Default value (-999999999) overwritten with 0

Hi,

I encountred an issue. When I answer a question sometimes functions/variables get created (a simple example would be a gender question with values male (1) and female (2) which will for example be answered “male” which corresponds to value 1). A pre-requisite to answer this question would then be earlier in the survey a “question” if the surveyed subject agrees to be surveyed (which they can decline). If they agreed at first, and some data is filled out (like the gender, etc.) and then they “change their mind”, the question about gender (and many other questions) will disappear (since the condition → agreeing to be surveyed; was not met). However, the gender variable which by default unfilled has value -999999999 got overwritten by 1 and then when it “disappeared” got updated to 0 automatically (the value 0 is visible after making an export of this interview). What we expect to happen and should happen is that the variables “return” back to their “default” value (-999999999) and won’t get changed to 0. Is there a way how we could ensure that the variables reset to their default values (-999999999) before they were filled, as if they were never filled or attempted to be filled, without having to manually overwriting it ourselves?

Please let me know if you know how to fix this or if you have any advice for me.

Hello Michal,

If a question is categorical with values 1 and 2 defined as answer options, you should never be getting 0 in that column in the data export, regardless of what sequence of edits the interviewer or respondent does to that question.

You claim that you observe otherwise. Please create a simple questionnaire with just the questions needed to reproduce the issue, import to the demo server and demonstrate that the exported data contains an invalid code for that category after your edits (and describe exactly what they are).

If you suspect a bug in the software, please write the exact version of the software you are using.

Best, Sergiy