Situation:
Assume I have a study with about 30 questions of type array. Each question will have the exact same set of subquestions.

Problem:
A change in the set of subquestions has to be manually applied to each of the 30 questions. This is tedious but can also lead to inconsistencies.

Question:
Is there a way to maintain a single instance of a subquestion set, then "link" all question to use this?

Also:
I know about label sets, but I would still have to re-apply the label set to all 30 questions if there was a change to the set.

System:
Version 3.14.11+180926

Maybe usage of {Q1_SQ01.question} in subquestion SQ01 of Q2 and Q3 etc ?

Interesting, I'll give this a try. However, this will not work if the number of subquestions changes (i.e., one is removed or added), correct?

Yes, as Denis suggests, place the sub-question text in equation questions and reference those equation questions in the actual sub-questions.

Yes,
if the number changes, it will not work.
Therefore there as a long time request to include "dynamic lists", something like a question type "list" that does nothing than hold a list of items, and a set of functions to manipulate it (add, remove items, randomize the list, add one list to another one, and so on)

To avoid problems after adding or removing items, you may use something like
{eqItemList_1.question} with subquestion relevance "!is_empty(eqItemList_1.question)"

This way you can create questions with (let's say) 40 items to be on the save side, but only existing items are displayed.

On the other side: I see this as an academic question.
Of course, nice to have this option, but to be honest: How many times do you change your items?
I should say, once, after the pretest (maybe twice).
And to insert the set of items to your 30 questions doesn't take more than 10-15 minutes (have your items prepared in a text editor and "quick add" them, just copy and paste)

Joffm

Thanks for the additional information, Joffm.

I agree, usually there should be no more than one change. However, the organisation for which I do the content editing in LimeSurvey has a tendency of having at least three rounds of changes, often four or five. Plus, there isn't just one survey, but about 60.

You can see where the desire for a more dynamic solution comes from.

The "!is_empty()" therefore seems like a great option. I will try this out and see how it works for our purposes. The "dynamic lists", however, sound even nicer. But it sounds like I should not put my hopes up too high for that.

domsson wrote: Situation:
Assume I have a study with about 30 questions of type array. Each question will have the exact same set of subquestions.

Interesting, I'll give this a try. However, this will not work if the number of subquestions changes (i.e., one is removed or added), correct?

Unclear here …

Else : if you have to remove one , just remove the line (i don't understand)

I slightly better approach.
Use a hidden multiple question (M1) to store your items.
Set the default answers of the items you want to use to "checked" .
In your arrays now have the subquestions as
{M1_SQ001.question}
...
{M1_SQ040.question}

and use array filter "M1".

Joffm

A (partial) solution is to use label sets. So you create/change the label set and you just apply it to the questions.

Unfortunately, you still need to apply it to each question where this label set was used. So it saves a little bit of time (you don't have to change the subquestions one by one), but you still have to touch each question in each survey where you had used this label set.

Ideally a change in a label set would apply to any case where this label set was used. Today, as I understand, this is not the case. Once the label set is applied, it stays as it was applied.