Re: [ga] Whois Accuracy Study Launches

["'kent'" <kent@xxxxxxxxx>]

On Wed, Jun 10, 2009 at 09:35:23PM +0100, Debbie Garside wrote:
> I don't think this is overly stringent and much of the work that is being
> done e.g. sourcing of databases per country and running the list against
> them for matches will be just as cheap to do for 400 or 800 as it is for 3.
> In fact all of the companies I have dealt with in the past (admittedly only
> in the UK

This isn't a single nation study.  Costs go up when you have to work
worldwide.  Databases that exist in UK don't necessarily exist elsewhere; you
need the expertise to know which databases exist and how you access them.  If
you don't have a presence in the other country, you need to subcontract
people on the ground in those countries.  You need contacts, expertise,
experience in those countries. 

>  but I have been doing this since 1988) have a minimum charge that
> usually equates to the cost of processing 1,000 or 4,000 records.

How do you -- specifically -- verify a postal address in Beijing, or Tokyo,
or Hyderabad? How much would it cost? When it comes to phoning the
registrant, how do you speak to them? Are there cultural norms you need to
worry about? etc etc etc.  Multinational studies are a big deal.

> > There is a great deal of "on the internet no one knows your a dog"
> > uncertainty -- this study is the first one I have ever seen
> > that will actually address that issue head on, and I'm really
> > interested to see what the results will be :-)
> 
> I just think that more could be gleaned from this study that may prove
> useful in policy making.  How inaccurate the Whois database is as a whole is
> rather boring and useless information.

The scope of the project is limited, but I personally don't think that the
information will be useless, or even boring.

> What does it matter whether it is
> 10% or 20% inaccurate or 3.3% for .com or 0.9% for .org (as identified by
> GAO)
> If Russia has a higher level of .com inaccuracies than the UK what do you
> think that would tell you?

I don't know.  Probably nothing, but I'd have to really look at the details
of any such study. 

>  If you link the level's of inaccuracies to GDP
> data do you think you may find a trend?  Possible!

I think such correlations are generally pretty meaningless, to tell you the
truth: more "Lies, Damn Lies, and Statistics", and certainly not at all 
relevant to ICANN's mission.

>  What does 3.3%
> inaccuracies within .com's and 0.9% in .org's tell you?  Is there a
> difference in the registration procedures for these or is it that there is
> more scope for phishing with a .com?

I agree 100% it would be nice to know more.

[...]

> > Per country accuracy statistics are not the goal of the
> > study, only overall accuracy.
> 
> But there does seem to be a lot of emphasis on information regarding country
> profile within this methodology.

There is an entire science of designing samples so that they rigorously give
you the information you want at minimal cost.  If country A has 1% of total
registrations, and country B has 1% of registrations, and it costs 10 times
as much to verify a registrant in country A, then it is statistically
permissable, under certain circumstances and constraints, to just use data
from country B.  But those circumstances and constraints are very important,
if statistical validity is to be maintained.  Hence, the fairly detailed 
description of the sample methodology.

[...]

> That's as may be but their web page is down so I can't check them out
> www.norc.uchicago.edu/

Works for me.  Here's a sample from the list of current projects at NORC:


  Assessment of the Uses and Users of HealthierUS and Healthy People 2010

    HealthierUS and Healthy People 2010 are two important Federal initiatives
    aimed at improving the health of Americans through disease prevention and
    health promotion.  Though conceived during di... 

  Bank of Spain's Survey of Household Finances (EFF)

    Sponsored by Spain's Central Bank, NORC is carrying out the third round
    of the "Encuesta Financiera de las Familias" (EFF), Spain's Survey of
    Household Finances.  The EFF is the sole source of ... 

  Baseline Survey for "Mi Primer Empleo" Training Program, Honduras

    NORC has completed work on a baseline survey for an impact evaluation of
    the second round of the Mi Primer Empleo ("My First Job") Program in
    Honduras.  Mi Primer Empleo is a ... 

  Basic Theories and Models of Religious Change

    The Metanexus Institute, which advances research, education and outreach
    on the constructive engagement of science and religion, has made a grant
    under its Templeton Advanced Research Project to ... 

  Black Youth Culture Survey

    The Black Youth Culture Survey is a key part of a larger effort --The
    Black Youth Project-- which is examining the attitudes, resources, and
    culture of African American youth ages 15... 

Kent