On Wed, 24 Mar 2004 05:26:36 GMT, doubter
wrote:
In either of the above cases, once the customer has defined the needs, the
pollsters use various data sources (census figures, voter registration
data, assessed property values, car registration data, magazine
subscription data, to name but a few) to build the sampling model. It is
a difficult and precise science. This is the reason leaders such as
Gallup, Field, Harris and Nielsen are respected and their polls are
usually given credibility.

Difficult, yes... Precise -- highly doubtful. I'm studying
statistics now, and I'm highly dubious of the claims that, in real
life, a sample of 2000 to 3000 people gives me a 95% or 99% confidence
interval on a population of several hundred million. I'd have a lot
more faith in these studies if they spent the money on samples of
10,000 to 50,000 people. But even then, there is tremendous room for
subjective error in defining a "suitable" sample group, even among
trained statisticians.

Steve O.
Steven AATT Domain DDOOTT com
To send an e-mail, substitute @ for AATT, a . for DDOOTT, and OpComm for Domain

Standard Antiflame Disclaimer: Please don't flame me. I may actually *be* an idiot, but even idiots have feelings.