---
name: cite-a-dataset
category: school
description: Cite a dataset with creator, title, publisher, date, version, persistent identifier, access information, and relevant subset details. Use when data supports an assignment, paper, analysis, or reproducible result.
---

# cite-a-dataset

Cite the exact version and subset used, not merely the repository home page.

## When to use

- Use for public, licensed, restricted, archived, or self-created datasets.
- Do not expose restricted record locations, credentials, personal data, or prohibited metadata in a citation.

## Procedure

1. Check the required citation style, repository guidance, license, and data-use agreement.
2. Identify creators, dataset title, collection or project, publisher or repository, publication year, and release date.
3. Record the exact version, edition, snapshot, accession, or checksum used.
4. Prefer a DOI, accession, handle, or other persistent identifier over a transient download URL.
5. Add access date when the dataset changes over time or the style requires it.
6. Describe the subset, query, table, files, dates, or filters needed to locate the analyzed data.
7. Cite related software, codebook, methods paper, or data paper separately when it contributes distinct evidence.
8. Generate the reference in the required style and verify each field against the repository record.
9. Test the identifier and ensure the citation resolves without unauthorized access.

## Worked example

A student analyzed one annual release from a frequently updated government dataset. The citation includes the agency, dataset title, release date, version, DOI, and access date. The methods section separately records the table, years, filters, and downloaded-file checksum.

## Done

- A dataset citation contains verified creator, title, repository, date, version, persistent identifier, and required access details
- A reproducibility note records the exact subset or files used and confirms the identifier resolves within permission boundaries