I have a large survey, split into several different jobs in the export job section of SSI Sawtooth. For one job only, section j of my survey, several key variables have completely different values for every respondent after they are exported vs. how they appear in the view/edit data window.
In the view/edit data window all of the values appear correct, with an edit status of either 2 or 3. However after downloading the data in SPSS format (and using Stat/Transfer software to convert the data into a STATA file) the edit values for each respondent are suddenly completely different and seemingly random with values that are completely incorrect and out of the range of possible values that we even use, 12, 5, 6, 4, 7, 8 etc.
This same issue occurs for another key variable, sys-endtime. Completely different times appear in sys-endtime for this job compared to every other job.
Any ideas what might be causing this issue or possible ways that I could go about fixing it? When I go and download a job with just the identifying variable and sys-endtime and edit and no other variables, all of the edit values are fine. I was thinking about trying to use this and do a merge as a workaround solution, but I need sys-endtime to match on and it is incorrect.
Thanks very much,