This is a multi-part message in MIME format.
---------------------- multipart/alternative attachment
Larry,
>>Sarah, the only thing you've "proven" is that someone was playing fast =
and loose with the numbers from the exit polls.=20
"Proving" is done by attorneys and folks like you. Scientists would =
never be so arrogant as to presume to have "proven" anything. They =
formulate and test theories, and the modern scientist does this with =
heavy use of mathematical principles of probability, a.k.a. statistics.
Now, let's formalize this, in semi-lay terms ('cuz I don't want y'all to =
fall asleep on me):
PINKO COMMIE THEORY: Bush didn't really win. The election was "won" =
only through hacking of the non-paper vote.
HYPOTHESIS: The vote count will differ from exit polling in states where =
non-paper balloting is utilized.
DESIGN: Randomly sample states with non-paper balloting and states with =
paper balloting. Record the differential between Kerry's margins in the =
final vote tally and in the exit polls. Translate the data to actual - =
predicted margin, yielding "margin error." Test for differences between =
paper and non-paper states.
RESULTS: As stated before in "Who REALLY won" post. Non-paper states =
had a much greater margin error that statistically favored Bush. =
Moreover, paper states had no statistical margin error that would favor =
either candidate.
The hypothesis is therefore supported, providing support for the pinko =
commie theory. This does not "prove" the theory. It is merely a =
(strong) point in favor of the theory. Other issues and hypotheses can =
and should be examined, of course.
POSSIBLE PROBLEMS:
(1) The states might not have been randomly sampled, hence my question =
to Thump. Since I didn't collect the data, I don't know.
(2) Exit polling may not have been done the same way in all states. In =
particular, if there were differences in exit polling between the two =
*groups* (i.e. paper and non-paper), then the data could be screwed up.
FACTORS THAT ARE *NOT* PROBLEMATIC:
(1) Exit polls were not done at *all* polling locations (e.g. Conrad's =
point). However, if the polling locations were randomly selected, which =
would be consistent with the goals of an exit poll, then the exit poll =
data *should* predict the actual vote count with no net bias towards =
either candidate. The results would be +/- a certain margin for error. =
In the long run, with infinite sampling, the error would be zero. =
Statistical methods, such as I used, detect departures from zero bias. =
Tests get more sensitive with more extensive sampling. My testing =
reveals that there is a statistical departure between paper ballot and =
non-paper ballot states, assuming that the polling data were collected =
the same way in both sets of states. The question to be answered is =
what the source of that departure is.
Now, I would enjoy hearing clarification on the possible problems with =
the data. I suspect that these issues do not present as problems with =
the data, but again, I don't *know* that. They aren't my data.
LARRY'S ISSUES:
Larry has suggested that "someone was playing fast and loose with the =
numbers from the exit polls." That may be, Larry. That's why I would =
like clarification on the above. Please understand my conclusion, =
though, which is that this error between exit polling and vote count =
exists only in non-paper states and favors Bush. Why is that? Perhaps =
you are proposing a different conspiracy from the pinko leftist commie =
media?
LARRY'S PROPOSED CONSPIRACY??:
The entity conducting the exit polling wanted to produce data =
implicating that the non-paper states were guilty of election fraud that =
would favor Bush. They were less concerned with arriving at credible =
outcome conclusions that they could utilize to correctly "call" =
electoral outcomes, and thus they were willing to sacrifice their =
reputation for reliability to the greater cause of pointing a finger of =
blame towards the Republicans.
Well, like most theories, this is potentially testable. What hypotheses =
do you propose, Larry? What is your evidence?
Finally, David Love wrote, "I think arguing about whether Bush won or =
not absent any real evidence is unproductive." I agree. However, the =
exit poll data *are* evidence. And contrary to the suggestions of a =
few, it *does* matter what happened. I am the first to admit that Bush =
"won" the election and that it is pointless to dispute the election =
results. Bush will serve another 4 years, whether I like it or not. =
However, the bigger issue is election reform. Remember, everyone (you =
especially, Larry), the pendulum swings in two directions, not one. =
There will probably be a huge backlash of liberalism in decades to come, =
and the Republicans will then be out of power. Do all you Republicans =
REALLY want an election system that can be rigged by the party in power? =
Hey, just imagine paperless balloting, with no possibility for recount, =
in a political environment such as in the latter FDR administration, =
only with more rampant corruption. Do y'all REALLY want that? Think =
about it! Now is the time to nip all this paperless balloting in the =
bud, before our "Democracy" becomes a "Demockery."
Hey, all you veterans, didn't you fight wars in defense of Democracy? =
Where's your fight now, when Democracy is arguably on the chopping block =
in our own country?! I'm not suggesting you oppose Bush. I'm =
suggesting you oppose paperless balloting with no accountability!
Peace,
Sarah
PS Larry: "You wrote If you can't see that CBS blatantly attempted to =
do harm to Bush during an election, something is seriously wrong." In =
fact CBS is owned by Viacom, which has repeatedly stated it would prefer =
Bush as president (with deregulating and such...). Ya gots ta' cater to =
tha' boss, ya' know, and that's why CBS wore a muzzle and suppressed =
damning stories in the final stages of the election.
---------------------- multipart/alternative attachment
An HTML attachment was scrubbed...
URL: https://www.moypiano.com/ptg/pianotech.php/attachments/6b/3e/33/aa/attachment.htm
---------------------- multipart/alternative attachment--
This PTG archive page provided courtesy of Moy Piano Service, LLC