Frank Smutniak (Apr 17 2026 at 17:33): Frank Smutniak (Apr 17 2026 at 17:33):When creating a new dataset the author name is pre populated but without orcid and ROR. If the user is required to type in a name then orcid and ROR will be denoted. Is there any way to suppress auto-population?
Philip Durbin 🚀 (Apr 17 2026 at 17:41):Hi! I'm not sure I'm following. :sweat_smile:
Philip Durbin 🚀 (Apr 17 2026 at 17:41):Name is always auto-populated.
Philip Durbin 🚀 (Apr 17 2026 at 17:42):If you have an ORCID associated with your account, it should be auto-populated too.
Philip Durbin 🚀 (Apr 17 2026 at 17:42):You want to suppress auto-population of both?
Bryan Gee (Apr 17 2026 at 17:46):Being tangentially related to this issue, I assume Frank is talking about suppressing auto-population of both, although with the ability to link ORCIDs in the account profile, auto-suppressing affiliation is maybe more important. One of the issues we run into in the Texas Data Repository is that because the depositor automatically has an entry created, with the freeform affiliation from their account profile auto-populated, they don't change that non-ROR-linked affiliation to a ROR-linked affiliation (though I discovered that if you have the plug-in enabled and put the ROR URL string as the affiliation in your account profile, the plug-in will recognize that and "convert" to display when creating datasets), and thus we often get datasets where everyone but the depositor has a ROR-matched affiliation.
Philip Durbin 🚀 (Apr 17 2026 at 17:48):You have ROR set up through https://github.com/gdcc/dataverse-external-vocab-support ?
Bryan Gee (Apr 17 2026 at 17:52):Yup
Philip Durbin 🚀 (Apr 17 2026 at 17:55):I see. I'm definitely not an expert in this. You might want to ask at https://groups.google.com/g/dataverse-community and see if Jim Myers will respond.
Philip Durbin 🚀 (Apr 17 2026 at 17:55):You could also open an issue in that repo.
Frank Smutniak (Apr 17 2026 at 18:02):Thanks. I'll ask there.
Philip Durbin 🚀 (Apr 17 2026 at 18:09):https://groups.google.com/g/dataverse-community/c/4GRpethFFRg/m/ZUeb4qUJCQAJ looks great. Thanks!
Philip Durbin 🚀 (Apr 17 2026 at 19:27):To make this a bit more visual...
... the first author (me) has a non-ROR affiliation that was auto-populated from my user acount.
The second author has a ROR-matched affiliation.
This is what we're talking about, right? :sweat_smile:
Bryan Gee (Apr 17 2026 at 19:44):Yep, and it's worse for institutions whose names don't fit in the field. Below is what I see in TDR; the top one is just the auto-populated freeform text, and the second is manually ROR-matched by me, but you can't even tell that there's a difference because the name trails off, which probably makes it impossible for depositors to even know there's a difference.
Philip Durbin 🚀 (Apr 17 2026 at 19:50):Oh, yikes, I see what you mean. :grimacing:
Philip Durbin 🚀 (Apr 17 2026 at 19:51):I would suggest opening an issue for this.
Julian Gautier (Apr 17 2026 at 21:26):Hi Frank and Bryan. Great conversation here. Some of the design challenges you've described are also included in an evaluation that my colleagues and I and other folks in the Dataverse community shared in a Google Doc last August, especially in two tables in the "Next Steps" section that lists reasons why the Dataverse repository might not have recorded ORCIDs and RORs for author name:
The Harvard Dataverse repository filled the Author Name field with information from the user’s account, that account didn't include an ORCID, and the user didn’t know they could use the field to find and select the name from the field’s suggestions so that the repository would include a ORCID
The Harvard Dataverse repository filled the Author Affiliation field with information from the user’s account, user accounts don’t include RORs, and the user didn’t know they could use the field to find and select the organization from the field’s suggestions so that the repository would include a ROR
The report builds on usability testing and benchmarking we did for Harvard Dataverse, but it's really helpful to hear that folks from other repositories have noticed similar things.
I'm pushing in the coming months to revisit and iterate on these designs, in an effort we'll be tracking in the GitHub issue at https://github.com/IQSS/dataverse-pm/issues/473 as part of our NIH-funded work at Harvard, and it would be incredibly valuable for us to collaborate :pray:
I'll also be talking more about this in one of the Thursday presentations during the Dataverse Community Meeting, in case you and other folks reading will be in Barcelona :beach_umbrella:
Julian Gautier (Apr 17 2026 at 21:26):Oh, I'll add this to the Google Groups thread you started, too!
Laura Huis in 't Veld (Apr 28 2026 at 08:58):This is slightly related, so I'll post this here:
At DANS, we use the ROR CV in the affiliation field, but we haven't connected the ORCID to the author name field (yet?). This results in a misalignment of the 'Identifier' field. This lay-out has caused some confusion, because some users were thinking that the 'Identifier Type' field belonged to the Affiliation field.
A screenshot here for clarity:
Laura Huis in 't Veld (Apr 28 2026 at 09:00):Also, I tested the set-up on demo.dataverse.org, and found out that if you save an author name without ORCID, and come back later to edit the metadata, the fields for 'identifier' and 'identifier Type' are not present any more, so you can't add the missing info at a later point in time.
Philip Durbin 🚀 (Apr 28 2026 at 13:39):@Laura Huis in 't Veld interesting, thanks. I wonder if this problem exists in the new React UI but I wasn't able to check. If you want, please feel free to go ahead and create an issue for the old JSF UI at https://github.com/IQSS/dataverse/issues
Julian Gautier (Apr 28 2026 at 15:20):The last time I checked, the React UI can't use the external vocab functionality yet, so this problem couldn't exist there. Jim's mentioned this a few times and we chatted with @Ellen K and others about this a while back so it's on our radars. Although I haven't found a GitHub issue about letting React UI users choose terms (like names or organizations or keywords) from externally managed vocabularies.
Julian Gautier (Apr 28 2026 at 15:21):Laura Huis in 't Veld said:
This is slightly related, so I'll post this here:
At DANS, we use the ROR CV in the affiliation field, but we haven't connected the ORCID to the author name field (yet?). This results in a misalignment of the 'Identifier' field. This lay-out has caused some confusion, because some users were thinking that the 'Identifier Type' field belonged to the Affiliation field.
A screenshot here for clarity:
Thanks! We've seen this bug, too!
Julian Gautier (Apr 28 2026 at 15:23):Laura Huis in 't Veld said:
Also, I tested the set-up on demo.dataverse.org, and found out that if you save an author name without ORCID, and come back later to edit the metadata, the fields for 'identifier' and 'identifier Type' are not present any more, so you can't add the missing info at a later point in time.
I think you can, but folks need to first remove the name and add it again. Then those Identifier and Identifier Type fields show up.
It's super helpful to know that you didn't realize this. Have your users been confused about this, too?
Laura Huis in 't Veld (Apr 30 2026 at 12:38):Ah, yes removing it first works, I see now.
We don't have the connection with the ORCID API in place, I was just testing it to see how it works. So, no confused users, just me :).
Julian Gautier (Apr 30 2026 at 12:45):Ah sorry, I missed that you already mentioned that your repository isn't using the ORCID API connection.
Last updated: May 30 2026 at 09:11 UTC