Fred Stone wrote:

"chris.linthompson@gmail.com" <chris.linthompson@gmail.com> wrote in
news:1131159006.110574.278300@g44g2000cwa.googlegroups.com:

Fred Stone wrote:

"chris.linthompson@gmail.com" <chris.linthompson@gmail.com> wrote
in news:1131147813.836720.63620@g47g2000cwa.googlegroups.com:

Fred Stone wrote:

AC <mightymartianca@hotmail.com> wrote in
news:slrndmnf5e.ljo.mightymartianca@nobody.here:

On 4 Nov 2005 11:41:50 -0800,
chris.linthompson@gmail.com <chris.linthompson@gmail.com>
wrote:

Fred Stone wrote:

"A.Carlson" <amcarls@hotmail.com> wrote in
news:i0qlm11gu4boktqsh1mf9joc4l109v54o8@4ax.com:

On Fri, 4 Nov 2005 04:09:15 +0000 (UTC), Fred Stone
<fstone69@earthling.com> wrote:

"Ike" <accordiondocxyzxyzxyz@mindspring.com> wrote in
news:PIzaf.4606$m81.3502

"Fred Stone" <fstone69@earthling.com> wrote in message
news:Xns9703CF5963BB2fstone69@213.155.197.138...

bdbryant@mail.utexas.edu (Bobby D. Bryant) wrote in
news:dkebkp$isl$1 @geraldo.cc.utexas.edu:

On Thu, 03 Nov 2005, "maff" <maff91@yahoo.com> wrote:

On virtually every key measure of presidential
character and performance, the new survey found that
Bush has never been less popular with the American
people. Currently 39 percent approve of the job he
is doing as president

CBS (IIRC) announced their latest poll yesterday, and
it was only 35%.

Cheny was at 19%, down 9 points in the past month.

Congress was at 34%.

It's real easy to get results like that when your
survey sample includes only 27% Republicans and then
you weight them down to 23%.

There are less republicans than ever. maybe there are
only 23%. I sure wouldn't want to be caught dead in
that party since day one.

You're falsely generalizing from your own irrationality.
The numbers of republicans and democrats should be
roughly equal, according to the last election.

Do you really think that all Americans belong to one of
the two major parties or that people consistently vote
the party line?

No, and I note that "Independents" were the largest group
of respondents in that survey.

There are plenty of people who consider themselves
independent and plenty of people, even party members who
chose to sit out a given election for any number of
reasons.

Not to mention the fact that in some states you have to
register for one party or the other for primary purposes
but that does not obligate you as to how you vote in the
general election.

I'm sure that there are also at least a few people who
would declare themselves independent for poll purposes
but register one way or the other out of necessity due to
peculiarities of local voting laws.

The bottom line is that party affiliation can be a poor
indicator as to who will win in a given election and, by
the same token, who wins in a given election can be a
poor indicator of party affiliation.

While this is true, it is also true that Democrats do not
outnumber Republicans by more than 10%.

In fact, the percentages of Repubs vs Dems matters not one
whit, since the results are broken down by party. The really
interesting thing isn't that Democrats hate Bush; it's that
the level of disatisfaction among Republicans climbed so
high, so quickly.

Well, here's the key question. Does Fred generally agree
with the survey's results?

The results are not reliable, for the reasons that I've given.

Your reasons, I'm afraid, are not sufficient. For example, you
didn't comment on their rationale for weighting the samples as
they did.

Neither did you.

But then, I haven't taken issue with their methods.

So you just blindly accept whatever you read in the MSM.

Not hardly. But I tend to think it's more reliable when it
corroborates other published material.

To
simply assert they should not have done so tells us nothing.
Why, exactly, should they have not weighted the percentages as
they did?

I have given my reasons for believing that their weights were
inappropriate. Take them or leave them.

No, you made the assertion that they were inappropriate. Not quite
the same thing.

No, I explained why I made that assertion.

Here's what you said:

"It's real easy to get results like that when your survey sample
includes
only 27% Republicans and then you weight them down to 23%. "

and,

"And then they weighted the Republicans down by 4%. They didn't weight
the Democrats, and they weighted Independents UP by as much as they
pushed Republicans down.

It's real easy to get the result you want if you skew the sample by
more
than 10%."

and, "You're falsely generalizing from your own irrationality. The
numbers of
republicans and democrats should be roughly equal, according to the
last
election. "

and shortly after that you said,

"The results are not reliable, for the reasons that I've given."

Now, there's nowhere you explain *why* you're unhappy with their
weighting sceme.

You also are are on record as saying there's no
technical discussion of their weighting. Without a technical
discussion, there's no way to deal with these details- except you
seemed to wave your hands and do so.

What effect did the weighting have on the numbers when they were
broken down by political party, as opposed to the overall
numbers?

That analysis would require work that I'm not getting paid for. If
you'd like my detailed analysis, let me know and I'll give you my
paypal address.

Are you qualified to do the work? Are you a statistician?

Are you going to pay for the work you want done?

Evasion noted.

In any case, thanks for saying you based your conclusions on...an
analysis that was never done?.

On analysis that I *did* do. I didn't do the analysis that you're
asking for because it wasn't needed.

Your analysis consisted of "I don't like these numbers. What can I
fixate on?"

You's have to address these issues before your conclusions can
be accepted.

You've done nothing to address those issues either, so why should
we accept their conclusions?

Beacause as Dave Fritzinger has pointed out, this poll is exactly
in line with every other recent poll concerning the President's
popularity, approval rating, and perceived trustworthiness.

Most other recent polls don't put him as low as 35%.

Now if all the other polls agreed his approval was 35%, then I WOULD
know something was rotten. But they're all within just a few points,
aren't they- in the general vicinity of "sewer".

If you
say this one is flawed, you better be able to back up the
rationale, for this poll and several others. So far, you've
presented nothing to make us think this one is flawed.

Nothing I present will make you think. You already let the MSM tell
you what to believe.

Au contraire. I could point to numerous examples where I admitted I
was incorrect- sometimes by conservatives on matters of politics, too.
For as long as we've been sparring, I've never seen a similar
admission from you, Fred.