I would like to implement a complex question sequencing (~routing) mechanism, to ensure a better scientific quality of my work.
My ideal data result would be as follows:
|- 10 Task in each process)
|- 4 Question for each Task (from a set of 80), some being common to many Tasks
The 80 Questions all need to be answered twice (to prevent cheating), in a random order (pseudo-random is OK)... and overall, the 5x10 Tasks should all be covered by 2-3 relating answers at least.
What would be possible options to do this using LimeSurvey?
Do you have any alternative tool (preferably opensource) to recommend?
I thought about the following options, please let me know if any sounds easy/impossible to you!
option A) I create 50 groups (5x10), and put manually a subset of the 80 questions in each group, with randomisation on questions order. (downside: lots of preparatory work; upside: limesurvey handles it already)