Yeah, I kinda wondered about this. We'll see if the other polls follow suit. I do believe Obama is ahead, but I don't think it is by that much.

We'll see though in about six weeks. I still think McCain will win. It is hard for me to believe that with everything that is going on the American people will vote for someone with so little experience and with such a liberal worldview. But hey, I've been wrong before, and I'll be wrong again.

Rush has been talking about this the past few weeks. Thanks for the insight. I was getting a little depressed. Even or close up to November, I think McCain wins by 4-6 points.

Look at the internals some more, and you will see that only 76% of the respondents were even registered to vote.

What they are setting this election up for is to make it look like Obama should win easily, and then if McCain wins they can scream "voter fraud" and we are back to where we were in '00 and '04. The problem is that the only voter fraud that truly happened was when Christine Gregoire stole the Washington governor's chair, and the Democrat GOTV machine somewhere in the Midwest: Minnesota or thereabouts, slashed the tires of dozens of Republican GOTV vans. Oh and don't forget ACORN who seems to always find a lot of new and dead Democrat voters.

So let me see if I've got this striaght - the only way Obama gets a 9 point bump is by overrepresenting Democrats by 10%?

Why do I think that if all numbers were equal, McCain would be ahead at this point by 9 or 10 points?

Oh, they're being cute. They reference the vote as "Net Leaned Vote", which is a way of saying that they pressed those who initially said 'I don't know' to go one way or the other, skewing the actual number of undecideds. Also, given the democratic portion of voters in the 2006 election, this poll is claiming that the Democratic Party has increased party affiliation by 42% since 2006, in spite of the public's strong disapproval of Congress.

Note also that the ABC/WaPo people don't release internal support by party affiliation and independents. Hmmmm, I wonder why they would do that .... ?

Wizbang Mantra on Polls

When favoring McCain - "good, scientifc don't you know."

When favoring Obama -" bad, flawed don't you know."

What a hoot the next 5 weeks will be.

Here you go Arcman:

http://www.usatoday.com/news/nation/2005-01-24-tires-slashed_x.htm

It will only make the Obamaniacs worse when the messiah loses in November.

"They stole the election AGAIN! That's 3 times in a row!", will be heard among the sounds of liberal heads bursting as the pressures exceeds the tensile strength of the duct tape wrapped around them.

looking under the covers at the WaPo poll is interesting (I just reviewed all the sampling data and it is slanted as hell). But up-front in their depiction of the overall poll is this tidbit [emphasis mine]:

This Washington Post-ABC News poll was conducted by telephone September 19-22, 2008, among a random national sample of 1,082 adults, 916 registered voters and 780 likely voters. The survey includes additional interviews with randomly selected African Americans, for a total of 163 black respondents.

so in a sample of 1082 adults, 163 were specifically selected based on their being black. Worth noting:

even if NO OTHER blacks were included in the rest of the sampling, that's 15% of the sample...when blacks comprise only 12% of the overall population a 3% OVER SAMPLE for the Dems right there...since the African-American popluation is voting for Obama at near 100%

Yet in the sampling data on "Race" they claim only 12% "African American"...having ADDED 15% of (according to them) "randomly selected African Americans"

the African-American results were ADDED to the other non-racially based results, meaning you would expect that 12% of the OTHER results would naturally be African-American as well...meaning that up to 25% of the WaPo poll was African-American...and the African-American community is voting for Obama at nearly 100%.

Hell..I'm surpised McCain got within 9 points!

Actually, I don't believe any polls JFO. The last two elections showed how flawed they really are. I have no idea why they still waste their time with them. I do like Luntz real time reaction polling when the candidates speak. ww

Poor samples are like "movements"

The "movement" is the crap running down the legs of these biased pollsters who overweight their polls like a sagging pair of diapers.

But who's really surprised at these tactics?

Actually Justrand, what they did with the African-Americans, was they used that group for certain questions, then weighted their response to balance with the desired demographics. So the fact that ABC/WP oversampled blacks is not statistically significant, however it is a point of concern that they chose to include a demographic sector which does not fit the terms of 'pure' RDD methodology.

