They may have forgotten to publish.  They can look under Tools → Versions → View version history to see when the survey was published and compare that date/time against the start date/times in the response data.

Since you copied the latest correct version, your results are correct. It doesn’t tell you anything other than the latest version is correct.

Thanks -- that may be possible, but just to be clear, I didn’t copy “the latest correct version” -- I copied the version that had generated the issues (as far as I know), as no corrections were (consciously) made to the original survey after this problem was noticed.  But, again, I wouldn’t rule out the “publishing” issue, especially if no other alternative is plausible.

Thanks again. 

What you copied was the current version. It may or may not have been published (the likelihood is that it wasn’t published).

That’s my point: I copied the “current” version (which, as far as I know, is still having outputting problems, possibly due to not being published -- unfortunately, the history is less than revealing, since it mentions  “Survey flow changes saved” over a month ago as “changes since the previous version” and the Survey Flow appears to include the correction condition codes), but when I copied it, the new version (which I did not publish) did NOT have the same problem -- it outputted the condition codes correctly (whereas the one I copied from had not).  Perhaps copying a survey automatically incorporates changes made to the original, even if those changes weren’t originally published?

In other words, maybe there’s a previous, “published” version of my original survey that has the wrong embedded conditions codes (I don’t see any evidence of that, as the Survey Flows looks correct in the previous version history, but maybe).  At some point the incorrect survey got correctly revised, without being published (and thus failed to output the correct condition codes), and that’s what I copied over in to a new survey today, again without publishing the new survey -- except that that new survey works fine (it has the correct condition codes).  That all sounds possible -- it’s just a bit surprising.

Thanks again.