Posted on 08/04/2004 9:03:00 AM PDT by Big Otto
Tracking polls move a bit from day to day. The fact Rasmussen shows Kerry and Bush bouncing up and down a little may simply mean that they are being honest and reporting the normal statistical variations. Although I was getting the feeling they were looking at other polls over the last few days.
>>wow, when they lose in Nov, the DUmmies will say "It's good we lost so we can get stirred up and work harder in 2008."<<
O, yeah, right. Like they'll really admit they lost. I remember in 1984 when people were talking about the silent majority who were so dismayed by Reagen and the "conservative Democrats" that they stayed home. The claim was that Reagen only got under 30%. (50% of 58%)
This is one of those trick questions, right?
Frankly, ABC's poll immediately before the convention *did* set up a false low-water mark for Kerry. Weighing the persistently shows that Bush did not improve before the convention, and Kerry got no bounce after the convention.
Funny... the middle of a sentence got cut:
...Weighing the party affiliation so it tracks consistently shows...
DU'ers are simply stupid people who make us laugh.
From your link, Big Otto.Here's your BUMP, John Fn Kerry !!
Election 2004
Date Bush Kerry Today 46 46 Aug 3 45 48 Aug 2 46 47 Aug 1 45 49 July 31 46 47 July 30 45 48 July 29 45 48 July 28 45 48 July 27 46 47 July 26 46 46 July 25 46 47 July 24 45 47 July 23 45 48 July 22 45 48 July 21 47 46 July 20 47 45 July 19 47 45 July 18 46 46 July 17 46 47 July 16 46 48 Earlier Results for
Dates are release dates
Surveys conducted on preceding three nights
"This is one of those trick questions, right?"
I wish BUT the fact is ABC CBS CNN FOX and NBC still call Hanoi Kerry a war hero.
They leave out a war hero to the Viet Cong.
I agree. I roll my eye a lot. Especially the way Zogby gets bashed. There's a reason these people have jobs as pollsters, and it's not just some big conspiracy.
And by the way, I share your hesitation.
The insanity of this person is revealed by the way he/she identifies with the Kerry campaign.(see bold text) The morons who root for the Boston Red Sox write the same pathetic fashion.
I don't identify one friggin' iota with George Bush... I will vote for him because John Kerry is a nightmare for America if elected.
The Rasmussen JA numbers are high compared to other polls and the horse race numbers low, so 55% would be a good number to aim at, which would translate to 51% or something like that for the other polls.
The question is if things are different in this election cycle and there just aren't any convention bounces to be had (in which case things aren't so bad for Kerry), or if Kerry should have gotten a bounce, but didn't (which is obviously bad). We won't know until after the GOP convention.
I don't know how many have noticed, but I was just over at Rasmussen, and when I see the internals, it does appear that he is overweighting Democrats by 3-4% not only in the horse race but in some state polls. For example, his newly released Iowa poll has Kerry up 3, HOWEVER, Bush gets 1% more R's than Kerry gets D's and they are tied on independents. I remember reading yesterday similar discrepancies in the horse race poll.
" Bush is the choice for 88% of Republicans, 13% of Democrats, and 41% of unaffiliated voters. Kerry is preferred by 81% of Democrats, 9% of Republicans, and 42% of unaffiliated voters."
Here's what the Iowa poll said:
" Kerry has secured the vote from 87% of Iowa Democrats while Bush has done the same with 86% of the state's Republicans. Among those who are not affiliated with either political party, it's tied with each candidate earning 42% of the vote. Last month, Kerry held the lead among Iowa's undecided voters, 43% to 39%."
Haven't looked at all the other state polls but wouldn't be surprised that this trend was prominent in Rasmussen's polling. For the horse race numbers, a balanced weighting should put Bush up by 5 or so. Everyone who reads Rasmussen's polls should be aware of this bias.
Folks,
When one talks about the "accuracy" of a poll, what is meant?
Does that mean that it measured the opinion of the electorate as a whole? Does it mean it measured the opinion of registered voters? Or adults?
Sampling theory is well understood. If you sample carefully, you can project samples of hundreds to a population of millions. Millions will vote in November.
The partisan mix focus of our turnout model discussions is important, but it is just as important to understand that no one knows the right mix. When we berate the LA Times for "oversampling" Democrats, we don't know that they are wrong. That might be the turnout. Maybe 10% more Dems than GOP voters will turn out. If that proves true, we'll lose.
If Rasmussen is using the 2000 election's turnout model, we would expect him to have a different result. Ditto Gallup. The turnout model defined who is sampled and that yields the result.
There is also some considerable doubt that anyone really knows accurately what % of the electorate are registered GOP or an independent who usually votes that way. And vice versa for the Dems. We don't even know the national mix, let along the correct sampling mix. There are a lot of independents out there.
Anyway, what we've always said holds true. If the sampling mix is held constant, changes in the poll are meaningful.
Americans are so gullible. There is no way to conclusively find out the veracity of any of these polls, whether done by a neutral, conservative or liberal polling organization. We just have to "take their word for it".
Instead of Bullwinkle Moose pulling a rabbit out of his hat we've got pollsters pulling numbers out of their @$$ and Americans continually swallow this up like it was chicken soup for the soul. And lo and behold, it's a virtual "dead heat", just like Florida 2000 and just like some want it so that reality remains hog tied.
Can't folks see through this simple ploy or are they just too unwilling to admit that they've been played?
"Haven't looked at all the other state polls but wouldn't be surprised that this trend was prominent in Rasmussen's polling. For the horse race numbers, a balanced weighting should put Bush up by 5 or so. Everyone who reads Rasmussen's polls should be aware of this bias."
I don't understand your point here. Please explain. Thanks.
Reread the post, it should be quite clear. Rasmussen divulged that Bush was beating Kerry by getting more of his base, and beating Kerry by getting more D's, then they are both tied in getting I's. The only way then that the poll can be tied is that Rasmussen is underweighting R's, and by the numbers, it looks like a significant underweight. He was doing the same thing in the IA poll and come to find out in today's polls, he didn't post the internals. Wonder why.
Thanks for the explanation. It wasn't clear to me, which is why I asked for clarification.
Rasmussen's assumptions include a loose LV screen and that Nader won't be a factor, both favorable for Kerry. His national tracking poll appears to be about 2 points favorable to Kerry and the state polls about 4 points.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.