[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Survey on FLOSS

On Wed, Mar 30, 2005 at 04:34:39PM +0100, MJ Ray wrote:
> As a statistician, I am worried that this is yet another
> under-tested survey form about this important topic being
> released into the world without a good pilot study first. Where
> is the pilot study report? I can't figure it out from the project
> status page, it's not on the "dissemination" page and there's no
> "downloads" page.

As somebody who understands and hates statisticians and statistics, I
echo the sentiment. This is a terrible survey.

> Some of the questions (18 for example) seem to assume
> a particular bias in the respondant. In others (12 and 38),
> some of the range of possible answers are not covered. The
> "discrimination" page is one-sided and not justified.

I find several of the questions impossible to accurately answer
(either because they call for knowledge about my activities that I
have no reason to record, or because they list a set of options which
lack one relevant to me), and I disagree with the assumed premises
behind several of them. None of them had escape options for "none of
the above" or "tell us why this question sucked", or similar.

Also, this survey is riddled with the "How important do you think this
is?" style questions, which are notoriously bad at collecting accurate
information. And to round it off, different questions give different
sets of responses, using different form layouts, and even different
numbers of responses. As a parting shot I was particularly impressed
by one question where I was asked to rank the importance of "Other".

I've seen better surveys from teenage students. A detailed critique
would be a waste of my time; I suggest that the people responsible ask
their relevant professors to do it.

And I hate this stupid term "FLOSS". It's a blind attempt to lump two
entirely discrete groups of people together, who have radically
different motivations. A very strange choice of sample set. I would
expect the two groups to give very different answers to several of the

  .''`.  ** Debian GNU/Linux ** | Andrew Suffield
 : :' :  http://www.debian.org/ |
 `. `'                          |
   `-             -><-          |

Attachment: signature.asc
Description: Digital signature

Reply to: