README for FAIR data

This week we had Zuzanna Zagrodska and Alain Danet telling us about how to make README files to think about making your data FAIR:

• Findable: this is making sure the data is somewhere searchable on the internet or that, upon publication, you clearly link to a doi on a permanent repository where the data is.
• Accessible: this is making sure anybody can download the data and read it on their machine (e.g. the data is freely accessible).
• Interoperable: this is making sure the data is can be modified, improved etc. This works by providing information on how it was collected, which format it is stored in etc.
• Reusable: this is making sure other users can reuse the data for their own work (licence etc.)

Note that I find that these points all overlap quiet a lot but they make for a nice acronym (yay!) and insist on the use of sharing some kind of format that humans and machines can read and understand to reuse your data.

To do that, the best way is to write a README file, as some metadata file, that you can share along your data. This README file should contain all the informations relevant to your data (what is it? why was it collected? how? by whom? with what? can it be shared? etc.). No need to be fancy when writing all that information but it’s important to be clear! Bullet points will do the job perfectly.

You can find their slides here to have a more detailed look along with examples on how to actually write these README files.

Zuzanna also made a great point saying that we end up writing these files in a rush just before submitting the paper to review and that can make it annoying/create friction/sloppy/etc. She thus rightfully suggested to write these README files from scratch even before starting to collect data! This allows you to think more clearly about what data you need and what could be some caveats with it etc. and should hopefully make your work more easy (definitely no rush at submission time)!