Data Documentation | Nineteenth-Century Data Collective

Examine the dataset closely to understand what it is, how the files interrelate, and what information is needed to reuse.

Check for quality assurance and usability issues such as missing data, ambiguous headings, code execution failures, and data presentation concerns
Try to detect and extract any “hidden documentation” inherent to the data files that may facilitate reuse or expose unintended information
Determine if the documentation of the data is sufficient for a user with similar qualifications to the researcher’s to understand and reuse the data. If not, create additional documentation (e.g., a data dictionary)

Key Ethical Considerations

If working with human data, is this research done with and not on communities and populations involved?
- Are there authoritative group representatives who should be contacted?
Are there labels or other descriptive indicators that could be applied to better represent or protect an identified group of people impacted by this dataset? (Example: TK labels)

Acceptable formats include txt, rtf, or mdl. For more information about READMEs, see Cornell University’s Guide to writing “readme” style metadata.

Include complete documentation about your dataset
- Provenance Information
- Licensing Information
- Descriptions of Study Level and/or Data Level information:
  - Study-level data documentation:
    - Provenance of information gathered in this dataset
    - Methods used to acquire, organize, or normalize this data
    - Descriptions of any transformations that this data has undergone
  - Data-level documentation:
    - Contents of data files
    - Definitions of all variables, labels, descriptions, units of measurement used in data files
    - Defined labels or codes used to indicate missing data (ie no blank spaces, but what would indicate unknown or missing)
Include participant information and codebooks (if applicable)
Check the accessibility of all files
- Each file can be opened without proprietary software
- There are robust descriptions in plain text of data files and any images.

How are your files (and folders) named?

Are names consistent?
Are the names meaningful?
Have you documented each of the file (and folder) names?
Do the file and folder names reflect versioning? (i.e. updated files are named in a way that is findable without opening up the files themselves)