Hi there,

so I did some poking around and failed to find a real solution for this specific situation. Maybe someone knows if/how this is possible:

Here is the situation:
We are in a school, so multiple classes, multiple teachers and multiple subject.
Every student is supposed to complete a survey for every subject after it ends. So the combination of student/teacher/subject is unique. However the same student may have had classes in the same subject, but with different teachers. And a teacher may teach several subject and of course multiple students - I think you get the gist...
The specific class-name uniquely identifies a class and starting-date of the course, e.g. "Test160101" would specify a test-course that started on 01-01-16.
Now we need a survey for the instance of teacher "Test-Teacher" teaching "Test-Subject" in the course "Test160101".
It's the same survey for every course, teacher and subject - but we need to be able to link it to this specific combination Course-Subject-Teacher. Not to a specific student, though! We need the survey to be anonymous for them.

Creating single students with individual mailadresses effectively kills all anonymity, so we had this idea: When a survey is needed, we create a bunch (n) of Dummy-Users with no name or mailadress but with preset values for the fields "Course", "Subject", and "Teacher" I've added to the table. We then print out the n Tokens and give them out randomly. With this the survey are linked to the course, subject and teacher, but not the individual student. However that's not how the export of tokens works...
We only get a csv with all the users. Filters only work for mailadress and such, but not for the fields we added.

Of course I can go aheand and convert the csv, filter values and so on - but this needs to be done for several hundreds of students and also by people who don't even know excel's first name.

Also the export of the results is not a bit better...

Almost forgot: The students have to take this multiple times, so with our "solution" we have to add the same students again, with changed values in our own fields. For aproximately 20-30 surveys per student on an average attendance-time and 30 students per class with 50 classes each year... Need I say more?

Well, excuse my wall of text.
Can someone tell me, if what we're trying to do is reasonable?
And if so: Please tell me how - I'm at a loss.

Now that I read that again:
I've of course thought about making it public and letting people enter their course, subject and teacher.
The problem here is: People

With this system, students can rate a specific combination several times or rate a course the did not visit.
Also if someone has a typo in - for example - the teachers name, the results are lost and can not be correctly related to the teacher when viewing the results.

Hi,
I have a little weird suggestion, ubusing the email input field.
If I understand correctly you create a specific number participants, lets say 30 for each combination of course/teacher/subject.
What if you give it a "valid" email address with this scheme: course@subject.teacher (e.g. Test160101@math.smith)
If you want to export a spefic list, just use "Filter by email address".
Sure its a workaround and only a first suggestion.
Greets
Stefan

Hi Stefan,

that could work for starters.
I've tried something similar, but your idea is a bit more to the point.
Of course I would prefer a solution to a workaround, but it's nice to have a fallback. I'll be trying this out tomorrow morning.

Like I said: I won't be the only person using that. If I where this would not be a problem - but try teaching that to your average user...

Thanks for the help!

Alex

I can't get your issue out of my mind.

Your solution should be something very user friedly. The admin page is a little bit far away to be friendly for users never use software.
I regret you need a solution apart from limesurvey for the distribution of the tokens.

If you are using a filesharing or a cloud for documents, maybe it is easier to provide a document with lets say 10000 tokens. This tokens are generated with the dummy function. Then export them once in this document.
The teachers choose how many tokens they need, insert their data (course, teacher, subject) and print it.

I don't know who will analyze the data, but it is easy to export the answers and merge them with the actual list of used tokens to get the combinations.

Now the question is, how should this document look like?
It could be a database, which you can access over web with a user interface for the teachers.

I'm familiar with excel and its visual basic, so I would make an interface in an excel document like this:


In the background is a hidden sheet/database with the tokens and empty columns for teacher, course and subjet.
By clicking "Generate tokens", the additional colums get populated with the input fields beginning with the first "empty" row.


In another sheet the tokens are displayed and can be printed out.


For the evaluation of the data just export the answers with limesurvey (e.g. to exel) and add the missing combinations with a function like VLOOKUP in excel.

Sorry for this excursion to excel, but that is what I'm dealing with. Other skills could provide other solutions.

Greets
Stefan