View previous topic :: View next topic |
Author |
Message |
trent1111
Joined: 08 Nov 2015 Posts: 2
|
Posted: Fri May 06, 2016 6:44 am Post subject: Random pairings with no repeats for a whole semester |
|
|
I have classes of 25-35 university esl students and I would like to find a way to pair up students randomly--but with no repeats. Each student would get a new partner every time throughout the whole 15-week semester. Students would only be paired with someone they've already worked with when they've already worked with every other student in the class or other factors mandate repeat pairings.
So anything that would enable me to do that would not only have to be able to pair students at random but would also have to remember which students have been paired with each other to prevent repeats over the course of the whole semester. It's hard because it has to deal quirks like absent students, etc.
There are plenty of “random team generators” on the internet, but I couldn’t find any that solve the problem of repeats over the course of a whole semester. There are some complicated suggestions using Excel, but none that I could find really matched my specific needs.
Anyone know how I might do what I need to do? |
|
Back to top |
|
|
adventious
Joined: 23 Nov 2015 Posts: 237 Location: In the wide
|
Posted: Fri May 06, 2016 8:59 am Post subject: |
|
|
What a fantastically geeky challenge!
Given an end result is all will be paired, why the randomization?
Yet random is fun, so...
My first thought was a "dance card" and a hole-punch. Every student has a unique card with designation for all other cards and drawing lots for pairings yet eliminated.
Hey...wait a minute...Is this a Microsoft/Alphabet interview question? |
|
Back to top |
|
|
wangdaning
Joined: 22 Jan 2008 Posts: 3154
|
Posted: Sat May 07, 2016 3:36 am Post subject: |
|
|
You cannot answer this question unless you can fix the number of students in the class and guarantee they will all attend every lesson. Good luck with both, as I know I have little luck in this area. |
|
Back to top |
|
|
trent1111
Joined: 08 Nov 2015 Posts: 2
|
Posted: Sun May 08, 2016 3:49 am Post subject: |
|
|
wangdaning wrote: |
You cannot answer this question unless you can fix the number of students in the class and guarantee they will all attend every lesson. Good luck with both, as I know I have little luck in this area. |
Exactly! That's my major sticking point. Really? No way out? Depressing... |
|
Back to top |
|
|
wangdaning
Joined: 22 Jan 2008 Posts: 3154
|
Posted: Sun May 08, 2016 10:39 am Post subject: |
|
|
I guess you could do a spreadsheet with all possible attendees name's on the top row and in the first column then just randomly select who is with who each day and put an x in the pairing intersects. Then you will know who has and hasn't been paired. Don't think there is a good magic trick for this as the attendance and numbers vary. |
|
Back to top |
|
|
MI6agent
Joined: 16 Apr 2016 Posts: 87 Location: Dark Web
|
Posted: Sun May 08, 2016 1:35 pm Post subject: |
|
|
trent1111 wrote: |
wangdaning wrote: |
You cannot answer this question unless you can fix the number of students in the class and guarantee they will all attend every lesson. Good luck with both, as I know I have little luck in this area. |
Exactly! That's my major sticking point. Really? No way out? Depressing... |
Your problem is not easy and not straightforward. You need a good background in statistics and probability to solve it.
With 30 students you need at least 435 combinations to pair them in 2.
The rule is known as permutations in statistics;
The formula is: n!/r!(n-r)
Where
n = 30
n! = 30x29x28x27x........................x1
r =2
r! = 2x1 = 2
You might try this web link to check if you pair your students in a group of 2.
In the 'Options' box put the list of your students (first try 10 students with their surnames and see how it will work)
In the 'Quantity' box put number 2.
Then click the button 'Refresh' to create the pair of students, you need to click the box for a new pair and so on... try it and let me know.
https://www.randomlists.com/list-randomizer?items=A,B,C,D,E,F,G,H,K,L,M,N,&qty=2 |
|
Back to top |
|
|
wangdaning
Joined: 22 Jan 2008 Posts: 3154
|
Posted: Mon May 09, 2016 1:41 am Post subject: |
|
|
The question is easily answered if student numbers and attendance is constant, but it is not workable in real life. |
|
Back to top |
|
|
Mr. Leafy
Joined: 24 Apr 2012 Posts: 246 Location: North of the Wall
|
Posted: Mon Jun 20, 2016 2:48 am Post subject: Re: Random pairings with no repeats for a whole semester |
|
|
trent1111 wrote: |
I have classes of 25-35 university esl students and I would like to find a way to pair up students randomly--but with no repeats. Each student would get a new partner every time throughout the whole 15-week semester. Students would only be paired with someone they've already worked with when they've already worked with every other student in the class or other factors mandate repeat pairings. |
I saw this one presented when I worked in Japan and it's worked well for me. It does need fairly stable attendance patterns, but can handle some absences. (You'll need to look at the appendix to understand the poor written description)
Then feel nostalgic about the names in the example sheet. |
|
Back to top |
|
|
|