13, 000(?) National Non-responders; 70,000 (?) State Non-Responders

What does Mitofsky know about these people?

TIA recently quoted Mystery Pollster as having stated (prior to the election) what pollsters are supposed to observe about non-responders. According to Mystery Pollster, pollsters are supposed to record the gender, race, and approximate age of non-responders.

If I recall correctly, response rates were about 50% which would make the non-responders number about 13,000 nationally and 70,000 statewide.

Until a few days ago, I presumed we knew nothing about these people. But if Mystery Pollster is right, we should know enough based on gender, race and age to make some reasonable predictions of who these people were.

And I would think that predictions based on this information would be far more inclined to support or refute rBr, than say, the analysis that Lizzie recommends.

So why didn't Mitofsky release any stats on these people when he posited the rBr hypothesis?

In a court of law, if you have information that only you possess and you refuse to share it with the opposing side or the jury or the judge, the presumption is that such information is harmful to your case.

I believe that should be the presumption here until such information is forthcoming.

And not to TIA's statement.

Mike

http://www.mysterypollster.com/main/2004/11/exit_polls_what.html

Third paragraph under how exit polls work.

TIA quoted MP exactly.

What he states could be just as true for the LA Times and other exit polls, but not for NEP

Mike

He is discussing Mitofski.

Still no word...

:rofl:

N.B. It's this type of post that makes it so obvious that there was election fraud. Any one with eyes to see and ears to hear plus a brain comes to this conclusion quickly after a few hours of review.

The presumption that EM have something to hid screams out from all the post electoin interactions.

The government declares eminent domain when it wants to build a major public works project that it argues is essential. Why can't EM be compelled to give up this information. The detail collected would answer the question that has already been answered by TIA statistically.

Choice 1:
"...obvious that there was election fraud. Any one with eyes to see and ears to hear plus a brain comes to this conclusion quickly after a few hours of review."

<Yup, look at the big picture, take EVERYTHING into consideration, etc. Now, THAT sounds like common sense.>
--------------------------
Choice 2:
"When the philosopher's* argument becomes tedious, complicated, and opaque, it is usually a sign that he is attempting to prove as true to the intellect what is plainly false to common sense."
Edward Abbey
http://www.giga-usa.com/quotes/topics/common_sense_t001.htm

<So, instead, let's complicate (and thus obfuscate**) it, and focus on only one specific idea (rBr), which can NEVER possibly be proved or disproved through direct data/observation.>

*I would also think this applies to the statistician - especially if all they have to work with is inferred data. ;)

**Definition - obfuscate: "To make so confused or opaque as to be difficult to perceive or understand: “A great effort was made . . . to obscure or obfuscate the truth” (Robert Conquest)." <I think a "great effort..." is being made promoting rBr too. ;) >
---------------------------

Now which approach is more likely to reveal the truth?

rBr is about as real as George and Martha's son in "Whose Afraid of Virginia Wolf" (The Albee quote above is perfect).

George (in a rage to Martha after she brings up the fictitious son):

"about the sprout, the little bugger, our son. If you start in on this other business, Martha, I warn you."

The process of exploding fictions can, indeed, generate a great threat to the intentionally or unintentionally mis-informed.
:hi:

Auto, I was thinking about you when I posted this

Thank you very much."

Here is my favorite line of the month on DU (from the initial post):

"In a court of law, if you have information that only you possess and you refuse to share it with the opposing side or the jury or the judge, the presumption is that such information is harmful to your case."

It combines the clear logic of the election fraud accusation with the ultimate goal, a court of law!

:hi:

But if you are so inclined, I am wondering if we can dig up enough exact (or at least close) numbers on the non-responders so that you can make some reasonable predictions as to what the make up of non-responders had to be to produce a GWB victory.

Some possible questions:

First of all, do we know the true numbers of non-resonders?

Was the 50% non-response number for GWB right?

Was the 56% response number for Kerry right?

Assuming those were the numbers, how many non-responders would have had to be men to get GWB's numbers to produce a victory?

What is the likelihood that there were enough men in the non-responders group?

What predictions could you make about race and age?

Just curious as inquiring minds want to know.

Remember, it is ONLY inferred data, and can NEVER be backed up with direct data (ask the rr's why, and who they voted for - that time is forever past). So, wherever the supposed rBr support comes from (National or local) it should NOT be given much credence. The onus should always be on the rBr proponents to verify the hypothesis, and they can't, and NEVER will be able to.
<support for this position here:>
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x363706#363731

So, if we say, "<show me the National/State data and that> "would be far more inclined to support or refute rBr", and they do, we have fallen in a trap. Could Mitofsky be setting us up, and is going to present this data at the meeting in mid May?
----------------------

In contrast, there is much more data (inferred and direct) that supports a stolen election (regardless what Febble, etc. say).

Also, historically, significant exit poll discrepancies have proved election fraud - NOT rDWCr (reluctant "Declared Winning Candidate" responders).

I'll bet that jerk that tried to steal the election in the Ukraine; and was subsequently defeated in a "re-do" (after the exit polls pointed to foul play) wished he would've thought of the rDWCr hypothesis. ;)

It seems to make sense that we should try to prove, statistically if we can, how unlikely it is from the numbers we know.

I understand what you're saying (and it MAY even be unavoidable), but I do have concerns with that approach:

Even though it can never be proved, trying to "disprove it" gives it (implies) more value than the idea really has.

And, if more value than is really there is implied (by the opposition (our "side") appearing to take it seriously), if they come up with more support (even though it is still non-substantive), their case will appear even stronger. (I think part of the second USCV report may have given such an opening to Febble, et. al; however, I do recognize they may have had no choice - considering all the surrounding circumstances/environment. Personally, I would've preferred them to take the "merit-less case" and YOU must "prove" it, it's NOT our job to "disprove" it first approach.)
<Support for this idea here:>
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x363706#363731

I imagine trial attys run into similar situations. The opposition introduces merit-less/inappropriate "evidence", etc., and if it is not "objected" to, and, instead, an attempt is made to refute it, or it is even just accepted, (i.e. taken "seriously"), it is giving more substance than it should have (which is zero).

Just bringing up anything calls some undue attention to it which may inadvertantly give it some unwarranted acceptance.

But TIA is already on the March. See his post a little while ago. I am having trouble following it but he is obviously refuting rBr with it.