The “legacy data” folder in each data set contains questions that are no longer fielded in the relevant survey.
Why do we remove questions?
We tend to remove questions for one of the following reasons:
The question is no longer deemed sufficiently relevant to today’s digital consumers and hence it’s replaced by a different question
There’s now a better or more comprehensive way to ask the question and hence it’s replaced by a newer question addressing the same topic more effectively
How should I use legacy data?
Once a question has been removed from one of our surveys, it's best to avoid using it in any analyses or audiences. That said, there are some occasions in which it can be useful to do so. If do you plan to use legacy data, be mindful of the following:
If a question has been replaced by a newer version looking at the same topic, you should avoid trending the two on a like-for-like basis. Any wave-on-wave shifts are likely to be the result of changes in question format or wording rather than a reflection of a real world change.
When building audiences, you can’t combine current and legacy questions where there are no overlapping waves using AND.
When building audiences, you can combine current and legacy questions where there are no overlapping waves using OR. However, changes in question format or wording mean that the audience may be noticeably different in size from one wave to the next.