Data Management: Data Documentation

Documenting Data

To make sense of the data later, there should be somewhere that has

• Lists of variables names (what variable goes with what measurement)
• Definitions of variables (e.g. units of measure, exact wording of a question)
• Codebooks for variable values (e.g. Is "male" entered as M or as 1?)

Where to Put the Data Documentation

Data documentation could be stored within the same file as the data itself,  could be stored in a separate codebook, or both.

Advantages of data documentation within the data file

Specialized data analysis software often includes tools that make it easy to store this documentation as part of the data file. Storing the documentation within the file makes it easier to understand the data, because a separate codebook does not have to be consulted, and the data are hard to separate from the documentation.

Advantages of a separate code book

When the data are stored in a specialized, proprietary filetype, long-term access to the data could be hampered (see the filetypes section). A plain text readme file with the data documentation is less reliant on a particular computer system and more easily understood by a human being.

Metadata Standards

For some types of data, standards or conventions exist for what documentation should be maintained or how the data should be represented. These conventions make it easier for people working in the same discipline to understand and share data across research groups. The list below provides links to some examples of metadata standards for particular types of data. Other standards are developing for other research communities.