Sooner or later a form asks for your ORCID iD. Then a repository hands you a DOI, and a funder's template wants an institution identifier too. It's easy to treat these as paperwork. They're actually the plumbing that lets your work be found, cited, and credited to you instead of to the other person who shares your name.
This guide explains the three identifiers you'll meet most often, what each one is for, and a short setup list you can finish in an afternoon. It's written for researchers, and for the librarians and research staff who get asked about this every semester.
What a persistent identifier is
A persistent identifier is a permanent label for something: a person, a digital object, or an organization. The label stays the same even when the thing moves. A web address breaks when a lab changes its site or a department reorganizes. An identifier is meant to keep working, because a registry keeps track of where the thing currently lives and sends people there.
That's the practical point. Names are ambiguous ("J. Smith" has written a lot of papers), and links rot. Identifiers fix both. They're also a big part of what makes data findable and reusable in the FAIR sense, since a machine can match an identifier exactly where it could only guess at a name.
ORCID: an identifier for you
An ORCID iD is a 16-digit number that belongs to one researcher, written as a link like https://orcid.org/0000-0000-0000-0000. It's free to register, and you decide what appears on your record and who can see it. Publishers, funders, and repositories use it to connect your name to your work without confusion over spelling, name changes, or common surnames.
A few habits make an ORCID record worth having:
- Register with a personal email as well as your work one. Add your institutional address, but keep a second address on the account so you can still get in after you change jobs.
- Fill in the basics. Your name, current affiliation, and a short list of works. An empty record doesn't help anyone.
- Connect it to the systems you use. Many journals, repositories, and funders let you sign in with ORCID or import works from it, which saves retyping and keeps the record current.
- Check what's visible. Each item on the record has its own visibility setting. Make the ones you want public, and leave the rest.
- Use it everywhere. Put the iD on your CV, your website, your email signature, and every submission form that asks.
One more tip for people who support researchers: ask whether your institution has a membership or a sign-in integration with ORCID. If it does, the setup is smoother, and you'll have a ready answer when faculty ask why they should bother.
DOIs: an identifier for the work
A DOI, or digital object identifier, is a permanent identifier for a thing you produced: a journal article, a dataset, a report, a piece of software. It looks like 10.xxxx/something, and when you put https://doi.org/ in front of it, the link goes to wherever the object currently lives. If the host moves the file, the registry gets updated and the DOI keeps working.
Different organizations issue DOIs for different kinds of things. Journal publishers typically register article DOIs through Crossref. Repositories that hold datasets and software often register theirs through DataCite. You don't normally need to know which; you need to know that the DOI exists and that you've used it.
For data, this is the step that turns "a file on a server" into something you can cite. A dataset with a DOI has a stable home and a standard citation. A dataset without one is hard to point to, and harder to give credit for. Most researchers don't request a DOI directly. They get one when they deposit in a repository, which is one more reason to pick the repository carefully (the repository guide covers how).
Two things trip people up. First, a DOI belongs to one version or one record, so check how your repository handles updates and whether it gives a separate DOI to each version and one to the whole collection. Second, once a DOI is registered, the record behind it shouldn't just vanish. Many repositories keep a "tombstone" page that explains what was withdrawn and why, which is better than a dead link.
ROR: an identifier for the institution
The third one is newer to most people. A ROR ID identifies a research organization, such as a university or a lab. It exists because affiliation strings are a mess: "Seton Hall," "Seton Hall Univ.," and "SHU" are one place spelled three ways, and a database can't always tell. Some forms and metadata schemas now ask for a ROR ID next to the institution's name. If you manage records for your campus, look yours up once and keep the ID where your team can find it.
How they fit together
Picture a dataset record in a repository. It has a DOI for the dataset, ORCID iDs for the people who made it, and a ROR ID for their institutions. A machine reading that record can tell exactly who made what and where, with no guessing. When another researcher cites the dataset, the citation points back to the same DOI, and the credit lands on the right people. That's the whole system, and it's why a README asks for the contact's ORCID and the exact citation string.
A setup checklist
- Register for an ORCID iD if you don't have one, and add a second email.
- Fill in your affiliation and add your main works. Import them from a connected source where you can, then check them by hand.
- Put your iD on your CV, website, and email signature.
- Enter it each time a journal, funder, or repository asks, so it travels with every record.
- Deposit your data in a repository that registers DOIs, and copy the DOI into your paper's data availability statement.
- Look up your institution's ROR ID and save it for forms.
- Once a year, open your ORCID record and tidy it.
Common mistakes
- Making a second ORCID iD. If you can't remember your login, use the account recovery tools instead of registering again. Duplicates split your record in two.
- Leaving the record empty. An iD with no works attached won't connect your name to anything.
- Citing a URL when a DOI exists. Use the DOI link. It's the one that keeps working.
- Treating the DOI as the whole job. A DOI points to a record, and that record still needs a title, creators, and a README that explains the files.
- Using only your work email. When you move on, the account goes with it.
If you support researchers
An identifier workshop is one of the easier things a library or research office can run, and attendance tends to be good because everyone has hit a form that asks for one. A single session can walk faculty and students through ORCID setup live, show where a DOI comes from, and send them home with the checklist above. Pairing it with a research data services launch gives your new service a concrete first event. I help campuses plan sessions like this as part of a data services engagement. [Placeholder: add whether your institution has an ORCID membership or integration, so this guide can point readers to local help.]
Common questions
What is a persistent identifier? A permanent, unique label for a person, a digital object, or an organization that doesn't change even if the web address does. ORCID iDs identify people, DOIs identify articles, datasets, and software, and ROR IDs identify institutions.
Is an ORCID iD the same as a DOI? No. An ORCID iD identifies you, the researcher. A DOI identifies a thing you produced, such as an article or a dataset. The two work together: a repository record can list your ORCID iD as the creator of a dataset that has a DOI.
Do I have to pay for an ORCID iD? No. Registering for an ORCID iD is free for individuals. You control what's on the record and who can see it.
How do I get a DOI for my dataset? You usually don't request one yourself. A repository assigns it when you deposit. Many general-purpose and discipline repositories do this automatically, and your library may be able to mint DOIs through a service it already subscribes to.