[cc-community] cc usage stats

Luis Villa luis.villa at gmail.com
Mon Nov 27 16:17:22 EST 2006


On 11/27/06, Mike Linksvayer <ml at creativecommons.org> wrote:
> On Mon, 2006-11-27 at 03:30 -0600, León Felipe Sánchez Ambía wrote:
> > What would be great besides having the information from time to time
> > is to have a tool or that somebody could tell us how to generate
> > these statistics on our own. It seems the tool available via open
> > business is not completely accurate as it doesn't include all countries.
>
> There's code the 'stats' directory of the svn repository for
> https://sourceforge.net/projects/cctools
>
> Would probably take a fair amount of effort and knowledge to run, but
> you're welcome to, and to submit patches. :)
>
> Note also that any single number can and will wildly fluctuate depending
> what search engine cluster(?) you might be hitting.
>
> The numbers we've been putting out every six months are what I believe
> at the time to be lower bounds, but are still very hand-wavy.
>
> As I said in an earlier reply to Alek, another round will be coming at
> some point in December.  These will probably include some jurisdiction
> estimates, but these will be even more hand-wavy than the aggregate
> estimates.

For the purposes of estimating percentages of license by type (as
opposed to total # of licenses), or for the purposes of estimating
trend in total # of licenses, have you looked at doing a rolling
estimate from multiple sources, a la the multiple-source political
polls at pollster.com:

http://www.pollster.com/polls/?state=VA&race=senate_race

The rolling average of a number of skewed/individually broken polls
can be quite useful for viewing trends even when the individual
sources are known to be broken/not statistically very valid.

Luis


More information about the cc-community mailing list