I am more intrigued by the fact that the ABC/WP people claim 1082 adults, 916 registered voters, and 780 likely voters, when in the internals of their poll it shows that only 76% of the respondents said that they were registered to vote. 76% of 1082 is 822, so where did the other 94 "registered voters" in their poll come from? There is a serious conflict in their claims right there, and bad enough to invalidate everything they claim.

Thanks for pointing out how flawed the poll is. I'm wondering why they keep trying to paint a false picture of Obama being ahead. Are they trying to discourage McCain/Palin voters from voting? Is there a psychology to making people think they should vote with the majority? Does this help them cover up the ballot box stuffing that they are going to get ACORN to help them with? Hmmm... I'm wondering.

They need justification to drum that 'the Sarah Palin bump is over' or 'Its the economy, stupid.' as a preface to a spot where they can fawn over Obama again.

Therese, that's it! They want to depress McCain/Palin turnout, no doubt about it. I do believe that Obama is slightly ahead, but not by 9 points, no how, no way!

Disapproval with Congress is high, DJ, but that has largely to do with Democrats not doing what they said they would, which is basically the opposite of what Republicans want to do. Party affiliation has shifted in their favour since the midterms as support for the President and his policies has plummeted.

Numbers and citations Hyper?

Numbers and citations Hyper?

Ah, c'mon now, SCSIwuzzy. That's just downright mean. Can't you just let him have his moment?

Hyperbolist -

Today's number show that Bush's approval rating is 19% according to American research Group.

17% approve of his handling of the economy - WOW! That's a lot of very rich or very stupid people.

The Democratic Congress has been ineffective for one reason - republican obstruction.

Arcman -

No, the Dems will not shout "voter fraud." That a right-wing trick.

Dems will scream election fraud. There is a HUGE difference. Election fraud is when entire communities are disenfranchised or caged, like Rove is so effective at doing.

Voter fraud is what Ann Coulter did. She was a registered voter in New York, I believe, but she voted in Florida where she was not registered.

What is it with you guys and polls? Do you refuse to even try to understand what you are talking about (or report the numbers correctly), or do you really not understand them?

This poll is weighted from +10 to +16 in favor of the Democrats. That, folks, is what's called stacking the deck.

Umm, no. The poll respondents gave their party affiliation. How would you like them to weight it?

Look at the internals some more, and you will see that only 76% of the respondents were even registered to vote.

Wrong, it's 84.6%. Of the 1082 people polled, 916 are registered voters, 780 are likely voters (72%).

So let me see if I've got this striaght - the only way Obama gets a 9 point bump is by overrepresenting Democrats by 10%?

You say they are over-representing, yet they have 38% responding as affiliated with the Democratic party. Since exit polls have reported that 37-39% of voters were Democrats in the last four elections, on what basis do you conclude that 38% Democratic is over-representing? The difference between this poll and past elections is the loss of Republican-affiliated voters to the independent category. While this drift may not be as dramatic as the poll indicates (I think this poll is a bit of an outlier, myself), there certainly has been a drift. I'll leave it to you to wonder why so many more people prefer to think of themselves as independent than Republican.

They reference the vote as "Net Leaned Vote", which is a way of saying that they pressed those who initially said 'I don't know' to go one way or the other, skewing the actual number of undecideds.

No, the present both sets of data. They ask the leaners to commit so we have an idea of where they might end up when they vote (of course, voting is six weeks away so that can be taken with a grain of salt).

Also, given the democratic portion of voters in the 2006 election, this poll is claiming that the Democratic Party has increased party affiliation by 42% since 2006, in spite of the public's strong disapproval of Congress.

No it isn't. You're lying again. The Democratic affiliation in this poll is 38%, making it a 0% change since 2006. Leaners may be pressed to say which way they lean, but that doesn't mean they are affiliating themselves with the party. You are comparing numbers from different questions, keeping in line with your consistently dishonest poll analysis.

I am more intrigued by the fact that the ABC/WP people claim 1082 adults, 916 registered voters, and 780 likely voters, when in the internals of their poll it shows that only 76% of the respondents said that they were registered to vote. 76% of 1082 is 822, so where did the other 94 "registered voters" in their poll come from? There is a serious conflict in their claims right there, and bad enough to invalidate everything they claim.

It would help if you actually read the questions.

905. Are you registered to vote at your present address, or not?

94 people are registered, but not at their present address. Not tough to figure out.

Anytime a lefty votes it is fraud.

