As anyone who has worked with data knows, documentation is crucial. Datasets typically cannot explain themselves--that is, one cannot interpret the data without additional information.
The most common form of data documentation is a code book. Typically these list dataset contents variable by variable, offering information for each. This information can include variable labels, which explain the nature of the variable, along with value labels (for a gender variable, explaining that "1" indicates male and "2" indicates female.) Other information, like missing value codes, is also usually provided.
|
