To document a dataset’s source on a Mac, create a plain-text README beside the data and record the provider, original page or identifier, download date, version, license, and any changes you made. Keep that note with the project files so the source can be found again even if a Finder tag, bookmark, or download history is lost. This guide is for a Mac user working with a local project folder.

Create a source note beside the dataset

In Finder, open the project folder and choose File > New Folder (or press Shift-Command-N) if you need a folder for the original file and its documentation. Name it clearly, such as data-source, or save the note in the folder that already contains the dataset. Apple documents this Finder folder command; keeping metadata close to the data follows government guidance for recording dataset information. Organize your files in the Finder on Mac, Record information about data sets you share with others

Open TextEdit from Applications or Spotlight. Choose Format > Make Plain Text, then create a new document. Plain text is a practical choice because it can be opened without the same word-processing app and is easy to read alongside the dataset; Cornell Data Services recommends plain text for data README files. Save it in the same folder as the dataset as README.txt or README.md. If TextEdit’s save dialog proposes a rich-text extension, check the filename and format before saving; menu wording and dialogs can vary by macOS version. Writing READMEs for Research Data

Record enough detail to identify the exact source

Use a compact record like the following, replacing each prompt with what you can verify. Leave a field marked “unknown” if you cannot confirm it; do not guess.

  • Dataset title and filename:
  • Original provider or creator:
  • Source page URL or persistent identifier:
  • Version, release, or publication date:
  • Date obtained (YYYY-MM-DD):
  • License or reuse terms, as stated by the provider:
  • Coverage or contents relevant to this project:
  • Changes made to the downloaded copy:
  • Related files or documentation:

The source page URL should point to the provider’s dataset page when available, rather than only to a temporary download link. Add an identifier such as a DOI if the provider supplies one. Dataset catalog standards distinguish a dataset’s landing page, publisher, license, version, and provenance; recording these separately makes it easier to tell where a copy came from and which release it represents. Data standards: Dataset

Record the date you downloaded or received the file, not as if it were the date the dataset was created. If you can verify a provider’s release or update date, label it separately. Use the provider’s actual title and spell out the organization rather than relying on an ambiguous abbreviation. GOV.UK’s guidance recommends identifying who created a dataset, when it was created, and a distinct identifier, and advises using “unknown” when information is unavailable. Record information about data sets you share with others

Describe what happened after download

A source record should distinguish the provider’s original data from the version used in your project. State whether you renamed files, filtered rows, combined sources, converted formats, or corrected values. If the data is unchanged, say so only if you know that. For a derived file, identify the input filename and summarize the operation; link to a script or analysis note if it contains the reproducible steps. Sandia’s data documentation guidance specifically recommends naming the original source for derivative data and listing processing steps. Data Documentation – FAIR(ER) Data Sharing

For example, write “Filtered to rows dated 2022–2024; saved as sales_2022_2024.csv” only when that accurately describes your work. Do not replace the original file silently: retain an untouched copy when feasible and give transformed outputs distinct filenames. If the provider’s page lists a license or usage restrictions, record the stated wording or link to it; the note helps locate terms but does not replace checking whether your planned use is permitted.

Use Finder tags only as a shortcut

A Finder tag can help you locate the dataset or README later. Select the item in Finder, choose File > Tags, or Control-click it and choose a tag. Apple says tags can be used to find tagged items from the Finder sidebar or search; tags are labels for finding files, not a substitute for documenting provider, release, and changes in the README. Use tags to organize files on Mac

Before considering the source documented, open the README from the project folder and check that its filename matches the file you used, its URL or identifier is specific enough to retrieve the source, and each date has a clear meaning. If you cannot identify an exact version, state that limitation in the note. A short, accurate record is more useful than filling every field with assumptions.

Sources