Dear community,
We are currently facing a challenge in a long-term survey over 4 - 6 weeks that originates from an unfortunate set of circumstances:
We need to let a few participants retake their in-progress survey because due to issues with embedded data, some conditional questions were skipped which should have been displayed.
Although this is a perfect scenario for retake links, our participants may lose their retake sessions over time after clearing browser data or by using other devices so that we are afraid of receiving multiple unintended retake sessions per participant.
Is there any way or approach that would allow our participants to retake their survey that guarantees a single new session? Because lots of questions were already answered and the new/ missing ones are scattered over different areas controlled by embedded data that was wrong, initially, but got fixed now so that the retake flow would yield the correct set of questions.
Currently, we only see two options: Letting our participants use retake links and risk multiple unfinished responses and frustrated participants or providing new individual survey links meaning that all previous answers would be lost and need to be answered again.
We also thought of using new links and manually copying and pasting the previous answers ourselves. However, the new missing questions are scattered over the survey so that we could not proceed without leaving questions unanswered again. The survey flow additionally prevents us from letting participants go back in the survey so that we ourselves also could not go back.
Any ideas or suggestions would be much appreciated. Thank you for your support!
Best regards,
Daniel
Wow, sounds complicated! Is your plan to provide a "retake" link from the Data & Analysis tab? You should be able to make a Contact List (Panel) with the links. Therefore it won't matter if they use a new device (or if their cache is clear) because they'll be starting over with the retake link, but with the data intact. You can even send them the Retake as New Response...so you have a separate response in case you still have issues with the questions.
Another option is to download the responses and create a Contact List from their responses. Then send them a NEW survey link but do it via the contact list, pulling in any (and all) answered items. Show those values and let them edit them OR, just skip the questions that you already have answers to. Can you do that anyway? Can you just have them take a "new" survey, but only ask the questions they didn't answer because of the issue? you could Show them what they answered and NOT allow them to edit those previous answers...much easier really.
Thanks for sharing your ideas, mklubeck!
We can basically try everything that is possible in our running survey. We would probably try to avoid creating a new one as only a few participants need to restart and as my colleagues would have more effort with evaluating the results.
All our survey sessions belong to individual survey links so that no anonymous links were used. So, we could very well create new individual links for our affected contacts but wouldn't their original answers in the already started sessions be lost?
Ideally, our testers could use any new link (individual, retake, ...) that has all previous answers (they can be displayed to the people again but could also be skipped) and that lets them answer the new questions that were originally hidden by accident. They should then be able to pause and resume working on their survey regardless of the device without starting new sessions if they should use another device. Because we would expect interrupted sessions due to longer waiting times in an end-to-end test.
You indicate and I read similar info in other threads, that existing answers could be piped into new survey sessions by using the contact list? Would we need to adjust all survey questions for some default answers accordingly or how would that work? As we have quite some questions, it would cause some effort to transfer all info manually.
However, if we could use an export of the affected in-progress answers, for instance, and could use these to create a contact list that has all the individual default answers based on the export, that would be the ideal solution for us, I guess.
Hi DanielDrewitz ,
I think your best option is to share the Retake Response link (not the Retake as New Response) with each respondent who needs to complete the survey again. When you share the link, carefully explain that it does not behave like the original survey link. Everything must be done on the same browser and device; they cannot start the survey in one browser or device and complete complete it on a different browser or device.
The Retake Response link only allows one completed response per link, but if you need to keep the Responses in Progress clean, do the following. As each updated response is submitted, check the Responses In Progress and delete any duplicates that were started from the same Retake Response link (though they shouldn't be able to be submitted anyway).
Best of luck!
Hi MatthewM, thanks for your suggestion.
I am afraid that we would want to avoid any danger of lost responses or the need to let participants go through the hassle of identifying the device or session that still has their most important new answer.
Because we have inexperienced people in there and also want to avoid losing participants because that would introduce huge delays. Our main goal would be minimising any efforts and risks for our participants even if that yields additional manual effort for ourselves.
However, I thought of using the retake links indeed, asking all people to finish until the original position again in one go (which is no problem), then creating new individual links and transferring all answers manually. Quite some effort, but should be working.
This way, we could share individual links again that have all answers and are not prone to different devices and such things. Because due to our long waiting times within the survey, the risk that participants try to continue with another device is quite high, unfortunately.
I definitely appreciate the emphasis on minimizing the risk of error on the respondent side. So you're staying that people who need to complete the survey again are likely to start the retake on one device and finish it on a different one?
I'll defer to your familiarity with the project and the amount of manual work that the workaround requires. My assumption was that the number of questions hidden by accident in the original version would be small compared to the number that were answered, so it wouldn't take very long for someone to complete the survey again, but I could be mistaken.
DanielDrewitz One note. If the response is "in progress" currently, you can "close" the response, saving it as if completed. Then provide the retake link, so all of the data they completed to that point would be there.
And yes, you could 1. Close the surveys, 2. Export those responses 3. create a contact list from them 4. send to those who need to retake (this is where the "new" survey comes in - but you don't have to. you could instead add "default" values to your questions = to the imported embedded data from the contact list. If there is no data, the default would be blank).
THAT all said (it will require non-trivial work on the back-end, but make it very easy for the respondent), if you only have a few that have this issue, you might instead: 1. Close the surveys, 2. click on the retake link yourself, 3. Call the respondent and ask them the questions they missed, and you type in their responses.
Just a note about closing in progress responses: make sure that your Incomplete Survey Responses setting is turned to "Record", not "Delete".
When the above setting is set to "Delete", in progress responses will be deleted by Qualtrics without any warning when closed. See Closed In-Progress Responses should NOT be Deleted when Survey Options are Delete after 1 week — Qualtrics Community
Leave a Reply
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.