Show random subsample of groups

1 month 1 week ago #192667 by lobster
Hi there,

I have a survey with 6 groups and want each participant to see randomly chosen 3 of them. I tried putting them all in the same randomization group and using a counter to limit the number shown to 3. However this counter does not behave like I would expect it to.

Anyone know a good solution on how to show a random sample of groups only?

LimeSurvey Partners
1 month 1 week ago #192670 by Joffm
Hi, lobster,

1. with javascript
create a question of type "short text" and enter this snippet
```<script type="text/javascript" charset="utf-8">

function shuffle(array) {
var currentIndex = array.length, temporaryValue, randomIndex;

// While there remain elements to shuffle...
while (0 !== currentIndex) {

// Pick a remaining element...
randomIndex = Math.floor(Math.random() * currentIndex);
currentIndex -= 1;

// And swap it with the current element.
temporaryValue = array[currentIndex];
array[currentIndex] = array[randomIndex];
array[randomIndex] = temporaryValue;
}

return array;
}

// Fill the array
var arr = [1,2,3,4,5,6];
arr = shuffle(arr);
anumbers = arr.slice(0,3).join('');
\$('#question{QID} input[type="text"]').val("#"+anumbers);
\$('#question{QID}').hide();
});
</script>```

You will get a string like

Now you can display your groups using the relevance equation:
group 1: strpos(Q1,"1")>0
group 2: strpos(Q1,"2")>0
...
Now you see the reason for the "#": To be able to distinguish between the index "0" and the result "0", meaning "not found".

2. Without javascript
There are only 20 different combination of 3 groups out of 6.
1: 1,2,3
2: 1,2,4
3: 1,2,5
4: 1,2,6
5: 1,3,4
6: 1,3,5
7: 1,3,6
8: 1,4,5
9: 1,4,6
10: 1,5,6
11: 2,3,4
12: 2,3,5
13: 2,3,6
14: 2,4,5
15: 2,4,6
16: 2,5,6
17: 3,4,5
18: 3,4,6
19: 3,5,6
20: 4,5,6

Create a random number 1-20 and display the groups accordingly.
Group 1: random<11
Group 2: (random<5) OR ((random>10) AND (random<17))
Group 3: (random==1) OR ((random>4) AND (random<8)) OR ((random>16) AND (random<20))
...

Joffm

Volunteers are not paid.
Not because they are worthless, but because they are priceless
Attachments:
The following user(s) said Thank You: gabrieljenik

1 month 1 week ago #192676 by Joffm
3. Use QCode.gseq
https://manual.limesurvey.org/Expression_Manager_-_presentation#Access_to_variables
Set the relevance equation of each group to
"G1Q1.gseq lt 4"
"G2Q1.gseq lt 4"
where GxQ1 is a question in this group which is aleays displayed.
...
As you read in the manual the count starts at "0". So you have to adapt the "4" of my example.

Joffm

I always take for granted that the groups have the same randomization group name.

Volunteers are not paid.
Not because they are worthless, but because they are priceless

1 month 1 week ago #192694 by okalmus
Hi Joffm,

thank you very much for all your help. I went with the random number method, as it seemed the most straightforward to implement right away. I will look into the other options for future reference for sure.

Thanks again.