Hi,

I know that tokens are not stored together with the answers (in anonymous surveys), but I thought that I could use a passthrough field to pass a token through an anonymous token-based survey to another survey (using the autoload URL feature at the end). Isn't a passthrough field something that passes a value without storing it permanently??

I started on writing a documentation page about that, then stopped and never started again. But there may be some valuable information in there for you:

docs.google.com/document/d/1J4dCuS2Egxv4...ZfQ/edit?usp=sharing

Please add your findings/questions and refer to it here (since Google docs does not sent notifications when someone has editted)

Greetings from Holland,

Tammo ter Hark

Passing the token through the survey would allow to make conclusions on who filled in the survey and thus this would be a breach of anonymity. So it is good that you can't pass the token to the next survey. So no, you can't use the token to pass through and want a survey to be annoymous.
(If this would be possible, there would be ways how you could find out who did fill in the questionnaire, thus it is not annonymous).

Thank you very much Tammo - I might just add something to your documentation page (if I come up with something that can be useful to others). Right now I am looking into different solutions so maybe I will end up with something completely different.

Also thanks to you holch. Are passthrough field values stored somewhere in the database?? I thought they just passed through the system without leaving a trace... Could you or someone else please confirm if a passthrough field value is stored or not? (and if so, where is it stored - it is not in the results)

Hi Sweden,

ik he solution as I described, they are stored in the database. That way you can glue the result tables together when analysing the data. Else you wouldn't know who entered what.

Tammo

Thanks again Tammo. I actually don't want to link the data between survey1 and survey2 because survey1 is anonymous.

That is also why I need to know if passthrough fields are stored somewhere - there isn't much information on how they work in the manual.... maybe the creator of the passthrough fields knows?

There is actuually (as far as I know) no special passthrough trick, it is just a variable that you define yourseld and that you handover to a following survey. If you create a random variable (using the Expression Manager) in survey 1, store that in survey 1 and pass that over to survey 2 and store it in survey 2, you will be able to combine the result tables while maintaining anonimity.

I think that that is what you want to accomplish.

Tammo

Thanks Tammo. Survey2 is not anonymous so I don't think I can use a random variable that is stored in both survey1 and survey2. I believe it would ruin the anonymous part of survey1, but maybe I don't understand it correctly.

Anyway, I really hope that someone can confirm that a passthrough field value is not stored somewhere. That would make me very happy

If survey 1 is anonymous and survey 2 is not, you can just start survey 2 from the end URL of survey 1. That way no one will be able to connect the results. You do not need a variable then. Just use the URL of survey 2 as the end URL of survey 1.

Tammo

Thanks, that is also what I have in mind, but I do need to pass through the token somehow. The post from holch made me wonder if passthrough field values are stored somewhere in the system. Hope someone can answer this question.

You can't have you cake and eat it: when survey 1 is anonymous and survey 2 is not and you combine the results, survey 1 will not be anonymous anymore.

There is no "magic passthru field", it is just a variable that you define yourself. Can be a random variable.

Only when both are anonymous, the combined result can be anonymous. You can combine them with a random variable. For combining two tables, the combining variable must be stored in both tables. Otherwise there is no way of finding out how to combine.

If survey 1 needs to be completely anonymous, then working with tokens will also be difficult, because for tokens you need to register an email address (OK, there are trick circumventing that...).

Maybe I misunderstand you, but without more information I cannot see it otherwise..

I do think you misunderstood me, but it is my fault - I could have been more clear. I don't need to combine the result from the two surveys. In fact, I want to prevent that they can be combined. That is why I want to know if the passthrough field leaves a trace (e.g. stored somewhere in the database). Most program can store values temporary without leaving traces (even without using magic!), so I hope it truely is a passthrough field.

Yes, tokens can actually be created real easy without emails and that part is working as it should (it was easy to export and import tokens so they are identical).

I think I will get my cake in the end, but it might take some time before I can eat it Thanks again for your time Tammo - much appreciated.