Give us yer f.cking data

[sTeamTraen]

... as Bob Geldof didn't (even with paraphrasing) say.

For those with fancy-pants academic access: https://journals.lww.com/anesthesia-ana ... .1600.aspx

For the plebs, the final draft text: https://docs.google.com/document/d/1Nd1 ... mO8ay/edit

(Yes, I got to write a paper with the redoubtable John Carlisle. This is an actual #AchievementUnlocked™)

[science_fox]

In my field of Mass Spectrometry we are already seeing this - most journals require the raw data files to eb uploaded to a repository eg https://www.ebi.ac.uk/pride/ . I'm not sure anyone is routinely checking them for manipulation though.
And congrats on the paper!
Still difficult to get these sorts of opinions published!

[shpalman]

Physics is doing this, mainly using Zenodo.

Sergey Frolov gave a talk (because I asked him to) on open data in condensed matter physics (and what happens when you don't have it) and it's on his youtube channel at https://youtu.be/_3t0iIs8xxc

Phrases like "data are available from the authors on reasonable request" are no longer considered acceptable.

[dyqik]

NASA requires that scientific data from it's missions are released within 24 hours of collection.

It's a little difficult for us to do that with VLBI data though

[Gfamily]

NASA requires that scientific data from it's missions are released within 24 hours of collection.

It's a little difficult for us to do that with VLBI data though

You release it correLATER !! (I think this is a VLBI joke, but I'm not an expert)

[dyqik]

NASA requires that scientific data from it's missions are released within 24 hours of collection.

It's a little difficult for us to do that with VLBI data though

You release it correLATER !! (I think this is a VLBI joke, but I'm not an expert)

It works

[kerrya1]

Phrases like "data are available from the authors on reasonable request" are no longer considered acceptable.

If only this were actually true. I'm currently working on assessing the compliance of grant holders with funder FAIR data requirements. This should be easy to do by extracting the Data Access/Availability Statement (DAS) from a publication, following the persistent identifier to the landing page and checking that minimal metadata standards have been met.

The majority of papers still don't include an easily identifiable DAS, and a lot of those that do contain some variation of the old "on request" statement.

Things are very, very slowly getting better but until funders and publishers actually enforce requirements for a meaningful DAS, and employers improve recognition and reward for researchers who put in the effort to make their data FAIR then I don't think things are really going to improve majorly.

On the plus side I am currently working on a project which should inform updated UKRI (all councils) guidance and requirements om data management and sharing.

[shpalman]

Phrases like "data are available from the authors on reasonable request" are no longer considered acceptable.

If only this were actually true. I'm currently working on assessing the compliance of grant holders with funder FAIR data requirements. This should be easy to do by extracting the Data Access/Availability Statement (DAS) from a publication, following the persistent identifier to the landing page and checking that minimal metadata standards have been met.

The majority of papers still don't include an easily identifiable DAS, and a lot of those that do contain some variation of the old "on request" statement.

Things are very, very slowly getting better but until funders and publishers actually enforce requirements for a meaningful DAS, and employers improve recognition and reward for researchers who put in the effort to make their data FAIR then I don't think things are really going to improve majorly.

On the plus side I am currently working on a project which should inform updated UKRI (all councils) guidance and requirements om data management and sharing.

Is there something official regarding "minimal metadata standards"?

[kerrya1]

There is no universal minimal metadata standard which is required by all funders / publishers and there probably won't ever be. The closest thing is probably the Datacite Metadata Schema (https://datacite-metadata-schema.readthedocs.io/en/4.5/) simply because DOIs have become the most common of persistent identifier in use currently.

The problem is that the Datacite schema is designed to be used for any type of data from any discipline, this means that it is good at FA (Findable, Accessible) but less good for IR (Interoperable, Reusable) as these facets are often dependent on data format and disciplinary standards / norms.

Having said that if we could get everyone using the Datacite scheme M & R properties, and doing it well, then that would be a huge step forward,