Is there a way to "freeze" a non-prefilled answer/prevent the interviewer from changing their answers?

Hi there!

I have a questionnaire in which respondents can be monetarily/materially compensated depending on their answer to a question. The issue is that I do not want the interviewer to be able to go back to that question and change the answer to it. Is there any way to make that the answer to a question in Survey Solutions cannot be changed? So, in other words, the question can only be answered once?

I found a similar topic to mine here in the forum, but it was never answered: Name, Gender, Age should be freezed. I also saw that other survey softwares have this feature under the name of “forward-only survey”.

Thanks a lot!

No, there is no way to prevent interviewers to changes answers. The whole purpose of a survey is that interviewers are the ones recording answers coming from a respondent, and interviewers can make a mistake - mistype something, mishear something, respondent changed their answer after thinking about it a bit, etc etc.
Although I (only kinda) understand the need for some answers to be locked, we haven’t so far come up with a robust way to figure out when is answer ‘final’ and can be locked…

1 Like

Adding on Zurab’s answer:

As some sort of ‘work-around’ you can use the paradata to investigate if interviewers do indeed change the answer to a question. It will be too late for the particular interview itself, but you can discuss with respective interviewers before they continue with fieldwork.

1 Like

Thank you very much for both answers! I think Peter’s work-around is already good enough for us!!

Catching up on the “too late to do something” idea. Depending on what questions you ask. If the rewarded answer is not obvious to the interviewer himself, then you can reward the respondents later once the answers are sent to the server. As long as the interviewers don’t know what answers you reward or what questions matter.

Clearly, if the question is something like “How much is 2 plus 3?” that would not be satisfying this requirement. Something like “Are there any three positive integers a, b, and c satisfy the equation a^n + b^n = c^n for any integer value of n greater than 2?”

If you pursue this way, your questionnaire may be programmed to display a certain code at the end of the interview, based on which the respondent can claim (let’s say a month later) a reward. Or you don’t display anything immediately, and just follow up later with the “winning” respondents using the contact information you acquired.

Perhaps there are other possibilities, but we know so little about your project. If this is a WB project involving rewarding the participants of the study, please make sure you run it thoroughly by the ethics committee. If this is not a WB project, it is still a good thing to do.

1 Like