Yes, you are spot on!
I am indeed using an array(text) for the first question Spent. And three 5-point-choice array questions for Quality, Usage, Importance. Spent is the number of hours spent working on a product in my case.
Your "2. Questions with subquestions (multiple, multiple numerical,..) -> the text of the subquestion." is still unclear to me. Could/should I use another question type in my case? If so, which one would that be?
If not, I suppose there is no way to get rid of those [ ], or is there?!
(Btw, had figured out it refered to this x-scale just after my update and had it filled like [test], like your [Text 1], and then it started to make some more sense to me.)
Suggestion for development: if nothing sits between [ and ], then remove [] from .question result?
(You could argue that if there is a single column only [Text 1] does not add much value.)
Thanks for the tip about label sets! Will have a look at it later. (So far, I created 4 XML import files using some Excel interface.)
*** Thanks again for all your fast responses. Highly appreciated!! ***
___________________________________________
P.S. I am pretty new to LimeSurvey. Have played with it 5 days by now. I am sorry for not having mentioned array(text) specifically. Didn't understand its impact. Now, I do
P.P.S. I skipped using array(number) as it caused the survey to slow down tremendously to the point it almost became impossible to enter a number.
P.P.P.S. My initial setup contained a list of all 9 departments x 25 products which I then filtered using relevance that linked to department. The filtering took like 10 seconds in Google Chrome (so a user woull see the full list before it collapsed), but got almost completely stuck in Microsoft Edge (with multiple "not responding messages"). That's why I "invented" this approach. Besides, 9 x 25 x 4 questions would hit the (possible) limit of 1000 columns. Initially I started with 30 products. Products itself are not fixed yet, and will "depend" on the level of aggregation applied.