Please help us help you and fill where relevant:
Your LimeSurvey version: [see right hand bottom of your LimeSurvey admin screen]
Own server or LimeSurvey hosting: University of Fribourg,  Centre NTE , CH-1700 Fribourg
Survey theme/template:
==================
Dear all,

I am conducting a study in which the first I ask about the the perception of a list of countries on two dimensions - warmth and competence, then based on their rating I will select one country to pipe into the following questions. Specifically, I have a list of 5 countries and the first question is to ask about “rather warm”-“rather cold” for the 5 countries and then, in a second question, about “rather competent”-“rather less competent”. And from that, I would like to make up the matrix for each respondent that randomly pipe in one country which is (1) rated highest on both questions; (2) rated lowest on both; or (3&4) rated high in one and low in another.
Can I do that with Limesurvey?If someone has a clue that helps me resolve this, I would be very thankful.

Thank you again for your time and support & Kindest regards,
Bee

 

Please, never send lsq or lsg files.
Only lss.
Copy the survey, remove everything not relevant for your problem, export as lss.

Joffm

Thank you for your remarks. This is my first time here. I attached again the lss file. Thanks.

Fine,

now, what is your definition of 

which is (1) rated highest on both questions;

Which country is highest:  1st: 2, 2nd: 2 or 1st: 1, 2nd: 3 or 1st: 3, 2nd: 1
Of course you can calculate something like a mean and then the minimum/ maximum of the mean
and for the last 1st high. 2nd low you calculate the mean with a reversed second scale

If you don't have a clear preference what is the "highest country", (see question above) you may select at random if there are ties.

And did I understand correctly?
You want to select 4 countries
1. "high-high"
2. "low-low"
3. "high-low"
4. "low-high"

Then you select one of them at random.

​​​​​​​Joffm

I do not have strict preference what is the highest, but more about the 2x2 matrix exactly like you mentioned: high-high; low-low; high-low; low-high and then select one at random.

I do not have strict preference

Maybe you have a soft preference.
You see this is the crucial point, the definition of the "winner" or "loser".

In case of ties could one of the two scales be weighted a bit higher?

Joffm

i see your point, but i really do not have one. When it is a must, i'd choose the second scale - competent. Thank you.

Okay,
I'll try something.

Till tomorrow.

Joffm

Hi,
here is an example.
I used the assessment values to be able to calculate the high/low countries
In the first scale I used
10,20,30,40,50
To prefer the second scale a bit here the values are
8,19,30,41,52

So a rating of  1 and 2 gets a sum of 29,
but 2 and 1 gets a sum of 28. So this is selected.

The rest are some equations to calculate the max/min values of LOW/LOW, HIGH/HIGH, HIGH/LOW and LOW/HIGH
at the end one country is randomly selected and the name is displayed.

At the moment all questions of group GCalc are visible. After testing you have to hide them.
 

Joffm

By the way: You should think about your way to select countries
with this rating
 
I get
LOW/LOW: China & Philippines
HIGH/HIGH: USA
LOW/HIGH: Singapore & Germany
HIGH/LOW: China & Philippines

Well, it doesn't matter, as you select only one country at random.


 

Hi Joffm, you are a wonderful life saver. I cannot thank you enough.

Bee

Dear Joffm,

I still have a small issue, which i hope that you could help me again. I looked at the code and if i understand correctly: if the sum value for any country equals one of the four values for eqLowLow/HighHigh/LowHigh/HighLow, that country will be listed in eqCountries, then we just take a random country out of that eqCoutries. Is that right?
If so, my problem is that i cannot know which country matrix a respondent is assigned for, i.e. whether it is the case of LL/HH/LH or HL. For it is also important to see later how many responses i got in each of these four cases. Is there a way we can work around that?

Thank you again.
Bee

You are absolutely right.

But that was not the requirement in your question.

But you can insert it easily.
Don't join all countries, but separately (eqCountries1,  eqCountries2,  eqCountries3,  eqCountries4) 
Create a random number (1-4) to define which of your four categories you use.
Rest equal. Select one country at random from this category.

Joffm