[gnso-rpm-wg] Critique of INTA survey

[George Kirikos]

Hi Kurt,

Thanks for mostly agreeing with my analysis. However:

On Wed, Aug 30, 2017 at 1:47 PM, Kurt Pritz <kurt at kjpritz.com> wrote:
> There was one conclusion I could draw. It states that UDRP and Sunrise were
> the favored rights protection mechanisms, used to a major or moderate extent
> by 67% and 64% of the respondents respectively. The next most utilized RPMs
> were Trademark Claims and URS (by 36% and 27% respectively). To me this says
> that, to those who are in-the-know, Sunrise is a highly-valued RPM and,
> therefore, should be continued. (Sorry, George) (see slides 15 and 51)

The first part of your conclusion is correct (obviously anyone who
personally benefits from "front of the line" privileges is going to
value it), but the second part (therefore, that it should be
continued) is NOT correct. As a PDP, our job is to weigh the benefits
against the costs of policy choices amongst ALL stakeholders, not just
ones receiving benefits.

Thus, if that was "the one conclusion (you) could draw", and it's now
debunked, then we're left with the truth, that no conclusions can be
drawn from it --- it's for entertainment value only, i.e. it's an
advocacy piece, marketing fluff, not a scientifically-valid survey
that would endure any serious peer review from those in the field of
statistics.

To be clear, I tried to keep yesterday's email as short as possible
(remember, it was a response to a very long document), and didn't
point out every flaw with the survey. To point out another, note that
on page 6 it notes that 67% of responses were from USA and Canada.
However, INTA's own website states that:

https://www.inta.org/Membership/Pages/Membership.aspx

"63% of our member organizations are outside of North America."

This further reinforces my point that it was an unrepresentative
sample. As we know from election polling, the survey companies make
adjustments in weighting to attempt to compensate for the
unrepresentative samples (e.g. if too many men were sampled relative
to the known proportion, they'd lower the weights accordingly, etc.).
No attempts were made to do this (nor could they credibly have done
so, given the small sample size, and lack of randomness).

This is a classic case of "If you torture the data long enough, it
will confess to anything."

Sincerely,

George Kirikos
416-588-0269
http://www.leap.com/