[gnso-rpm-wg] Critique of INTA survey
icann at leap.com
Thu Aug 31 02:48:52 UTC 2017
On Wed, Aug 30, 2017 at 10:07 PM, icannlists <icannlists at winston.com> wrote:
> I've read with some amusement your various critiques of the INTA survey. However, I could not find even imperfect surveys that you have submitted that tend to support your various positions taken on this list and on calls. I hope you will be as open minded about what we can learn from what you view as a flawed survey as we have all been about your positions taken based on non-surveys. However, if you have submitted surveys that tend to back your various positions -especially scientifically perfect ones - and I simply overlooked them, could you send them round again now? Thanks.
If you saw a specific fault in the analysis I provided (or Kurt's),
please be specific/precise about what you feel is incorrect. If you
accept the analysis as correct, and your only point is "nobody's
perfect", that's a weak argument. Bad data is often worse than no data
at all, because bad data can lead one towards making bad policy
choices, and indeed even embolden that decision-making (because
there's a pretense that it was supported by data).
For instance, if there was an automotive transportation policy survey,
and 80% of the non-random responses (all 33 of them) were from Rolls
Royce owners (an unrepresentative small sample), it should be obvious
that the survey would likely lead policymakers to make incorrect
decisions for the population of 100 million+ drivers.
An invalid survey (due to a non-random sample combined with a far too
small sample size) is just that --- there's little to learn from it
(except perhaps how to design better surveys).
I've not submitted any survey data. I don't have the budget of ICANN
or INTA supporting my participation in this PDP. My positions are
backed by sound and logical reasoning, with deep analysis of the data
that this group has had access to (e.g. answers from the TMCH
operator, information from The Analysis Group, etc.).
More information about the gnso-rpm-wg