[libreoffice-users] Re: Engaging users: initial results of the survey

Hi, Charles,

On 11/11/13 4:19 AM, Charles-H. Schulz wrote:
> Hello Alex,
>
> Le Mon, 11 Nov 2013 09:05:46 +0100,
> Alex Thurgood <alex.thurgood@gmail.com> a écrit :
>
> > Le 10/11/2013 19:46, Charles-H. Schulz a écrit :
> >
> > Hi Charles,
> >
> > Whilst I appreciate the effort in designing such a survey and the
> > objective behind it, I too, must admit that the survey was not worded
> > in a way in which I felt comfortable responding. Indeed, it seemed to
> > be distinctly biased towards getting the participants to answer in a
> > given direction.
>
> Next time I'm  sure you can join us in the weeks during which we
> discussed the survey on the marketing and project's list :-)

I hope you are making a list of the concerns voiced in this thread, and 
the other thread about the survey.  That will give you additional points 
to look at for the next survey.

> > Additionally, I would question the statistical relevance of 600
> > responses, when the project is alleged to have tens/hundreds of
> > thousands of users. If only 600 hundred people took the time to
> > respond, what does this say about :
> >
> > - penetration of the product;
>
> I honestly would not  think there's relevant data for this in the
> survey and from the respondents.
>
> > - reach of the survey;
>
> Good question with no easy answer. The survey was localized in 5
> languages aside English. The link was posted here and on the several
> other users mailing lists. The word was spread on the Facebook
> LibreOffice page and Google+ and to a lesser extent on Twitter.
>
> Once I've said this I guess I didn't say much. Here are two additional
> facts:
> - the survey was not translated (and not propagated) to three countries
>    where we know we have an active community and anywhere between a non
>    trivial number of users up to a large number of them: Russia, Japan,
>    Brazil. Judging by the survey results, it would seem that their
>    impact has been minimal or virtually non-existent. So the survey
>    reached out to some categories of users, but not all of them. I'd be
>    however confident in stating that the users that responded are
>    representative of the LibreOffice users in general.

Let's work with Alex's comment there were 600 responses to the survey. 
You often read how LO has thousands and thousands of users.  Just to 
make it simple, let's say there's 100,000 users.  That's probably 
miniscule to MSO and possibly even WordPerfect.

That means, at best you got the opinions of .6% of the users. 
Personally, I would never consider that to be representative of the user 
base, especially when you noted in the next paragraph of the limitation 
of the survey's distribution.  I would seriously consider junking this 
survey's results, using it as a learning experience, and doing a better 
survey.

Really, all you have is the opinions of the users of the mailing list, 
not users in general.

> - the survey could have had a bigger and much deeper outreach if  it
>    had been pushed directly to the users, say at the installation phase
>    or even through a mechanism allowing users to respond to it via the
>    StartCenter. That was obviously not the case, so in the end we
>    reached out the users who are on the project's mailing list and
>    connected to us through our social networks. This leaves out plenty
>    of users irrespective of their language.
>
>
> > - design of the survey;
>
> What would you like to know? The survey was designed in order to be
> progressive in its questioning as should be all the surveys. Beyond
> that, don't look too much into survey methodologies, I'm not sure they
> are that sophisticated, unless of course you would like to get a
> particular answer in advance, and that's precisely what  we wanted to
> avoid.

As I and someone else pointed out, you limited the range of the 
responses.  When you do that, you automatically color/bias the results 
of any survey.  Easy to correlate the data received, but no guarantee of 
accuracy.  Which is why, so often, election results don't match the 
polls and surveys.   :-)

> > - length and time for which the survey ran ?
>
> The survey  started on the 31st of October and expired yesterday.
>
> Best,
>
> Charles.
>
>
> > Alex


--
Ken

Mac OS X 10.8.5
Firefox 24.0
Thunderbird 17.0.8
LibreOffice 4.1.2.3


--
To unsubscribe e-mail to: users+unsubscribe@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted