Hello,

What I will be asking you was most probably asked many times, and it might have been the easiest procedure on LimeSurvey, but since I am a little bit nervous and out of time, I could not find my answer and was too anxious to check the forum more, so many apologies to those who will see this kind of post here again and again.

I do not have any experience in coding nor do I know any program, so if you consider my lack of knowledge while answering, I will be more than grateful.

I designed a survey consisting of 8 lists (groups) each of which is assigned to participants randomly (I managed to form a relevance equation by reading the forum and fortunately it works).
I have already collected data from 103 people, but it seems that the relevance equation only appoints people to different lists and does not keep a track of the number of people across the lists, which is a big problem for my research.

I need equal sizes of people in each list (25ish or 30ish, so I'll continue to collect data to reach this) by still assigning participants to lists randomly.

Is there any way to do this by writing a code? If yes, would you mind explaining it to me or if a post explaining this clearly exists, could you help me find it? I have seen some posts, but they included some Java explanations, which is beyond my knowledge, I must say.

This code can also put a number limit to lists like when 25 people have done a list, that list won't be able to be accessed anymore, or it may be a code that assigns an equal number of participants to each list randomly if possible.

I hope there is some way I can figure this out. Thank you so much for your time and help in advance. I hope my questions won't be burdensome at all.

Have a nice day!

Hi,
of course you cannot expect an even distribution. You know about Normal distribution resp. Gauss distribution?
As far as I understand: Your survey is running already.
So you cannot change anything in the structure.

What you can do:
1. Wait until you have your goal of 25 respondents in all groups.
It's usual that in some groups you will get an overquote.
Here you have the chance to remove bad quality responses (happy clicker, speedster, ...)

2. If one groups is full, change the relevance of that group to "0" and add this former relevance to the least filled group.
Example:
Initially:
Group 1: random==1
Group 2: random==2
Group 3: random=
Group 4: random==4

Now, let's say: There are 25 answers in group2, group 4 is least filled (only 21)
Group 1: random==1
Group 2: 0
Group 3: random=
Group 4: random==2 OR random==4

Meaning you have to check permanently.

Joffm

Hi,

Thank you so much for your response. It means a lot.
What you explained makes sense and I think it is my only way out, so I'll follow this procedure. However, I have another concern and so a question in mind regarding the relevance.
While I was designing this survey, at the beginning I wrote the codes under group relevance so that each question of that group has the same equation as exemplified below and in the picture.

Group 1: ((random==1))
Group 2: ((random==2))
.
.
Group 8: ((random==)


Yet, later I realized each question also has a part where you can put equations and as I did not know the difference between them, to make sure that my groups could be distributed randomly, I also wrote the group relevance under relevance equation part of each question as below.



Now what you told me above is to change the group relevance to zero, but now I am curious about whether I should also change the relevance equation part of each question in that group to zero. If I don't, would that be a problem?

Thank you so so much for your interest. I hope you could have time to answer this question of mine as well.

Hi,
if there is a relevance equation on the entire group you don't need to set the same relevance equation to the questions.

The reason for both is:
Show Group 1 only to men (group relevance equation "Q1=="man")
Show Group 2 only to women (group relevance equation "Q1=="female")

But now inside group 2
question 2.1 is only to young women (question relevance equation "Q2=="young")
question 2.2 is only to old women (question relevance equation "Q2=="old")

Joffm

By the way:
I think your random number is an integer. So no quotes needed in the group relevance equation, just "random==1"

I see and now understand the difference between them clearly. I will also delete those quotes. Thank you so much.

I think I could not express myself enough, but my concern is more like this: let's say my group 2 reached enough number of people, so I will deactivate it. I will write O under group equation and all the questions in Group 2 will be deactivated, which is what I wanted. However, I did not know what you said at the beginning, so I also set the same relevance equation for each question (in total: 128 questions) in Group 2. Do I have to change all those equations to zero as well? Or changing only group equation to zero but keeping the relevance equation as ((random==2)) under each question would be enough? I am afraid there will be a clash or something and deleting them from each question will also be tiring. What would you suggest?

Thank you so much again,
yk