Re: [nc-whois] Whois Survey data please....

["Paul M. Kane" <Paul.Kane@reacto.com>]

I agree with Oscar  .... and had assumed that in the preliminary review
it would be statistical analysis .. ( I studied statistics at Uni for 3
"interesting" years <smile>) so here's a sample with an error of 0.5%
....

Best

Paul
 
 

Sector	Commercial	Individual	ISP	Other	Non-Com	Regrar/ry	Gov.	Totals
Population	1048	1017	230	219	207	130	35	2886
Sample	216	215	125	122	118	88	31	915
Error	0.05	0.05	0.05	0.05	0.05	0.05	0.05

"Cade,Marilyn S - LGA" wrote:

I of course agree that the sample should be statistical! MC

-----Original Message-----
From: Oscar A. Robles-Garay [mailto:orobles@nic.mx]
Sent: Monday, October 08, 2001 11:33 AM
To: Cade,Marilyn S - LGA; 'YJ Park (MINC)'; Paul M. Kane; WHOIS Committee - DNSO
Subject: RE: [nc-whois] Whois Survey data please....

I could agree on you suggestion but let me propose to use a statistical formal sample. I recall (it was many years ago, when I used to be an engineer) there is a formula, depending on the "expected error": 5%, 2%, 1%, 0.5%. All we need is the total popultation (numbers of survays answered).