Field note
Which questions do these notes answer?
Every published note about profiling a dataset, writing field documentation, recording missing-value and allowed-value semantics, reviewing with real readers, freezing editions, and handing a dictionary on. The index below links all of them.
Every published note
- What belongs in a data dictionary?Which details should each field include?
- Data dictionary versus codebookIs a data dictionary the same as a codebook?
- Document missing-value codesWhy should missing values have separate meanings?
- Document allowed values and labelsHow should categorical codes be documented?
- Write clearer field definitionsWhat makes a field definition useful?
- Document measurement unitsWhen should a data dictionary include units?
- Describe dataset population and universeWhat does one record represent?
- Version and maintain a data dictionaryHow should a data dictionary change over time, and how do I keep it from going stale?
- Test whether a reader understands a fieldHow can I review data-dictionary clarity?
- Offline data-dictionary workspaceCan I document a dataset without an account or upload?
- Create a dictionary from CSV structureWhat can be observed from a CSV safely?
- Turn JSON Schema into reader documentationCan JSON Schema start a data dictionary?
- Passive data-dictionary handoff packetWhat should a portable dictionary packet contain?
- Why discard source rows after profiling?Why does the studio avoid saving imported records?
- Data dictionaries for research handoffWhat should a handoff dictionary explain to a researcher or a receiving analyst?
- Observed versus declared data typesIs a detected type the field's true type?
- Lint a data dictionaryWhat can deterministic documentation linting find?
- Use safe examples in field documentationShould a data dictionary include example values?
- Track field owner and review statusWhy record ownership and status?
- Back up a local data-dictionary projectHow do I protect work saved in the browser?
- Spreadsheet-safe dictionary CSVHow should CSV exports handle formula-like cells?
Policy and product pages
- How to create a useful data dictionaryWhat is the complete workflow for creating a data dictionary readers can use?
- End to end worked exampleHow does the whole studio work on one real dataset?
- Codebookery FAQWhat should I know before using the studio?
- Product limitationsWhat does Codebookery not do?
- Accessibility statementHow is the studio designed for access?
- Security and local-data modelWhere is project information stored?
- Terms of useWhat terms apply when you use Codebookery?
- SupportHow can I get support?
- LicenseMay I redistribute the studio?
- About the productWho provides Codebookery?
- Local data privacy boundaryDoes local-first mean private in every circumstance?
Every published note
The notes below cover the whole workflow: reading structure from a file, writing definitions readers can act on, recording what missing values and allowed codes mean, testing comprehension with a real reader, freezing editions, and handing a dictionary on. Each answers one question and links to the station in the studio where that work is done. Together they are the reasoning behind the product rather than a feature list.
Policy and product pages
The policy pages are kept separate because they answer different questions: what the studio does not do, where project content is stored, what the local-first boundary covers, the accessibility position, the terms, the licence, and how to reach support. Each fact is stated once, in the page that owns it, rather than repeated across the notes.
Where to start
If you are new, read the complete workflow first, then follow the end to end worked example, which runs the whole studio on the fictional Northbank tree survey and hands you the file it was built from. After that, read the note for whichever part of your own dataset is hardest, which is usually missing values or allowed codes.