I’m admin of digital repository in the Scientific Institute.
According to the rules of research grants, an article (with results) can be published in the subscription journal (it would have DOI from publisher), but at the same time the AAM version must be placed in an open access institute repository with a separate DOI.
How can I register in CrossRef such an AAM article placed in the repository. As “posted content” - analogous to the preprint?
Thanks for your question.
We’ve discussed this at length recently, and our current position is that AAMs deposited in institutional repositories should not be registered at all.
This is because (unlike preprints) they are not likely to be cited in a manner that is distinct from the VoR journal articles, and therefore they should not have a distinct DOI.
“Mandatory criteria for repositories:
• Use of PIDs for the deposited versions of the publications (with versioning, for example in case of revisions), such as DOI (preferable), URN, or Handle.
• High quality article level metadata in standard interoperable non-proprietary format, under a CC0 public domain dedication. This must include information on the DOI (or other PIDs) both of the original publication and the deposited version, on the version deposited (AAM/VoR), and on the Open Access status and the license of the deposited version.”
(from “Plan S Principles”)
The grantee must comply with the requirements, that’s why I wonder, how to do it?
Our understanding of the Plan S criteria is that using the DOI assigned to the article’s VoR for the equivalent AAM would be sufficient. They’re not distinctly citable, so a single, unique DOI (registered by the publisher of the VoR) for both instances is the expected usage.
That said, a DOI is not often necessary by that standard, because many repository platforms create a Handle ID for each item in the repository as a default, and that alone would meet the Plan S criteria.
According to “Plan S Principles” requirements, the AAM/VoR version must to be published in open access.
At my Institute, grantee is going to publish research results in a subscription journal, so publisher will not publish VoR version in open access.
For the article’s author the only way is to publish his AAM version in institutional repository in open access (after aggreement with publisher).
So he need a DOI number for this article.
Yes, absolutely the AAM should be in an open repository in that situation. But, that doesn’t mean that version in the open repository needs its own DOI. It can still share a DOI with the VoR version, even when that VoR version is not open access.
Crossref DOIs don’t identify every discrete instance of a work. They’re citation identifiers, so they identify each instance with a unique citation.
The AAM version in the repository and the VoR version on the publisher’s website/platform (even if that’s behind a paywall, etc.) would be cited in exactly the same way by a researcher who was referencing the work. Therefore, they should both be identified by the same DOI, and in this case that should be the DOI registered by the publisher.
I’m going to pragmatically disagree here - ideally AAM metadata would propagate as part of the relationship metadata in the VoR. It doesn’t. Until those updates are widely enabled in platforms, DOIs need to be assigned to AAM in order that GreenOA works. While that is happening VoR and AAM DOI can be linked via updates to relations metadata. We really need the output from the new JAV WG from NISO to get a bit more clarity on how to think all this through BUT I might suggest that given the ongoing changes in the publisher landscape, AAM might become MORE, rather than LESS important - outline of these thoughts presented to Open Repositories 2021 here: 10.5281/zenodo.4945533
Thanks for your feedback. I agree that collecting AAM metadata as part of the VoR record would be very useful.
From Crossref’s perspective, the question of “should this thing get its own DOI?” is always dependent on the answer to “will it have a unique citation?”.
Our best understanding at the moment is that the typical researcher who’s citing a paper will generate the same citation whether they found that paper’s AAM in a repository or whether they found the paper’s VoR on the publisher’s site.
That doesn’t mean that the VoR is more important or that the AAM is less important. It means they are functionally, from the perspective of their citations, identical objects. The DOI is an identifier, not a stamp of approval - it doesn’t indicate which version is most important, most useful, most accessible, etc. Those will all necessarily vary from one user to the next and from one publication to another, and that’s just outside the scope of what a DOI can do.
You can think of DOIs for AAMs in institutional repositories working in the same way as DOIs in PubMed. There are many articles (some with full text, some just abstracts or metadata records) on PubMed. But, PubMed doesn’t register a new, distinct DOI for the PubMed version of each article. They use the DOI that was registered for the VoR by the publisher. They also create their own PubMed ID for each item, the same way that Institutional Repositories typically register a Handle ID or repository-specific PID for each of their items. A user wouldn’t specifically cite the PubMed version of the article in a way that was distinct from citing the VoR, but they could use either the PubMed ID or the DOI (or both) to identify it.
I hope that helps clarify.