Hi all,

I'm working on a survey with some advanced skip logic.

I know Lime Survey currently does not support skip logic for subquestions in arrays, so I'm wondering what more experienced users might suggest as a work around.

Essentially I need to set up 1 arrays of 10 questions, each of which is answered on a 1-5 likert scale.

Then I need a follow-up array with subquestions based on answers to the previous array. I only want subquestions based on those from the previous array that were answered 1-3. In addition, I want to limit the number of subquestions in the new array to 5, drawn at random.

Suggestions?

I know Lime Survey currently does not support skip logic for subquestions in arrays...

LimeSurvey 2.06 and newer does support relevance for array sub-questions.

Thanks. I was misinformed by dated forum posts.

Do you have any recommendations on where I can find a good tutorial for Relevance equations like the ones I might need to make this happen?

Have a look at Expression Manager and associated pages in the manual - www.limesurvey.org/manual/LimeSurvey_Manual .

Assuming a question code of "Q1" and subquestion codes of "SQ001, SQ002...", the relevance for row one of Q2 would be:


Sample survey attached:

Hi Tony,

I had dug around the manual, and found this section about randomization and relevanceStatus , but I can't quite seem to get it to work.

My first group of questions Q1, Q2, and Q3 are arrays with a combined total of 16 subquestions. Each of those with values >0 and <4 then becomes a valid subquestion for Q4 and Q5, each of which are arrays, set in their own separate groups.

I need no more than 5 subquestions for both Q4 and Q5, and I need them to be drawn at random from those valid subquestions in Q1, Q2, and Q3.

I'm getting the 5 subquestion limit just fine with the relevanceStatus and LT5 techniques in that tutorial, however my subquestions aren't being drawn at random, instead they're just the first 5 valid subquestions from the combined arrays of Q1, Q2, and Q3.

Any ideas where I'm going wrong?

In otherwords, I think, is there a way to create Randomization groups for subquestions?

Did you try randomizing the sub-question order in Q4 and Q5? (in question settings)

I did. The array still chose the first 5 valid subquestions from Group 1, it just shuffled their order. I also tried assigning all of the questions in all groups to the same randomization group, but that didn't seem to have any effect either.

The only solution I've figured out is to break each subquestion in the latter groups out into its own separate array. I can make 5 relevant subquestions from group 1 then reappear as individual questions that way. It's less than ideal, but not unacceptable.