Hi Gail and all,

Might you use the Date property of the DataCite record, with the dateType=”Updated” to record the as-of date? I’m not asserting this as a best practice, but rather asking if this makes sense to those following the discussion.

--Joan

 

Joan Starr

EZID Service Manager

California Digital Library

University of California Office of the President

wk: 510-987-0469

twitter: joan_starr

http://orcid.org/0000-0002-7285-027X

she, her, hers

 

 

Thanks for the very helpful discussion to date!

 

Caltech has been wrestling with what best practices to adopt in the event a dataset is deposited in CaltechData that indicates, in the DataCite metadata field for “related to” that it is identifical to another dataset. We recognize that two datasets may be identical at a given point in time (ergo, when one dataset is deposited in our repository) b ut are struggling with the curatorial task of disclosing that the datasets may diverse down the road.

 

We’ve been looking for guidance on how to populate the DataCite metadata template in cases of duplicate datasets when there is a need to deposit identical datasets in multiple locations (say to meet funder and institutional requirements). We also wonder how to guide users when they’ve duplicated a dataset but then added to it. It would be ideal to be able to couple the ‘identical’ assertion with an “as of” date to indicate the last know time the two datasets matched.

 

Appreciate any guidance or examples of how you all have expressed the relationships between datasets in cases where duplication was practiced.

 

Gail

 

Gail P. Clement  | Head of Research Services  | Caltech Library  | Mail Code 1-43  | Pasadena CA 91125-4300  | 626-395-1203

http://orcid.org/0000-0001-5494-4806 | library.caltech.edu

 

 

 

Hi Dan,

I’m sure the ESRC would be very happy with data they funded being shared as widely as possible, as long as it was also deposited with UKDS. Whether there is a more subject-specific archive or repository is another question. I doubt any repository can offer data management to a higher standard than UKDS, but some might provide their users with features that appeal to particular audiences relevant to the sub-discipline, provided of course that the researcher puts some effort into preparing the data so it can make use of these.

 

How you catalogue the multiple datasets is another issue, as others have pointed out. But if there is any difference at all between the deposited datasets they would need different identifiers anyway. There’s no need for researchers to assume they will publish one single definitive dataset from their research, any more than they publish one single definitive article.  And beyond contractual obligations etc. its sound advice to deposit the data wherever it will get the most interest.

 

Best wishes,