Hi everyone. I seem to always struggle with this issue but I actually think Qualtrics is broken. I'm trying to create 3 different end of survey messages for a survey to help us identify the right pool of people for product research.
I have the following survey flow which should work according to this KB. 
The problem is every time I test the survey, it's not working and keeps using the default end of survey message.
SO THEN I updated the end of survey message with a customised message and when I tested it, it showed the Qualtrics default end of survey message, not the new one. Is this a problem I can solve or is it definitely a Qualtrics issue cause I'm going in circles and it's very frustrating. 
Thanks in advance!
Different customised end of survey messages for different scenarios
Userlevel 3
+7
Userlevel 7
+30
Are there other parts of the Survey Flow not shown in your screen shot?
Have you tried testing this with just one End of Survey branch set up? I would suggest removing the first and the third one, and see how it behaves when you answer the survey in a way that should trigger the second EOS message. The reason I suggest removing the first EOS branch is because having conditions based on choices Not Selected sometimes doesn't work the way you expect it to.
Also, not sure if this is just a meaningless coincidence, but your Survey Flow looks different than mine. I have arrows between the objects in my Survey Flow. I don't know why they would be different but perhaps that's something Qualtrics Support could explain.
Userlevel 7
+37
I'm going to agree with MatthewM above re: testing with one option or with triggering only the second set of options. Your first set of criteria are "negative" (Is not Selected) and there are multiple of them. The combination of those two may be upsetting your logic flow from ever getting to that first option.
- Is it possible to break those two "Is Not Selected" options apart so that only one is being checked (even if they both have the same end of survey behavior attached to them?
- Is it possible to flag people with an Embedded data fields based on the same logic and then use that criterion to pass people down the respective paths? (Then you also have a data field to reconfirm where the system thought they should be going?)
- If even without any custom end-of-survey steps, you can't get the survey to save your end of survey behavior, I'd suggest (and I know this seems dumb) trying again after having been out of the system for a while (I know that sounds like a dumb answer, but we have an open issue here where Qualtrics isn't saving survey edits after a certain amount of edits have been applied, and it LOOKS like it's saving, but isn't, and we are still trying to figure out why our company network and Qualtrics are suddenly fighting each other. So... just ruling that kind of weirdness out.)
- Finally if you have ruled out the double-negatives and it's definitely not that the changes aren't saving, but you still aren't able to get any end-of-survey changes to be recognized, then maybe see if the Support team can help troubleshoot?
Userlevel 3
+7
Heya,
I think it's having the double negatives that's breaking the system. I removed everything in the survey flow that I was trying to do and thought I'd try something with skip logics and display logics in the survey itself which required "display if x or x is not selected" and it would show it even if you select x.
So I removed one of the negatives and it actually worked. HOWEVER, I need to display if another x isn't selected in a second question and both of those questions should lead to a "sorry..." message. So I've copied the same "sorry" message into a new question and did the "display if x is not selected" but again having two negatives isn't working. Which makes no sense cause it worked before I started playing around with the survey flow! I'm just getting angry at it now. Is this a Qualtrics bug, because it's such a waste of time to have to play around with these things.
Userlevel 7
+37
If one negative works, but two of them don't work, maybe just Duplicate the step that is checking TWO items, then remove one of the negatives from one, and the other negative from the other? (Leaving everything under it the same so you don't have to reset everything, just use the same end of survey.)
Then either:
- They both work because they are not combined or
- One works and the other doesn't, and you know there's something weird about whichever one is failing.
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.