What a Bounce: Bush and Kerry tied again

[dotComa]

http://www.rasmussenreports.com/Presidenti...acking_Poll.htm

Presidential Tracking Poll shows President George W. Bush with 48% of the vote and Senator John Kerry with 47%. The Tracking Poll is updated daily by noon Eastern.

Our latest data for Minnesota shows the President has moved into a tie for that Battleground State's 10 Electoral Votes. In California, Kerry still leads by eight.

Other data released yesterday shows that Zell Miller is now viewed through and entirely partisan prism--Bush voters love him and Kerry voters hate him. We also provided a brief analysis of the Bush Bounce in the polls.

Today, at 3:00 p.m. Eastern, we will provide commentary and analysis as to why our polls are currently showing a different result from the widely reported Time and Newsweek polls. At that time, we will also provide an update on who Americans think is winning the War on Terror along with other perspectives on the situation in Iraq.

-------------------------------------------

Rasmussen contends that the time/cnn poll was not a statistically valid group and when he fixed the data, the giant bounce Bush got was more like 3%.

[bishop]

personally, i was a bit in disbelief that bush supposedly received an 11% jump in his poll numbers immediately following his speech. he offered very little in terms of new ideas and details in his speech - i don't understand why the media gave it as much praise as it did.

you couple hypothesize about this poll almost endlessly.. but in general, i would simply say that the samples used in these polls are often too small to generate truly accurate statistics. if i'm not mistaken, their poll uses a sample of 3,000 voters.

i've read that voters are registering to vote at all-time levels for this election. many of these people are young, who generally vote liberal. just another reason to doubt the 11% jump in the time/cnn poll.

[Compugasm]

Originally posted by bishop,
if i'm not mistaken, their poll uses a sample of 3,000 voters.

That's actually quite a few. Most use barely 1k.

[katar]

Methodology: The TIME Poll was conducted August 31 – September 2 by telephone among a random sample of 1,316 adults, including 1,128 reported registered voters and 926 likely voters. The margin of error for registered voters is +/- 3% points, and +/- 4% points for likely voters. Schulman, Ronca, & Bucuvalas (SRBI) Public Affairs conducted the poll.

I am curious as to whether newly-registered voters (first-time voters) will have an effect in this election... Media pollsters wouldn't have these newly registered folks in their database. If this is a sizeable demographic (i don't know if it is), any poll is necessarily flawed.

I can't imagine there are many former Gore voters who will now vote for Dubya-- despite Bush's hair's breadth victory, American policy has radically, dramatically shifted in 4 years.

[Compugasm]

Originally posted by katar,
I am curious as to whether newly-registered voters (first-time voters) will have an effect in this election...

Maybe, were these highschoolers, aliens... What city, was this a phone survey, where did the phone numbers come from... blah, polls.

[Scribbler1]

Originally posted by katar,

I can't imagine there are many former Gore voters who will now vote for Dubya-- despite Bush's hair's breadth victory, American policy has radically, dramatically shifted in 4 years.

Now THAT is an EXCELLENT point, and one which has not been discussed much, as far as memory serves. Considering they didn't like Bush the first time, and many feel Bush stole the election from their man, I daresay they would probably vote for an overripe canteloupe if it ran against Bush.

Figuring the likelihood that most of Gore's supporters will vote for Kerry and a lot of Republicans (we have a couple here) are very distressed about Bush's actions, you have to wonder how valid these polls ARE. In fact, were they EVER valid, or did we just take them as such because we were told how "scientific" they were.

A lot of people lie, just for the sake of lying. Others lie for various valid reasons, and a lot of people will answer a question with what they THINK the questioner wants to hear.

We are constatntly told how accurate the polls are, by the pollsters themselves, but has any unbiased party ever examined the methodology of these polls and how valid the results are?
My feeling is that in the long run, polls mean squat except to those who use the results for their gain.

[Howard Roarke]

USA TODAY has Bush up 7
Newsweek Bush up 11
Time Bush up 11


Democrats would love for this not to be so, but it is, Bush has taken a convincing lead in the race.

Bush got a real bounce, Kerry didn't so now the Democrats spin, spin, spin.

[bishop]

i'll second that.. great point katar. (on a side note, it's very nice to be in the company of so many intelligent people here.)

imo, hardly anybody that voted for gore last time around will vote for bush. many of the former nader supporters seem like they're going to vote for kerry. young people are registering to vote at all-time highs - and history shows that they vote in a liberal direction. thousands of blue-collar workers are struggling now, more than ever before, and they traditionally vote democrat. (or at least it's logical for them to put some of the blame on the current leader for failing them.)

i do know ONE person who voted for gore the first time who's going to vote for bush. although, i know this person very well, and i can (unfortunately) say that this decision was largely guided by her boyfriend.

on the other hand, members of my family, who have NEVER voted for anyone but republicans plan to hold their noses and vote for kerry. for my grandfather, this will be the first time he's voting for a democrat. for my father, it'll be the second. for my mother, who's voting history is mixed/independent, she will either vote for kerry or abstain.

personally, what little credence i used to give electoral polls has evaporated. the only poll that matters is the one in november. in the meantime, everyone should buckle down and do what they can to support their candidate - and simply throw these poll numbers where they belong - in the garbage.

[Compugasm]

3min ago on the radio, they said it was dead even. Looks like there is an 11 point margin of error.

[bishop]

maybe we should just square the margin of error? +/- 16 points. that sounds a bit more accurate.

just wondering, these pollsters, do they divulge any additional statistics on their polls? do they give any info on variance, sample frequencies, etc?

[dotComa]

Originally posted by Howard Roarke,
USA TODAY has Bush up 7
Newsweek Bush up 11
Time Bush up 11


Democrats would love for this not to be so, but it is, Bush has taken a convincing lead in the race.

Bush got a real bounce, Kerry didn't so now the Democrats spin, spin, spin.

Howard, Time and Newsweek are from the same two polls who were incorrect.

CNN, and USATODAY have both said Bush jumped only two points from the convention. Kerry went down 5 on his own (7 point gap)

http://www.gallup.com
http://www.cnn.com/2004/ALLPOLITICS/09/06/...poll/index.html
http://www.usatoday.com/news/politicselect...9-06-poll_x.htm

[Scribbler1]

I have debated the relative worth of polls in a number of forums, and the answers were all pretty much the same. People parrot what the pollsters say and declare how scientific these polls are. As for me, knowing how diverse this country is (I believe the U.S. hs FIFTY-TWO registered political parties) and how people can change their minds at the drop of a hat, I don't believe asking 1,000 people about something in a country wilh a population of over 250 Million people (yes, I know they are not all voters, or even adults) can even REMOTELY qualify as a representative sample. You can use all the science you want, but with relatively NO raw data you don't have jack.