Standard variable naming in longitudinal projects
Source:vignettes/vig_05_standard_variable_names.qmd
Under development
Objectives: - Check existing standards and principles used by, for example, the IEA, OECD and Statistics Norway.
Although there is no compulsion, there are great advantages over time if the variables are named in a systematic way so that you can later create a pattern for which variables are the same over time, which have changed somewhat but approximately the same, etc. This enables trend displays.
General advice
- Use English as far as possible
- Underline to separate words, not lower/uppercase letters. But numbers vs letters can make a difference
- Avoid years. A survey can last over the New Year. Is it the year of publication, data acquisition, or the “litter” you have indicated?!…
- Use either cycle numbering, or preferably cohort (year of birth), which is more stable. Alternatively, more information in the dataset’s metadata.