The left will try to steal another election. Only this time, there will be riots because the media will report that whites stole it from Obama. Scary but true. ww

Are they trying to discourage McCain/Palin voters from voting?

Actually, reporting that Obama is so far ahead would likely have the opposite effect. Obama voters would not feel the urgency, and McCain voters would.

mantis: "Umm, no. The poll respondents gave their party affiliation. How would you like them to weight it?"

mantis...they don't just call a random number of people and use ALL the results (thus letting the chips fall where they may)

They DO throw away any results that are outside the ratios of people they are LOOKING for. They were LOOKING for the results they got!

They DO throw away any results that are outside the ratios of people they are LOOKING for. They were LOOKING for the results they got!

So they deliberately eliminate Republican affiliated voters from the survey in order to skew the results? You do understand that pollsters actually strive to be accurate (yes, they often fail), as polls with greater accuracy have a better reputation and more attention is paid to them, right?

And if the polls were deliberately skewed, how come party affiliation shifts up and down with each poll? If they had a pre-determined fake number they wanted, why doesn't it stay static?

And, is your tin foil hat too tight? You seem to have problems with your left finger and the Caps Lock key on your keyboard.

The only mistake I made, mantis, was taking you for a mature adult and serious about the issue.

And yet, you are the one making 'tin hat' jokes. That is truly sad ...

So instead of replying to his response to your attempt at mitigating Obama's clear lead in a poll, you wag your finger at him for making fun of a pretty stupid post--this stuff is your bread and butter, why are you taking your ball and going home, DJ?

I'll post it here since DJ is now disemvoweling my comments on his thread, as he knows he's caught in a lie.

Frank Newport of Gallup writes,

We've had many inquiries and comments about the latest Gallup Poll trial heat results on the presidential race. Our editorial team will be responding to as many of the issues raised as possible here over the next day or two.

One question that comes up frequently (and apparently is based on various statements bouncing around the Net) concerns the party identification of the respondents in our sample. The supposition on the part of some is that these party identification figures from poll to poll should be constant and the same as some standard established from previous polling.

That's simply not the correct way to look at party identification. At Gallup (as is the case for many other polling firms), we ask party identification at the end of the survey using this wording: " In politics, as of today, do you consider yourself a Republican, a Democrat, or an independent? " Our experience tells us that this is not a fixed demographic measure (like age or gender or ethnicity), but rather is a variable in and of itself. While many Americans are hard-core Republicans or hard-core Democrats and never would call themselves anything different, there is a group of Americans who have no firm party allegiance and whose political identification can and does shift during an election season.

In fact, if one candidate is doing particularly well, it is usually the case that more people in the sample will identify with that candidate's party. Thus, if Kerry is having a good period of time in the campaign (as was the case after the Democratic primaries last February and March, and again in June and July of this summer), then more people will identify as Democrats at the end of the questionnaire when we ask with which party they identify "as of today." If Bush is doing better, as he is now, then more people at the end of the questionnaire will identify as Republicans.

Furthermore, there are no Census or official figures on party identification nationally. A number of states do not require party registration, and what a person calls himself or herself can vary significantly from week to week or month to month.

So it is incorrect to say that a poll's showing one candidate to be ahead is the result of the fact that there are too many members of his party in the sample. In fact, that there are more people identifying with a leading candidate's party is a result of the same forces that are pushing that candidate into the lead.

One final note. Gallup (and other reputable pollsters) do carefully analyze the compositions of each sample on known demographic measures for which there are solid Census figures: age, gender, region of country, ethnicity, and education. And we do weight each sample to each of these if necessary, using complex and accepted statistical procedures. So our samples are remarkably constant from poll to poll on known demographic and regional measures. But in a political year we don't expect that samples will be the same from poll to poll in terms of party identification, any more than we expect the samples to be the same from poll to poll in terms of the choice of candidate for whom the respondents are voting.

The original is here, if you have a subscription.

The Freepers reprinted it here.

Mantis, I believe the word you are looking for, is "waaaaaah"

As to your continued - ahem - factually impaired opinion, I answered you fully in the article titled 'Errata'.

Go find an adult to help you this time.

Hyper, no running at all, but I do not repeat myself when it serves no purpose, especially when addressing a boor like mantis.

