Server Survey

Nick Barcet nick.barcet at
Fri Feb 22 17:48:11 UTC 2008

Nick Barcet wrote:
> A very dear friend of mine, working at a french statistical institute,
> will take a look at it for us, so I guess we will be covered on that side.

I just spoke with this person and reviewed the Survey. Here are the key
points from our talk :

- Online survey have a drastic fall rate if they take longer than 15m

- Our survey looks below this, because of the filtering we do.  Base
rule of thumb is
  * 30s for multiple choice or simple questions
  * 10s per item on rating questions
Based on this, our survey, without any filtering, should be taking 23m,
and be well below 15 after the filtering

- Product survey they conduct generally use a sample of 60 to 80 cases,
so our target of 100 is ok (I personally hope that we get more than this...)

- She thought our questions were clear and simple, which is a very good
point.  There should not be any problem in the data analysis the way we
set up the questions

- Warned me on the fact that open questions bring back very scarce reply
on the web: 3 words on average, so we should not expect too much from that.

On another front, I got the fix for the bug I encountered in LimeSurvey
today, so my setup is now ready for testing.  If you have already
volunteered, I'll be sending an email soon with details.  If you have
not yet, feel free to do so now :)

Last point we should be working on is defining a privacy statement for
the survey.  If anyone has a start point for this, go ahead and paste it
on the wiki page [1]



-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: OpenPGP digital signature
URL: <>

More information about the ubuntu-server mailing list