Check out the LimeSurvey source code on GitHub!

Limit of Participants and shown Question Groups

3 years 9 months ago #94942 by AndreasJanson
Hi Folks,

i have two questions with regards to the question groups.

1) My survey has for example 4 big question groups.

2) Each question group is linked with conditions, which refer to the experience of the participants with each question group.

3) The 4 big question groups itself are randomized.

My questions are:

Is it possible with LimeSurvey to only show one of the four randomized groups to the participants? Regardless, if the participant is experienced with all of the four research subjects. E.g., if one participant is knowledged with question group XY and YX, only one of the two groups (at random) is shown to the participant.

And furthermore, is it possible to limit the number of participants to each of the question groups? E.g., if 40 participants answered the question group XY, question group XY is not longer displayed to the participants?

Thanks in advance for helpful comments :)

If my problem is not clear, do not hesitate to ask.


PS: Version 2.00+ Build 130406

Please Log in to join the conversation.

3 years 9 months ago #94950 by orvil
Hi Andreas,

to understand your problem: there is one question - lets say Q1 - with four items at the beginning (within a first question group I suppose). Then there are four more groups with questions refering to these items of Q1.
If one or more items of Q1 are marked, (only) one refering group has to be displayed.

I think this can be done via EM.
Lets say Q1 has 4 items SQ001, SQ002 ... to be marked or not e.g. as a multiple choice or matrix question with experience ratings like very good =1, good=2, fairly=3 etc.

Step 1: create a first question of the type 'Equation' EQ1, to sum the number of marked items. The equation code to sum the number of fiting items will be
if you want to accept only 'very good' answers in a matrix type Q1 or in a multiple coice Q1. If you want to accept 'very good' and 'good' use
{if(Q1_SQ001==1 OR Q1_SQ001==2,1,0)+ ...

Step 2: create a second question EQ2 of type 'Equation', to create a random number between 1 and the number of chosen items

Step 3: create a third question of type 'Equation', name it EQ3, to create a string containing a number list of the chosen items
Please notice: there is no plus '+' between the if-statements, no space. This expression will construct a string like '34' if items three and four has been chosen. Or '1234' if all items are marked.

Step 4: now we have to select the character at the random position with a next equation. Create once more an equation type question EQ4 and use the formula:
this will select one digit at the position defined by the random number of EQ2. Eg if the selected items are 2,3,4 and the random number of EQ2 is 2 the resulting digit is '3'

At last you will have to edit the group validation according to the number:
For goup one the validation has to be EQ4==1
for group two the validation has to be EQ4==2
and so on

Good luck!

Best regards/Beste Grüße,
O. Villani

Please Log in to join the conversation.

3 years 9 months ago #94953 by AndreasJanson
Thanks Orvil!

Your thoughts are very helpful and go in the right direction... I guess ;)

The filter questions in front of the four big question groups (research objects) are not that complicated. They only ask, whether a participant is experienced with the research object with possible options to answer yes or no. If they answer yes for each object, there is a possibility to get specific question with regards to the research object (at random).

Though, it is quite hard for me to understand how i implement your equation questions, yet. But, i will now try to adapt it to my survey :)

However, my second question remains still unclear: Is there a possibilty to limit the max number of participants to each question group? Goal is to have after the survey a (min/max) number of participants for each research object. And i think there are some research objects that are more "popular" than others... in order to have respondents for each research object.

Thanks in advance,


Please Log in to join the conversation.

3 years 9 months ago - 3 years 9 months ago #94992 by orvil
Hi Andreas,

it's not so complicated but anyway I added a sample for you.

Btw. I had to do some recoding (I've used options instead of multiple choice so far). Here for all other visitors, who do not want to test the attached sample:

File Attachment:

File Name: limesurvey...mple.lss
File Size:22 KB

EQ1 is:
One has to use 'Y' instead of 1 for multiple choice questions.

Equivalent EQ3 is:
I forgot the curly brackets inside

And the validations are also working with quotation marks, like EQ4 == "1"

@your second question:
My only idea is to use quotas. I'm not sure if there is a way to access quota values by the EM.
Sorry for that.

P.S.: the equations are visible in the example file. You will have to hide them (advanced settings->Display->Always hide this question = 'Yes'

Best regards/Beste Grüße,
O. Villani
Last Edit: 3 years 9 months ago by orvil. Reason: addad some PS

Please Log in to join the conversation.

Imprint                   Privacy policy         General Terms & Conditions         Revocation information and revocation form