Hello-

I am planning on using Limesurvey for equal opportunities monitoring within a small charity. We have to ensure that this data is collected anonymously and is not able to be connected to individuals.

Due to the small size of the organisation it would be possible to identify some participants from observable characteristics (gender, age and ethnicity). If the entries are stored in a table then the other answers could be identified (which would include sexuality, religion and disability information which should be kept private).

Is there an option to only store the total number of times each answer has been given (eg 7 female, 5 male) for some or all of the questions?

Thanks

This is not possible with Limesurvey. Each interview is stored in one row of the database.

Thanks, I thought that would be the case.

I will add it to the suggested features list as I think it would be useful to some other users.

I don't think that this is very practical and thus will probably not be implemented (because I think the use case is very limited).

I am not a developer, but I think the way answers are stored would have to be changed completely and another problem is probably that you can't just write the answer to the database, but would have to make calculations with previous data.

I know that some institutions are very strict when it comes to this kind of data protection. In my opinion though, data protection is not only a technical aspect, but has to do with the process.

So if everyone is aware that they can't analyse the data on a one by one basis, then the current data model wouldn't be a big problem. You would take the data, do the crosstabs and analysis and then delete the original data. Of course, while you have the original data some "evil" person could still go into the database and eventually conect answers with people. The theoretical option is there and with the way data is stored in most survey systems this option is always given. But as I said, with the correct processes and the right people, there should be no problem even if the option exists.

But I know that this is not enough for some institutions. But I don't see any other solution with LS and I honestly don't know any other survey tool that would store results the way you want it (especially because this wouldn't allow cross tabulation, etc.).