Hyper,
That only shows one part of your claim, that approval of congress is low.
What about the second claim, that dem affiliation is up based on Bush after the midterms?

Mantis, I believe the word you are looking for, is "waaaaaah"

Actually, the word is liar.

I answered you fully in the article titled 'Errata'.

No, you didn't. You continued to lie:

This is a milder shift, but as I said before, this still shows that the Gallup Organization shifted party affiliation weights on a regular basis, invalidating the statistical impetus of voter responses.

Gallup did not shift party affiliation weights, as they don't weight by party affiliation at all. The response by those surveyed changed, as is known to happen around convention time.

You are a liar, and repeating your lie does not make it any less untrue.

Election fraud is when entire communities are disenfranchised or caged, like Rove is so effective at doing.

Voter fraud is what Ann Coulter did. She was a registered voter in New
York, I believe, but she voted in Florida where she was not registered.

SDW where are your sources to back your allegations against Rove and Coulter?

Well, SCSI, it sure as hell isn't up because of the inept Democratic congress. What other possible explanation could there be? Occam's Razor, and all that.

DJ, calling mantis a "boor" is hilarious. He has been more civil towards you than you to him and seems to know as much about the subject of polling as you do. You have repeatedly asserted that Gallup weights data according to party affiliation, which he denies. I have seen no evidence to support your assertion, and a lack of evidence either way would seem to support his.

You know what's "boorish"? Ending a conversation by removing the vowels from someone's posts when they disagree with you. No, this blog isn't a democracy, but you act less like a benevolent dictator and more like a spoiled brat with all the toys.

maggie, Coulter is currently under investigation for voter fraud, punishable by up to 5 years in prison.

Hyper:
Thanks for the link.

Now, where's the sources for Rove?

And by the way, clean it up.

Ohh noooo! FOX News is now in the tank for Obama too! Check out their poll...Obama is up 6 points on the FOX News poll! They too are leaning for Obama. Waaa waaa waaa boo hoo hoo. Unfair boo hooo.

hyperbolist, I don't bring disagreements out of threads to other places, and I am sure Kim, who took the time to write this thread, would like you to focus on it and not your spittle and frenzy on imagined injustices somewhere else.

You are nowhere near correct in your claims regarding the other issue, but this is not the place to hash it out.

As to Ms. Coulter, investigation is not proof, nor even a valid indicator of concern. To vcoin an old phrase, if that's all you have ...

Uh, I never said Rove committed any crimes, maggie.

A Fox News report that she's being investigated isn't a valid indicator of concern? If you applied that same threshold of proof to Barack Obama you would be voting for the guy. You're hilarious!

I left out all the other links I found because as they are from sources other than Fox or the right wing blogosphere, I knew nobody would bother to read them.

Actually, hyperbolist, I read CNN regularly, and I find a lot of useful information from AP/Ipsos. The source is a concern when it's an opinion piece or is obviously spin, but "investigation" is just that ... a look by the law.

It's not proof, or even evidence.

Hyper:

Since you volunteered for covering SDW in
allegations of Rove and Coulter, I assumed
you would do the same for sources about
Rove.
My bad for assuming.

This is a flawed poll that was spoken about all around talk radio today. The ABC/Wash Post Poll clearly was stacked with leaning members of a party, most specifically African-Americans - of which 90% in the country will be voting for Obama, and it's not for his foriegn policy, well if he had one.
From now on when these polls come out ABC/Wash Post will not be taken seriously - they impeached what little credibility they have as part of the Hussein Obama War Machine.

No worries. If I could find a slam-dunk for indicting Rove, I'd be the Attorney General in the Gore administration by now. (Har dee har har...)

Coulter, though, has her feet in hot water, because she voted in the wrong place. Trivial at face value, but laws is laws. She'd be the first to tell that to any 18 year old kid caught with fifty bucks worth of pot facing a jail sentence, except she'd be more of a jerk about it.

Hyper,
How about proof that Dem registration is up at all, compared to the activity of any other presidential year?

Not Occam's razor, just something other than what you hear in your echo chamber.

Hyper....

Since I've retired and am at home instead of away for 10 hours a day, I've gotten lots of calls from poll takers. The thing is, we have no way of knowing if the answers we give are those that are actually recorded. We just don't know do we? We might all be making false assumptions.