Dear all,

using version 2.72.6+ (171207), I need to track the time subjects spent on a page if they dropout (i.e., 30 seconds on the page before they left the page or closed it). Save timings only measures the time when subjects submit a page.

Anyone knows how to do that?
Help is highly appreciated

Best,
Andreas

No sure if this is even possible. You would have to record the time from opening the page to closing the page. Not sure if "leaving/closing" a page can trigger some Javascript and where to write this information, if the page is closed.

And not even sure if it make sense. Because I sometimes leave pages open for hours in the background. Then come back, or just close it because I don't need it anymore. So which case is it? Can't be determined.

What do you need this information for? Maybe we find a better solution?

You could place a script like this in the source of a hidden (via CSS) short-text question. This will increment the value of that question every second.

Hi Tpartner, great idea to count until the window is closed.

However, how and where do you store this?

Because a question on that same page won't be stored if someone closes the tab page or closes the browser? Or am I thinking wrong?

Oh, yeah, of course...I need to think about that some.

holch wrote: Because a question on that same page won't be stored if someone closes the tab page or closes the browser? Or am I thinking wrong?

The JavaScript would need to POST data or use a GET URL every x seconds. LimeSurvey would need a different URL generation / session concept to allow that in an easy way.
A JavaScript heartbeat script would be a nice thing. If the page/questionposition is available via JavaScript, an external script would be able to write respondent id, time and question/page position in a separate storage on the server.

@Jelo: I think there are far more important (basic) things to implement in LS before this very specific feature, don't you think?

E.g. randomization with fixed items, being able to choose the position of other, etc.

holch wrote: @Jelo: I think there are far more important (basic) things to implement in LS before this very specific feature, don't you think?

I have read things like:

Limesurvey is cheaper and have a lot more features than other surveytools

. Perhaps I got it wrong. But perhaps people writing such statements, are not using other survey tools at all. I only recognize that, when people tell me about features, which I haven't found. E.g. that LimeSurvey supports Skip-Logic. I'm was only aware of Display-Logic. But I was pointed to the feature-list, which states Skip-Logic.

www.limesurvey.org/editions-and-prices/l...rofessional-features

I thought "Skip-logic" is quite clear. Skip to a certain position in the survey. Which is not the same as hide all other questions.

Well, I think the Skip-Logic lies in the eye of the beholder. I don't know the exact definition of "skipping a question", but in my opinion, LS support that. You can skip questions, depending on the profile of the respondent or on previous answers. So I would personally say, this is correct.

What LS doesn't have is Jumps, which can be handy in some cases.

Actually I find the Skip-Logic quite simple, easy to understand and flexible. It has it's disadvantages when many questions need to be skipped or when respondents should jump to the end.

Overall, I think people who say "Limesurvey is cheaper and have a lot more features than other surveytools" they are partially correct. Some of the more basic survey tools have probably less advanced features.

I am often surprised that LS has some pretty advanced features (compared to paid basic tools), but then lacks some very basic features (real screen outs, etc.) that should be available to any tool that wants to cater to serious market researchers. Probably those features were implemented because one of the developers needed it for a client project.

But I am surprised that with a market research company being part of the LS GmbH such features as screenouts, flexible position of other fields, exclusion of randomization, subcategories for subquestions, etc. haven't been on the to do list so far.

I love Tpartners workarounds, because they allow you to do things with LS that can't be done out of the box. But i would rather like to be able to do such basic thinks out of the box. There will always be need for workarounds, because people come up with some crazy requirements that might be needed once or twice in a lifetime, these are the things that should be done with workarounds. As soon as these workarounds are used on a regular basis by many users, they should find their way into the tool.