Hi Thomas,

It seems there is a way to restart active respondents. From what I
can tell (I don't know all of the mechanics behind it) if you pass in
a valid I.Project, I.Session and I.Savepoint on the URL then the
survey starts you back at the screen you left off even if the survey
is still active.

To test this, do the following:

1. Take a survey using whatever URL you usually do - ie.
http://server/mrIWeb/mrIweb.dll?I.project=XXXXX&var1=1&var2=2

2. On the first page of the survey, view the source and look for the
hidden input fields I.Session and I.Savepoint (you already have
I.Project of course). Copy the values somewhere. The Session refers
the Value Cache file (.chj) that is created to hold the data for that
interview including the savepoints, quota information, etc. BTW -
you can also access the SessionToken in the Routing and the SavePoint
is just the question name for the question your are on.

3. Close your browser to simluate a resp closing it (though it
doesn't really matter if you do or not)

4. Open a new browser, and set your URL to something like this:
http://server/mrIWeb/mrIweb.dll?I.project=XXXXX&I.Session=
[sessionvalue]&I.Savepoint=[savepointvalue]

You should start right back up where you left off.

Now - obviously, you can't be asking resp's to do this to start their
survey back up but you could wrap the survey environemnt up with
something that helps automate and streamline it - getting the muddy
details out of the resps hands. I suppose you could maybe use a
combination of cookies and a jump page?

We've never really been pressured to come up with a solution for this
here. But, if you do come up with a nice way to use this facility
let us know.

Thanks.

Tyler Ball
Kantar Operations

--- In NADUG@yahoogroups.com, "Thomas Subias" <tsubias@...> wrote:
>
> Has anyone come up with a better way of handling interview timeouts
> than what Dimensions (running 3.5) offers by default?
>
> With the out-of-box setup, a respondent cannot return to an
> in-progress interview until the X minute timer has expired and the
> session goes from in progress to timed out. If we have X at 5, that
> is a long time to keep a respondent's attention. And we know that
as
> X->0 we'll take a performance hit. But as-is we're giving a less
than
> ideal experience to respondents, particularly as, under our old
survey
> system, there was no such limitation.
>
> Aside from tuning the X value, has anyone come up with a means to
> allow respondents to return to an interrupted interview immediately?
>
> Any input is appreciated.
>
> -Thomas
> KnowledgeNetworks
>