Data Curator

A data curator is an institutional role responsible for preparing, describing, and maintaining research data so it can be deposited, found, and reused after a project ends. The role is typically embedded in a university library, a dedicated research data service, or a research-computing unit, and it sits alongside — not in place of — the researcher who actually owns and generates the data. This entry covers the role — the person and where they sit in an institution; for the underlying activity itself (the metadata, cleaning, and preservation work), see data curation.

What a data curator does

Across published position descriptions and the research-data-management literature, four tasks recur consistently:

• Metadata quality. Selecting an appropriate metadata schema for the discipline (Dublin Core, DataCite, or a domain-specific schema), completing required fields, checking values against controlled vocabularies, and fixing gaps — missing units, ambiguous variable names, absent provenance — that would otherwise make a dataset hard to find or reuse correctly. See how to choose a metadata schema for a dataset.
• FAIR and deposit readiness. Checking a dataset against the FAIR (Findable, Accessible, Interoperable, Reusable) principles before deposit, and advising on which repository fits the discipline, licensing needs, and any funder mandate — see the CASRAI comparison of FAIR vs. CARE data principles.
• DMP and lifecycle advice. Helping researchers plan for curation and long-term reuse while a data management plan (DMP) is written, not only after data collection is finished — consistent with the Digital Curation Centre (DCC)‘s framing of digital curation as a full lifecycle (appraise/select, ingest, preservation action, access/reuse), not a single point-in-time task.
• Format and preservation checks. Verifying files are in a non-proprietary or well-documented format, checking that documentation is sufficient for someone outside the original team to use the data correctly, and flagging preservation risk (e.g., formats or software dependencies that are likely to go obsolete).

This is a distinct thing from the “Data curation” role in the CRediT Contributor Roles Taxonomy (ANSI/NISO Z39.104-2022), which credits a specific contribution to a specific published output (“management activities to annotate … scrub data and maintain research data … for initial use and later re-use”) rather than naming an ongoing institutional job. A person holding the institutional data curator role will often also receive CRediT “Data curation” credit as a co-author on a resulting data paper — but the two are not the same thing, and one can exist without the other.

Data curator vs. data steward vs. data manager

Be skeptical of any source that draws a clean, universal line between these three titles — the honest answer is that there isn’t one. Research-data-management job postings use “data curator,” “data steward,” “data manager,” and “data librarian” for functionally identical positions at some institutions and for genuinely different ones at others, with responsibilities varying substantially even where the title matches (see CASRAI’s guide on the data steward and data curator roles for the fuller discussion of that ambiguity).

Where an institution does use the terms deliberately rather than interchangeably, two patterns show up in practice:

• Curation as the later-lifecycle, hands-on end. “Curator” is used for the metadata, format, and preservation work described above, closest to deposit and reuse; “steward” is used more broadly for accountability across the full data lifecycle, including liaising with ethics/IRB review and keeping the DMP itself current — see data steward role (in DMP) for the narrower, DMP-contributor-metadata sense of “steward” used in the RDA DMP Common Standard.
• “Manager” as the broadest, most generalist label. Some European RDM programs deliberately adopted “data steward” because it reads as less discipline-specific than “librarian,” “curator,” or “scientist” — useful for one person supporting many disciplines with different data practices. “Data manager” is sometimes used the same way, or reserved for administrative/compliance oversight of a whole DMP portfolio rather than hands-on curation work.

In practice, read the actual position description rather than relying on the job title alone — this is standard advice from the library/RDM literature on the topic, not a CASRAI-specific judgment call.

Qualifications and reporting line

Advertised data curator positions most commonly require an ALA-accredited MLIS/MLS (or an equivalent subject master’s degree), demonstrated experience with metadata standards and schemas relevant to the discipline (Dublin Core, DataCite, or geospatial/domain-specific schemas), and familiarity with FAIR and, where applicable, CARE data principles. The role most often reports into a library’s research/scholarly data services unit (sometimes titled a Head of Data Services or equivalent), though at institutions with a centralized research-computing or research-office model it can instead sit there. It is common for the position to combine data curation with an adjacent specialism — GIS, a particular discipline’s data formats, or repository administration — rather than existing as a fully generic, discipline-agnostic role.

References

• ANSI/NISO Z39.104-2022, CRediT — Contributor Roles Taxonomy, “Data curation” role definition (credit.niso.org/contributor-roles/data-curation).
• Digital Curation Centre, “What is digital curation?” (dcc.ac.uk/about/digital-curation) — curation lifecycle model.
• CASRAI, Data Stewardship and the Data Curator Role in Research — fuller discussion of role-title ambiguity across RDM job postings.