OK, so I have imported your LSS file and had a look at the quesitons in the second group, because the error starts from there. The second group should only be shown if they work in Belgium, if they are 18 years or older and if they agree to participate.
The crazy part: Instead of solving this with a simple "quota" to screen people out that do not match the profile to continue, someone thought it would be a good idea to write a condition in ALL of the questions. Not even in all following groups, which would have at least be a lot quicker (only 14 conditions and not for each and every question after the first 3).
So in my opinion people jump to the end of the survey, because the conditions to show any questions from group 2 onwards are not met. Simple as that. So either they are not 18 yet, or they don't work in Belgium, or they do not agree to participate or 2 or 3 of these conditions.
But to be honest, this whole survey is messed up in my opinion.
Code:
((CF01.NAOK == "Y") and (CF02.NAOK == "Y") and (CF03_1.NAOK == "Y"))
I didn't look at every single questions, but I have looked at a few and they all have the condition above, so I can see a pattern here.
I also don't understand why CF03 is a multiple answer question and not a "list (radio)" because people should only be able to mark one of the two answer options. Instead of setting a maximum answer of 1, a list radio would have done this. Maybe there is a reasoning behind it, but I don't see why a list radio should not work. Because the way it is now, you can tick both answers for CF03 at first, only to later get the error message that you can only choose one. Not the greatest usability.
So I tested the survey and not surprisingly, if you either say "no" to working in Belgium or being older than 18 (why not directly ask the age brakets here and avoid to have a second question about age is beyond me, but that is another issue) or saying you don't want to participate, you make it to the end of the survey as if you have completed it all. Because actually you HAVE.
This is exaclty what the instructions in this survey are, because of the conditions applied to each single question.
You hide all questions from group 2 onwards if one of the 3 conditions is not met. So LS is doing what it is asked an considers the survey to be finished.
What you actually want to do is to create a quota based on these 3 first questions and "screen people out" if they do not meet one of the three conditions. This will give a different end page and respondents that say that they do not work in Belgium, or are not 18 or above or do not agree with participating will not count as a complete survey.
There is nothing wrong with the survey, it does what it is told to. But the survey was not planned very well. That is the issue here.
While looking at your survey I found that there seems to be a lot of "garbage" in the code, e.g. """ appears so often and this should actually be _"_. I feel like there is a lot of cleaning necessary. Not sure where this comes from, but probably copied from somewhere in the wrong way. But that just as a side note, this should not be the cause of your described issue. But it might cause other issues.
Code:
<span style="font-family:&quot;Arial&quot;,sans-serif">
Another thing: That introduction is a little Monster and while I understand that ethic commitees want certain things to be mentioned, this thing is insane. Who is supposed to read all this? Either people don't read it and go straight to the survey or they stop mid way and drop out. In either way this text then has not helped at all. Those ethic commitees need to get real. Not sure where those people live. It certainly isn't earth.
By the way, there is so much in this text that doesn't even apply, that I am not sure if this text and thus the agreement with it is even legally binding, because it referrs to things that are written as if this was a paper and pencil survey (Signature, pages, etc.). But anyway, this is off topic again.
Topic Author
