Anonymizing the research team during peer review.

[Matthew Harp]

Our Dataverse team was mentioned in a Twitter thread today. 

https://twitter.com/mserenity/status/1433148939662860289 

If I'm reading it correctly, and I'd like to think I'm reading it correctly, they are asking about how do you keep you and your team anonymous if you share your code/data during the peer review process. Unless you are using a third party resource I think using our Dataverse at a minimum gives away our institution. Is it possible to keep your whole dataset 'anonymous' when using a private link? 

Thanks for your thoughts,

Matt Harp

Research Data Management Librarian

ASU Library

https://dataverse.asu.edu

[Sebastian Karcher]

Hi Matt,

as it happens, this has just become possible with version 5.6 (https://github.com/IQSS/dataverse/releases/tag/v5.6 ) though this indeed can't mask the institution of a dataverse instance and I don't see how that would be possible -- you'd have to basically have an unbranded dataverse at a completely different URL. I guess you could copy the data to Harvard DV for peer review and then use the ASU copy for publication?

Best,

Sebastian

--
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 on the web visit https://groups.google.com/d/msgid/dataverse-community/172baa2f-11d5-47a7-973a-d9aa1cabbcbfn%40googlegroups.com.

--

Sebastian Karcher, PhD
www.sebastiankarcher.com

[Sherry Lake]

Ability to anonymize URLs is in the 5.6 release of Dataverse:

https://github.com/IQSS/dataverse/releases/tag/v5.6

Best, 

Sherry Lake

On Wed, Sep 1, 2021 at 3:44 PM Matthew Harp <mh...@asu.edu> wrote:

--

[Eunice Soh]

Hello, does anyone know if the aynonymised URL feature is enabled in https://demo.dataverse.org/? Would be interesting to try it out. 

[Philip Durbin]

Anonymized Access is not yet enabled on https://demo.dataverse.org but we're working on it and you can track the progress at https://github.com/IQSS/dataverse.harvard.edu/issues/119

So far we've come up with a long list of metadata fields we think we'd like to anonymize. If the experiment works well on the demo site, we'll most likely enable the feature on https://dataverse.harvard.edu

Hope this helps,

Phil

p.s. Buried in that Twitter thread is this fantastic tweet: https://twitter.com/Mammals_Suck/status/1433120274052833290

"Matt, talk to the awesome peeps over at @ASULibraries! @mserenity & @nightowlres set me up with a sweet dataverse under the umbrella of the project so all the datasets underlying different publications all get curated together. IT IS FREAKING AWESOME! https://dataverse.asu.edu/dataverse/MarchMammalMadness "

To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/75a49f3d-a52c-46a3-85f8-486fca400eb1n%40googlegroups.com.

--

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

[Matthew Harp]

Thanks everyone for your responses and information! This has been an interesting conversation and has been very helpful! Twitter can be quite the conversation starter. :-)

Kind regards,

Matt Harp

[Philipp at UiT]

At DataverseNO we have activated the URL for Anonymized Access feature in our Dataverse installation, but we currently still follow our own routines for anonymization including creating a new version of the dataset in an anonymized collection. The latter is possible to the extent that DataverseNO is a national repository, so even though someone would see that the anonymized dataset shared through private URL is deposited in DataverseNO, they wouldn't see which institutional collection it belongs to. The procedure is described in our curation guidelines; see the field "Related Publication": https://site.uit.no/dataverseno/admin-en/curatorguide/#metadata. Please let me know if you want to test the URL for Anonymized Access feature in our demo installation; I could create a test sub-dataverse and grant you curator access.

Best, Philipp

[sbar...@g.harvard.edu]

If anyone does need to test this feature and how it works without installing it first, it is now available on demo

Cheers

[Miguel Silva]

while reading through this conversation, I need to add the following in regard to "anonymity" and "anonymization".

I understand the need for "backward compatibility", a term widely used in computer sciences for the deployment of software solutions, however, with the current state and availability of this internet, asking for  "anonymity" is .... a "longer path to travel" .... one less trustworthy, one much slower. (...) This poses additional challenges in particular when the main concern is one of intellectual property (Patent).

 Other solutions exist that address and solve the same concerns while simultaneously offering similar, and even the same levels of protection of intellectual property. 

In regards to personal and even intimate data, (the hardest subjects and theme when on a public conversation \ chat ) the requirement is for real-time notification and real-time sharing of those who are using the data with the ones that originated the data.  A  good rule of thumb when thinking about data ownership ( anonymity) is: every time one finds the need for usage of some type of encryption or impose some type of  "access restriction", it means there's another, simpler alternative, as a solution to that specific problem. 

[Julian Gautier]

Thanks for the insights, Miguel!

You wrote that other solutions exist that address and solve the same concerns (about protecting intellectual property) while simultaneously offering similar, and even the same levels of protection of intellectual property. Could you share any of those solutions?

You also wrote that "in regards to personal and even intimate data, (the hardest subjects and theme when on a public conversation \ chat ) the requirement is for real-time notification and real-time sharing of those who are using the data with the ones that originated the data. Are you saying that it's important that the dataset owners (the ones that originated the data) should be notified in real-time about the people reviewing the data during a peer review process?