Paligal
(178 posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 02:56 PM
Original message |
Anyone have DEFINITIVE exit poll data? |
|
We are all citing exit poll data from various other posts, yet nobody states where they got their numbers from originally. More confusing, many of these numbers differ from poster to poster. Does anyone know where we can get the DEFINITIVE raw exit poll numbers (as in, not adjusted, as they did after the election). I would like to see the same numbers that the networks looked at on November 2. Can anyone help me? If we stand any chance of being heard, we must have credible evidence, and that means referencing our data source every time that we post exit polls as the support for our argument. Thanks everybody!
|
papau
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 03:15 PM
Response to Original message |
1. SLATE (MSNBC) is source of 2pm and 4pm numbers |
|
Edited on Fri Nov-26-04 03:17 PM by papau
Definitive numbers are a secret- as they are the post 6pm EST results but pre-correction to make them fit actual reported vote totals.
The fellow that did them is claiming that the only numbers officially released are those that he uses to make projections which are only the "post-correction to best fit with "actual" " numbers. (Do not look for logic in these statements) - and those really fit well with the results! LOL :-) :-)
The other earlier numbers "are not his fault"
I do love CYA!
I love even more the ability of our - "not controlled by the right wing GOP - we just act that way" - media to say nothing about this, while talking about the Ukraine exit polls! Media around the world is having fun laughing at our "free press" as they listen to our lectures about our freedoms!
:-)
The analysis done at DU should be read as based on the 4pm Slate released exit poll data unless the post says otherwise.
|
righteous1
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 03:19 PM
Response to Reply #1 |
|
The primary reason nobody releases raw numbers is because they are exactly that RAW. They mean absolutely nothing, and unless you know the model to apply to those numbers they can be extraordinarily misleading. They are about as accurate as if you called 10 people randomly in your city and asked them who they voted for pres. and then took those numbers and determined who won the election
|
papau
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 03:28 PM
Response to Reply #2 |
4. I agree the model is key - but there was a model at 2pm, and at 4pm |
|
that produced results before the model was tweaked to get best fit.
Now why do folks think that a tweaking that is outside of MOE on a dozen key states might need a review?
I guess too many folks had a course or two in Stat while in University!
:-)
|
Paligal
(178 posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 03:35 PM
Response to Reply #2 |
5. But isn't that the data the networks use? |
|
The raw exit polls are raw, but haven't they been correct to a couple of points in the past? Why not now? If I called ten people randomly, and got half for Kerry, half for Bush, it would actually give me an idea, yes. And if then they reported that Bush won 70% and Kerry 30%, let's say, I would wonder. Ten people is not a big enough sampling, obviously, but if we sample a thousand RANDOMLY, it should give me a fairly good idea, actually, I would think.
|
righteous1
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 04:52 PM
Response to Reply #5 |
|
because you could have "inadvertently" called 7people that happen to be evangelical Christians. Well maybe two of those voted for Kerry but they should have voted for Bush. Would indicate landslide Kerry but unless you modeled them it might just break 50/50. Crude example but it illustrates my point
|
jamboi
(1000+ posts)
Send PM |
Profile |
Ignore
|
Fri Nov-26-04 03:22 PM
Response to Original message |
3. Zogby called for the release of these on Countdown w/ KO the other night |
|
And for a complete investigation of both this and election fraud. He called for a "blue ribbon panel". Also there have been fairly complete writeups of the non-"corrected" exit polls on progressive websites. Look around and you'll find them.
|
DU
AdBot (1000+ posts) |
Fri Apr 26th 2024, 09:30 PM
Response to Original message |