Hi everyone,

I've been building my survey for the past couple of months and am now stuck on a critical point. 

For my survey I would like to have the following structure (fake survey topic for clarity):

G1
Q1 - do you like bananas? Yes/No
Q2 - do you like apples? Yes/No
Q3 - do you like raspberries? Yes/No
(...and so on until...)
Q10 - do you like peaches? Yes / No
(Note: I cannot make this into an array as each question includes a long description etc. so I do need these to remain separate questions.)

Then, I have 1 follow-up question group per "fruit", which contains various questions:

G2 - follow-up questions about bananas (Q1,Q2,Q3...)
G3 - follow-up questions about apples (Q1,Q2,Q3...)
G4 - follow-up questions about raspberries (Q1,Q2,Q3...)
(...and so on until...)
G11 - follow-up questions about peaches (Q1,Q2,Q3...)

I only want to ask follow-up questions about bananas to people who said they liked bananas i.e. answered "Yes" for bananas (and same for all other fruits).
I've achieved this by using a relevance equation.

MY QUESTION IS:
- I don't want the survey to be too long so I only want people to see 2 FOLLOW-UP GROUPS maximum, out of the total number of fruits they said "Yes" to.
- In other words, I want to randomly pick 2 question groups between (G2;...;G11) for people to answer, but only if they answered "Yes" to the relevant question in G1.
- For example if someone says they like A, B, and C but not D, I want them to be asked about A/B or A/C or B/C but not D.
- Of course, if people only liked 1 or 2 fruits then they should be asked about all the fruits they said yes to. If they liked none of them, skip G2-G11 altogether.
Is this even possible? Let me know if you need more details from me, and thanks in advance for any help or suggestions at all!! 

PS: I'm thinking a solution might be to "hide" any follow-up question group in excess of 2. Since all my follow-up groups are already in random order, and some may not be relevant to the participant in the first place (if they didn't click "yes" previously for that fruit), I have one idea. It would be to include a relevance equation that says something like: "if out of (Q2;...;Q11), we've already displayed 2 groups, then hide this group". But IDK if/how to do this...

UPDATE IN CASE ANYONE HAS THE SAME ISSUE:

I found out about a way to resolve my problem!!

I pretty much used the solution outlined in this article: toolsforresearch.com/limesurvey-randomly...of-y-question-groups

I just had to adapt it a bit to my own requirements as I have a LOT of different conditions, randomisations, etc. going on and interacting with one another in my survey.