Hi!
I'm trying to create a survey that will randomly attribute each participant to one of four groups. Each group should then view a different stimuli (text). The rest of the survey should be the same for everyone.

As far as I can tell, the randomization function allows us to present the questions in a random order, but not to create separate groups? Is there any way to do this? I'm very unfamiliar with limesurvey, so any help would be much appreciated.

Thank you!!

You will need to load a hidden (via CSS) equation question with a random number 1-4 and then place relevance/conditions on your "stimuli" questions based on the hidden random number.

Thank you for your reply! Is there any place I can learn how to do this? A walk through, tutorial, video...? I've been trying to read the manual but it's very confusing!

I found another question about this and followed the instructions:
create a question of type "equation" (question code? let's say "randomnumber")

question text: {if(!is_empty(randomnumber),rand(1,2,3,4),randomnumber)}

Question 1, relevance equation: "randomnumber==1"
Question 2, relevance equation: "randomnumber==2"
Question 3, relevance equation: "randomnumber="
Question 4, relevance equation: "randomnumber==4"

However the randomized question doesn't show up when I preview the survey. Does anyone have any idea of what I'm doing wrong?

The equation question text should be:


- www.limesurvey.org/manual/Expression_Man...#Access_to_Functions

Thank you! makes sense! however I still dont see the randomized questions showing up. Is there something I'm missing? Something to activate in the settings maybe?

Edit: I unhid the equation question but don't see any number show up when I preview. Is that normal?

No, can you attach a small sample survey (.lss file) containing only the relevant questions?

Here is the .lss, I removed all the other questions

I replaced
{if(!is_empty(randomnumber), rand(1,4), randomnumber)}
by
{rand(1,4)}
It works!
However, I haven't hidden that question yet, and while testing, I noticed that when I saw the number that showed up didn't always match the question it was assigned to (eg I saw the number 3 but the question that showed up was assigned randomnumber==1 )
I hope this is normal?
Thank you very much for you time in any case!

I do just have a final question though, I've tried it again and though the questions do show up randomly, the number that showed up didn't always match the question it was assigned to (eg I saw the number 3 but the question that showed up was assigned randomnumber==1 ).
In that case, how can I know which question the participants saw when I do the data analysis?

It's because Expression Manager sets the equation question both on page load and submit - that's why you put the IF statement in so the equation question is only loaded if it's empty.

Having said that, you had some unnecessary HTML and a typo in the equation question.

It should be:


Working sample survey attached:

tpartner wrote: Having said that, you had some unnecessary HTML and a typo in the equation question.

It's the reason of www.limesurvey.org/manual/Question_type_...ation_.28equation.29

clean textarea without HTML editor : less chance to add broken html

Thank you so much! I had copied the equation somewhere and didn't realize. It works now!

I've realized that I've enough participants in group 2, but not enough in groups 1, 3, and 4. Is there any way to rewrite
rand(1,4)
to exclude the number 2 and only randomly assign 1, 3, and 4?

Thank you! !

Set the equation to rand(1, 3) and then change the group relevance.

Group 1: randomnumber == '1'
Group 2: randomnumber == '99'
Group 3: randomnumber == '2'
Group 4: randomnumber == '3'