Researcher documentation/READMEs

[paul....@ubc.ca]

Hi Dataverse community,

I'm not sure if *all* of us have the same problem, but here at UBC it's sometimes challenging to get researchers to provide a basic text file to go with any data deposits. Because researchers already have to interact with Dataverse to produce at least minimal metadata, UBC decided to make a tiny utility so that you can generate a basic README from a data deposit.

Depending on the file(s) deposited, it may also churn through those and produce some summary stats, as for SPSS, Stata, etc.

Obviously this is not a complete subsititute for documentation, but it allows curators to produce a reasonably formatted document which they can then send back to depositors and say, "Hey, can you fill this out with a little more detail?" More importantly, because researchers *must* fill out fields when depositing data, you are at least assured of receiving *some* metadata, and if a researcher only has to fill out one thing there's a higher chance of success. Hopefully.

The utility is  available as part of the dataverse_utils suite (https://pypi.org/project/dataverse-utils/), and has the unremarkable name of `dv_readme_creator` (https://ubc-library-rc.github.io/dataverse_utils/scripts/#dv_readme_creator)

To give you an idea of what it looks like, here's sample of the Markdown it produces from this study in Borealis:https://borealisdata.ca/dataset.xhtml?persistentId=doi:10.5683/SP3/WSZ1W5

This one had a README already but I'm sure you get the idea. It also makes PDFs instead of plain text, although that may have less utility. But it *is* prettier.

Anyway, it's there if you need it or want it.

Paul Lesack
University of British Columbia Library

[Philip Durbin]

Hi Paul,

This is great. Thanks for letting us know about it.

I mentioned this the other day in a meeting for a project we're calling Trusted Data*. One thought we have is that the presence of a README could be one of many indicators of quality and trustworthiness. So, tools like yours can help with this.

I see that we already have dataverse-utils listed in the guides here: https://guides.dataverse.org/en/6.9/api/client-libraries.html#python. You are welcome to create a pull request, of course, to mention the new functionality or otherwise update what's written there.

Thanks,

Phil

* https://github.com/IQSS/dataverse-pm/issues/425

--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dataverse-community/fa62aa01-0c07-4c5a-9f15-d9810a43428en%40googlegroups.com.

--

Philip Durbin
Software Developer for http://dataverse.org
http://www.iq.harvard.edu/people/philip-durbin