Hi Guys,
I have a strange issue with few survey responses in my survey.

I have a survey with more than 2,100 responses and they are all ok.. only 3 respondents have answered questions that supposed to be hidden, based on their answers in a previous question, instead of the questions in the branch they chose.

conditions are set through the system and not with javascript, so I don't think it's a browser-related issue.

Can somebody help? Did anybody face the same issue before?

N.B: I've attached a screenshot of q1 & q2 in EM, where q2 should be hidden based on the answers of q1.

Thanks & Regards,
Medhat

Hi, medhat,

if there are three respondents who passed a condition you should very carefully investigate ALL their answers.,
Can you reproduce this issue entering the same data these respondents entered?


You can imagine that it is nearly unpossible to give advise without seeing your survey.

Last, but not least: which LS version?

Kind regards
Joffm

Thanks Joffm for your reply.

I couldn't reproduce the issue.
Even when trying with the same data, the survey works fine as it should be!

I'm using version 2.00+ ( I know it's a bit old, but I've used it in 100+ surveys with no issues)

Kind Regards,

medhat wrote: I have a survey with more than 2,100 responses and they are all ok.. only 3 respondents have answered questions that supposed to be hidden, based on their answers in a previous question, instead of the questions in the branch they chose.

When I had these kinds of issue is was mostly a browser environment issue. Limesurvey is still not recording the browserinfos via http string automatically. So it isn't possible to pinpoint the issue. I often came across such issues when using Limesurvey in combination with other survey tools. They captured the browserinfos without workaround javascript questions.

The most common issue was that a hidden question wasn't hidden in that browser. I had that e.g with a random number in a hidden field.
This questions wasn't hidden and people entered a different number into that field. Funny results we had.
Second most common issue was around sessions. Sometimes people stop and continued later. Sometimes the session handling gets lost and respondents still seem to be able to continue somewhere. But e.g. conditions no longer get the correct information.

Sometimes people just fiddle around with the survey and try crazy things. So manipulation with intention is also in play from time to time.

It is really uncommon, that you only have three responses with issues With 2100+ responses it still looks like a successful survey project.

jelo wrote: Limesurvey is still not recording the browserinfos via http string automatically.....

www.limesurvey.org/forum/plugins/107836-...tion-in-your-surveys ;)