Your quality improvement project uses a state discharge file and a national statistics page, and you cannot tell which needs a dataset citation and which needs a web page citation. Get it wrong, and readers cannot find your data. This guide gives the APA 7 dataset pattern and a worked example first, then covers statistics pages, registries and restricted-access data.
The version number, when there is one, sits in parentheses after the title and is not italic. The bracketed description tells readers the item is data rather than a document. Include the years the data cover in the title when the source does, because the release year and the data year often differ.
About the examples. Agencies, titles, versions and links below are fictional placeholders and do not match any real dataset. Use the details from the data you obtained, and check the source's documentation for a suggested citation, which you can adapt to APA.
The citation depends on what you actually touched. Decide first, then format.
| What you used | How to cite it |
|---|---|
| A downloadable file from an agency or repository | Dataset reference with [Data set] after the title |
| A number read from a statistics page or report table | Web page or report reference, with the table or page in the in-text citation |
| A registry or dashboard that updates | Dataset or web reference with a retrieval date |
| Data you collected yourself | Describe it in the methods section; cite a repository record only if you deposited it |
| Data in the supplement of a journal article | Cite the article |
Ask for your data sources to be described, cited and presented properly. You see the price before you pay, and revisions are free for 14 days after delivery.
Agencies such as the CDC, the WHO and CMS publish statistics as pages, reports and downloadable files, and the format you cite must match the version you used. The agency is the author. If a page shows a publication or last-updated date, use it; otherwise use "n.d." Do not reproduce a statistic without saying where it came from, and if a table gives the number, cite the table.
Keep primary data and summaries separate in your writing. If you analyzed a downloaded file yourself, cite the dataset. If you are quoting the agency's own published summary, cite that summary.
The methods section resource and the research methods guide explain how to describe data sources and how they were handled.
Numbered styles usually treat a dataset as an online source, so a fictional agency file would take these forms, adjusted to your manual.
Check whether your program's guide has a specific dataset pattern; when it does, use that. See also the guides to AMA and Vancouver.
| Mistake | Fix |
|---|---|
| Citing only the agency's home page | Cite the exact dataset, table or page |
| Leaving out the version number | Record it, since data files are revised |
| Using the download date as the publication year | Use the release year; put the download date only in a retrieval statement when needed |
| Treating your own project data as a published source | Describe it in the methods unless it sits in a repository |
| Reproducing a statistic with no source | Attach a citation to every figure |
Academic integrity. Show where every figure comes from, and honor both your institution's academic-integrity requirements and any terms attached to the data. For official examples, see the APA Style website.
Yes, cite wherever the statistic came from: the dataset if you computed it, or the agency page or table if you read it there.
Use it as a source of the correct details, then reformat it into APA so the whole reference list matches.
Describe collection, sample and handling in the methods section. A citation is needed only if the data are deposited in a repository that provides one.
Add one for data or pages designed to change, such as live dashboards, when no archived version exists. A fixed, versioned dataset does not need it.