Field notes
Writing that is not a project: arguments, close readings, and stories built from a dataset.
Projects have a question and a point at which they stop. Field notes do not. They are the pieces written because something was worth explaining, worth arguing, or worth showing from the data, and they are kept apart from the project record so that neither borrows credibility from the other.
The three kinds below are distinguished by what they claim rather than by length. An article argues a position and may reason past the evidence as long as it says so. An explainer takes one document, dataset, or convention and leaves a reader able to read the next one. A data story starts from a dataset and reports what the analysis found, including when that is nothing.
This section is new and thin. It is published in that state for the same reason unfinished research is: a category that exists and is visibly almost empty is more honest than one backfilled to look established.
Articles
Longer pieces that argue a position about public-interest data work rather than report a result.
Data stories
Pieces built around a specific dataset, where the analysis is the argument and the limits are part of the story.
Explainers
Pieces that take one method, dataset, or document and show a reader how to read it.