- Save the syntax for any derived variables
- Always save any derived variables to the dataset once you are happy with them so that you don't have to keep creating them.
- Develop a set of conventions for naming variables such as Age5 - age in 5 bands, Age7 - age in 7 bands etc.
- Develop a set of conventions for naming your files so that you can quickly find the syntax for the analysis you are looking for - for example adding a number at the end to show which the latest version is.
- Always switch the weights off before saving the data file.
- Always give the datafile a new name so as not to permanently overwrite the previous file.
- In handling very large dataset you may want to break the file up into smaller units - for this you need to think about what variables will be pertinent for the analysis and what classificatory variables you need.
