded on an ongoing basis during your research. It is recommended that an internal project standard should define how data will be annotated and stored. Meaningful naming of files and information in the individual documents (e.g. information about time, place and interviewee in an interview transcript) should also be included.

Research data are increasingly openly accessible via data archives, supplementary material in scientific journals and on the websites of research groups. Multidisciplinary and subject-specific archives may be located with the help of Re3data — Registry of Research Data Repositories.  Some data records can be found through search engines for data, e.g.:

• openaire explore
• search.datacite.org
• google dataset search
• data search (elsevier)

Research data must be cited in the same way as other publications in the spirit of good academic practice. We recommend the following details in the usual citation style: author, data record name, repository, version, persistent identifier.