[Rdap] data repository assistance please

[Daureen Nesdill]

Hi all,
The group went with http://reshare.ukdataservice.ac.uk/legal/ Which I have not had a chance to look review.

From this experience, I have two lingering questions:
How many scholarly publishers (and which ones) are establishing these types of criteria and providing “badges”?

Do data repositories regularly time-stamp and freeze data sets? Since I work with ELNs (electronic lab notebooks) I’m aware these systems generally do (FDA regulations)  provide time-stamp and freezing of notebooks. I’m wondering if ELNs’ time-stamp can substitute for repository time-stamp

Daureen


From: Rdap [mailto:rdap-bounces at asis.org] On Behalf Of Ruth Duerr
Sent: Thursday, April 27, 2017 5:47 PM
To: Research Data, Access and Preservation <rdap at asis.org>
Subject: Re: [Rdap] data repository assistance please

Hear! Hear!  +1


On Apr 27, 2017, at 5:19 PM, Sebastian Karcher <skarcher at maxwell.syr.edu<mailto:skarcher at maxwell.syr.edu>> wrote:

(I have a message in moderation, which can now be deleted, assuming this goes through).

I'd like to make a strong plug for going with ICSPR over half-baked solutions here. ICPSR has a huge amount of experience and significant technical infrastructure to handle highly sensitive data (and most things that are directly identifiable qualify). They charge money for the use of their virtual enclave because that's very expensive to maintain (which is why few other places do it).

The problem with putting a metadata entry in an IR is that you don't get preservation services, which is half the reason you put data in a repository in the first place. Even if we assume the researcher handles access requests, who will provide multiple back-ups, fixity checks/protection against link rot, file type migration if needed, etc.? The reason this is somewhat viable in Andrew's scenario is that the data are actually properly archived elsewhere. If that's not the case, this is, in my opinion, a no go, and repository managers should not allow, let alone promote, it.

Harvard Dataverse (as well as Zenodo, another self-deposit option that allows restrictions) is _not_ a robust solution, at this time, for sensitive data (this is planned for the 5.0 release of the software they use; they're currently at 4.6 [1]). While it is possible to restrict access to data, I am pretty sure that they are stored unencrypted on the server. I know that's not acceptable for sensitive data from an information security point of view and I _think_ (but am not a lawyer) you may actually break the law by using it for personally identified data.

So please make a pitch to your researcher for the value of curated, secure repositories like ICPSR for sensitive data.

My two (rather passionate) cents on this.
Sebastian


[1]
https://groups.google.com/d/msg/dataverse-community/ZQUQhVLiSoo/1yjhPFthAAAJ

________________________________
From: Rdap <rdap-bounces at asis.org<mailto:rdap-bounces at asis.org>> on behalf of Daureen Nesdill <daureen.nesdill at utah.edu<mailto:daureen.nesdill at utah.edu>>
Sent: Thursday, April 27, 2017 2:03 PM
To: Research Data, Access and Preservation
Subject: [Rdap] data repository assistance please

Hi all – Hope you had fun in Seattle.

I have a researcher in psych with spreadsheet data that cannot be de-identified. She wants to publish this data in a repository that freezes and time stamps the data, in addition to being only available to qualified researchers.

Any ideas?
We have investigated ICPSR, and re3data.

Daureen

Daureen Nesdill, MS, MLIS
Research Data Management Librarian
The Faculty Center @ the Marriott Library
daureen.nesdill at utah.edu<mailto:daureen.nesdill at utah.edu>
801-585-5975
ORCID http://orcid.org/0000-0003-0126-5038

_______________________________________________
Rdap mailing list
Rdap at mail.asis.org<mailto:Rdap at mail.asis.org>
http://mail.asis.org/mailman/listinfo/rdap

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kunverj.com/pipermail/rdap/attachments/20170501/2325f97f/attachment.html>