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

You made a survey without a survey statistician on your team.  Did you
send out a request for such a person on the mailing lists to advise you
before you put together the survey?  Did you have a clear and concise
question that you wanted to answer before you developed the survey
questions?  Did you run the questions by an aforementioned professional
in the staff and check for confirmation bias?
I am not a professional statistician, and that's just what I spotted.  I
have covered surveys as a journalist in my previous career, though.  And
I also am a veteran of setting up business projects.  A survey
statistician would have a lot more to say I am sure.  And we're not even
starting on the analysis.  In fact, I'd throw out the analysis and the
results and start anew.  First off, define "users" (end users,
evangelists, business users?) and state the overall purpose of your
survey in a single question.
I regret some of the tone of the previous e-mail (first e-mail prior to
coffee), but there's nothing here to work with.  You've got 300
self-selected users with at least two major questions in one survey that
you did not break out by region, sex, profession.  You want results, you
need good data underneath.

On 11/12/2013 7:04 AM, Charles-H. Schulz wrote:
> Le Tue, 12 Nov 2013 06:57:20 -0700,
> John Meyer <johnmeyer@pueblocomputing.com> a écrit :
>
> > On 11/10/2013 11:46 AM, Charles-H. Schulz wrote:
> > > Hello,
> > >
> > > As there were some exchanges about the survey here and as I
> > > advertised it on this mailing list as well, I thought you might be
> > > interested by my initial analysis:
> > > http://standardsandfreedom.net/index.php/2013/11/10/users-the-final-frontier/
> > >
> > > Thank you for your participation!
> >
> >
> > 1.  The survey seems to be a Self seLected Opinion Poll (SLOP), so I'm
> > taking it with a grain of salt the size of the Sears Tower.  There's
> > no margin of error included in the poll either and based upon the
> > sample as being from the mailing lists (where people are generally
> > active anyway) I'd say it's fairly skewed.
> > 2.  The conclusions are generic, wishy-washy and are based on guesses
> > and assumptions with no hard underlying data.  How much in
> > contributions has LibreOffice raised?  Does that fit in with what the
> > survey said? Where is the Quality Assurance in the web site?  And why
> > would an end user be interested in that?
> > 3.  User support and quality assurance do not require too much time or
> > technical knowledge.  Remind me not to hire you for either of those
> > tasks in my business.  Those are things that professional companies
> > hire entire other companies to do.
> >
> > I'd give this project an F in a freshman statistics class, and would
> > not base any strategy off of this "survey"
>
> Thanks John, I'll take it  from your comment that
>  1) you are either a survey professional and you only wait for the next
>  survey to contribute your time designing it
>
> and/or 
>
>  2) you will contribute the costs of hiring a market research firm the
>  next time we need a survey. 
>
> Allegedly, I and none of the other people who designed the survey are
> professional survey designers. 
>
> Best,


-- 
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