How to make data FAIR
1. Inform Yourself
Making data FAIR means making them findable, accessible, interoperable, and reusable.
2. Generate Metadata
Data documentation and metadata are essential for the reproducibility and reusability of research results. In the simplest sense, metadata is information about data and describes its basic characteristics (who?, what?, when?, where?, why?, how?).
To create metadata, you can create a machine-readable file (e.g., using XML) or a README-style PDF.
More on data documentation and metadata
Generate metadata from the beginning and save time at the end
3. Use a Controlled Vocabulary
Controlled vocabularies are lists of predefined, authorized terms that are commonly used and recognized in a discipline. You can use these in addition to metadata standards for naming your variables. Controlled vocabularies can be found at http://fairsharing.org/standards (search by subject).
4. Choose a Repository with a Persistent Identifiers
A persistent identifier is a permanent and unique reference (i.e., link) to a digital object, regardless of changes to its online location. Persistent identifiers can uniquely reference not only records and publications but also people (e.g., with ORCiD ID).
5. Publish Your Data in an Open Format
Choose a file format that can be edited with different software and is not dependent on paid software. Here you can find more information about open file formats.
6. Publish Your Data with an Open License
If you own the copyright to your data, you can use a license to inform how your data may be re-used. For example, you can provide the license in a README file, prominently on the download page or in a separate license.txt file.
Use:
Creative Commons license (note: use version 4.0 for data sets)
Open Data Commons license