In order to make your data material anonymous, you need to go through all your data and remove or edit identifiable information in such a way that individuals no longer can be recognized.
Usually this entails:
- deleting all directly identifiable data (such as names/lists of reference numbers/scrambling key)
- deleting, rewriting or grouping together indirectly identifiable data (i.e. background variables such as residence/work place/school, age and gender)
- deleting (or editing) photos and audio or video records
If you use a data processor, the data processor must also delete all personal information connected to the project which they possess.
Please note that in most cases you can lawfully keep an anonymous data set after the end of the project, since you are no longer processing personal information. In this case, you need to make sure that you have edited the data material adequately, so that individuals can no longer be identified. However, in some cases you do have to delete all your data. This may happen if you have promised the sample to do so, or when it is a demand from the data owner (e.g. Statistics Norway).
Publications should normally present the sample anonymously. However, in some cases there may be good scientific reasons to publish personally identifiable information, but this can only be done after consent from the person(s) in question.