Hi,
as far as I saw, only the videos are different for the four groups.

So I restructured your questionnaire,
improved the design of some question (in my opinion),
added some validations
and linked to the second questionnaire automatically by end-url according the selection.

If I am wrong and there are some text variations in the four groups, you may use micro-tayloring, like
{if(Randomisierung==1,"Das ist die Frage für die erste Gruppe",if(Randomisierung==2,"Das ist die Frage für die zweite Gruppe",if(Randomisierung=,"Das ist die Frage für die dritte Gruppe","Das ist die Frage für die vierte Gruppe")))}

This way all answers to one question regardless the group are in one column of the database, which is necessary if you want to calculate some statistical tests like t-Test or ANOVAs.
As I see your aim is to find differences between the videos (with/without gestures, with/without graphics)
and the splitting variable is the "Randomisierung".

Just have a look at it.


Joffm