I am running a 4 language survey and is having problems with incomplete surveys. After 4 weeks I have had 280 surveys, whereas 160 is incomplete.

A lot of people probably drop out in most surveys, but I have had reports that the survey stops loading after some questions. And I really suspect some error because people who have given very detailed answers are suddenly dropping out. Has anyone else had problems with this?

I have not yet updated my Limesurvey and is running Versjon 1.91+ Build 11379.

Thanks.

Hello,

I've also had some incomplete surveys though it wasn't that much. The first thing I would check is the Session lifetime in the global settings "General" view. If the session lifetime is too short some participants might be kicked out of the survery before the finish it.


Another idea might be to export the vv file of the survey (in the "Responses and statistics" view), open it with an editor or spreadsheet programm and try to find out where the participant stops and which is the last page of the unfinished survey.

If you have acess to the server of your LimeSurvey installation you might also have a look at the server logs. Perhaps you will see some entries in an access log that look strange to you or there are even errors which explain why the participants were stoped by force.

The best way to solve this is to ask users on what page of the survey they were running into problems and also which browser they are using.

You might also find this out by checking at which question people often stop answering.

If you are using tokens and your survey is set to be non anonymous, you can send another reminder, explaining the technical details and asking users to complete the survey they have started (they should be directed to the page where they left).

Thanks for your suggestions.

Session time is 3600s at the moment, which should be more than enough.

I have gone through the 50 last incomplete surveys, and I can see no real trend. Question group 3 and 5 are a bit overrepresented, but not much really.

Answers from Sweden are overrepresented though.

I updated my Limesurvey to Versjon 1.91+ Build 12170. It might help?

I'm not sure if the update will help much because there are no huge changes or know bugs that cause such a behavior.
Please check the server error log and compare the entries with the date/time of the incomplete answers. Maybe this gives some hints.

reidarborgersen wrote: I am running a 4 language survey and is having problems with incomplete surveys. After 4 weeks I have had 280 surveys, whereas 160 is incomplete.

The numbers do not indicate any problems to me. There are lots of complete responses so the survey seems to work fine.The high number of incomplete responses doesn't necesarrily indicate that there is a problem. Such an incomplete response is created even if people just click the link t the survey and have a look at the first page. Then they may not be interested in the mentioned topic and close the window. That alone already creates an incomplete response.When exporting data there should be a column "last page" at the beginning. That tells you how many pages they have opened from your survey. I assume that for most of the incomplete survey, the number is rather low which means they didn't really start the survey at all.

Mazi : please ! Don't answer 9 years after …

DenisChenu wrote: Mazi : please ! Don't answer 9 years after …

Why not? A user contacted me by email about this topic and I gave him this lengthy information. 
The details have not changed and I have only given additional details will may be helpful to several users. So I see no reason why this should not be done, even if the this is an old thread.