"thunder" wrote in message
...
On Mon, 06 Sep 2004 17:08:49 +0000, NOYB wrote:


That site also states that Rassmussen has explained that the Times and
Newsweek polls included too many republicans. That's pretty amazing
that
they know that Rassmussen is saying that already...because Rassmussen's
own website said that he'd give the reason for a discrepancy in the
different polls at 3 pm E.T. today. How does the author of your
electoral-vote website know what Rassmussen is going to say 2 hours
before
Rassmussen says it?

Liberal conspiracy? Perhaps, you should wait until 3 pm to see if he is
correct. By the by, others have noticed the faulty weighing:

http://www.mydd.com/story/2004/9/4/154842/1919

According to Chris Bowers (the author of that site you just listed), Harris,
Pew and NAES polls from earlier this year show that more people consider
themselves Democrat than Republican. He then tries to extrapolate the data
to reach the conclusion that Bush leads by only 5.6% rather than the 10 or
11 points shown in the Times and Newsweek polls. What a crock!

There are so many flaws with this theory, that it's laughable that any
serious person would even consider it. Here's just a few problems that I
see:

*The Harris, Pew and NAES polls could have been inaccurate for several
reasons...one of which could be that none of them considered whether the
people questioned were "likely voters".

*People might consider themselves "more Republican" since those three polls
were conducted.

*The assumption that just because "other" polls say that 94% of Republicans,
and only 82% of Democrats will vote for their party's nomination, doesn't
mean that *those* polls were accurate in the first place.

* You can't extrapolate data across multiple polls (which were polling
different questions, and at different points during the election year) to
conclude that a current Newsweek, or a current Time poll is inaccurate.
That's the most absurd reach that Bowers is trying to make

*mydd.com is a partisan Democratic site