...it turns out that the 2004 election was a "statistical dead heat". With just over four months remaining until voters weigh in at the polls, the new survey out Tuesday indicates Obama holds a narrow 5-point advantage among registered voters nationwide over the Arizona senator, 50 percent to 45 percent. That represents little change from a similar poll one month ago, when the presumptive Democratic presidential nominee held a 46-43 percent edge over McCain. CNN Polling Director Keating Holland notes Tuesday's survey confirms what a string of national polls released this month have shown: Obama holds a slight advantage over McCain, though not a big enough one to constitute a statistical lead. "Every standard telephone poll taken in June has shown Obama ahead of McCain, with nearly all of them showing Obama's margin somewhere between three and six points," Holland said. "In most of them, that margin is not enough to give him a lead in a statistical sense, but it appears that June has been a good month for Obama."
From the National Council On Public Polls:Certainly, if the gap between the two candidates is less than the sampling error margin, you should not say that one candidate is ahead of the other. You can say the race is "close," the race is "roughly even," or there is "little difference between the candidates." But it should not be called a "dead heat" unless the candidates are tied with the same percentages. And it certainly is not a “statistical tie” unless both candidates have the same exact percentages.
From CNN:
And just as certainly, when the gap between the two candidates is equal to or more than twice the error margin – 6 percentage points in our example – and if there are only two candidates and no undecided voters, you can say with confidence that the poll says Candidate A is clearly leading Candidate B.
When the gap between the two candidates is more than the error margin but less than twice the error margin, you should say that Candidate A "is ahead," "has an advantage" or "holds an edge." The story should mention that there is a small possibility that Candidate B is ahead of Candidate A. (CNN) — With the dust having finally settled after the prolonged Democratic presidential primary, a new CNN/Opinion Research Corporation poll shows Sens. John McCain and Barack Obama locked in a statistical dead heat in the race for the White House.
p.s. Yes, I was being being slightly facetious with the headline.
The poll, conducted June 26-29, surveyed 906 registered voters and carries a margin of error of plus or minus 3.5 percentage points.
7.01.2008
And Here I Thought John Kerry Lost...
@ 7:56 PM
Subscribe to:
Post Comments (Atom)
105 comments
Yea I noticed that earlier. It said dead heat but Obama has a pretty clear lead in their poll. Something tells me that this election is going to be a blow out with the press trying to spin it so that it's close all the way to election day.
All the recent polls carry the headline "dead-heat" with the candidates being in a statistical tie ; however, none of the articles mention that in all of those
"ties" Obama is in the lead.
Are you just noticing this now? The news agencies have always done this, even with polls during the primaries.
CNN always plays fast and loose with adjectives. Back when Hillary was still in the race, it said Obama's 6 point lead over McCain in some poll was "narrowly ahead", while Hillary's 7 point lead was "a wide margin".
Clearly, somewhere between 6 and 7 everything changes.
CNN, where "fair," "unbiased," "equal coverage," means everything MUST be 50-50 regardless of reality!
What I'm hoping for is to watch John King with his magic fingers saying "Well, Anderson, what we have here, is a route. Americans clearly want Change, and Barack Obama convinced them that he was the one to deliver it."
And then the pundits: "Yes, John. But it was so close just yesterday? What went wrong with the polls? Let's go to Bill Schneider to find out."
Bill: "Well, Anderson - as it turns out, the numbers were close, but they've consistently favored Obama! We've been calling it close because of the margin of error, but if you look all the way back to June, Obama's been leading by about 5 points or more for the entire election season!"
It will be fun to watch them relay this news as if it were news. We all, however, will happily send Nate boxes of chocolate, etc....
Soooo, maybe we need some new language for CNN and its ilk.
"Based on the 3 to 4 point lead that Obama consistently shows in the polls, we are unable to say that there is an underdog in the election fight. And there's also no overdog. Intead, it's dog eat dog."
The real dogs in this race are the mainstream media. And each one is trying to eat the other one's Milkbone.
Thanks for the guidance. Media outlets are more willing to call the race a dead heat because the suspense of an even match keeps up interest. You should compare the electoral projections of news sites with electoral projections of non-news sites. The last time I saw a comparison, there was a pretty incredible difference.
So, in the last month 7% of voters moved from undecided to decided in a 2:1 ratio for Obama, moving his lead outside the margin of error.
There's a dead heat for the press.
While you're at it, you could take a shot at CNN's ridiculous categorization of New Mexico as supposedly leaning towards McCain. They fail to justify this by telling us what the hell their reasoning is.
BTW, the headline on cnn.com/politics currently says "CNN Poll: Obama, McCain in statistical dead heat"
Weren't there people arguing in yesterday's thread on 'credible' poll sources that polls for the media may be considered unbiased? Well, maybe the polls itself are unbiased, but the media does surely have an interest in getting out certain poll results, namely such that show the race to be close.
Nate:
Hardly on topic, but I wanted to know how often the Senate Permathread and the Senate Poll Detail are going to be updated.
Yea, that ridiculous right-wing media...
The idea is to update the Senate polling once a week, usually on Sundays or Mondays.
To paraphrase the National Council on Public Polls: "Statistical dead heat" is, for all intents and purposes, a bulls**t term.
Aside from the media's "fuzzy math" on things like margin of error, they don't understand that a preponderance of polls all pointing in the same direction suggest that the race is moving in that direction.
It's much easier to take one poll and call it a trend and this appeals to the 24/7 media's fundamental laziness.
This reminds me of the 2006 Senate race in Rhode Island. Almost every recent poll showed Whitehouse leading Chafee by about five to seven points. But very close to Election Day, a poll came out showing Chafee down (or up, I can't recall) by a point.
Sure enough, all of the "experts" in the media jumped on the poll as a sign that Chafee was "closing the gap." He wasn't; the poll was an outlier and he lost by - say it with me, kids - seven points.
""They fail to justify this by telling us what the hell their reasoning is.""
CNN has Iowa as a tossup too.
Let's not entertain CNN though...they can't even get their ticker headlines grammatically correct.
CNN claims its estimate "is based on several factors, including polling, voting trends and ad spending."
I wonder what they mean by voting trends. I can only guess that, if you drew a trendline from, say, 1992 to 2004, you could make the argument that New Mexico is trending Republican.
Or maybe they're just assuming that New Mexico is Arizona Jr. and will vote like its papa.
well, when you step outside Olbermann, Stewart and Colbert, it is an odd world. Of course the above are partisan (and amusing) but the CNN/ABC crowd is on crack. it is indeed an odd world where Obama's consistent 5-15 point lead for the past months is spun as a dead heat, just as the press touted Clinton's viability for two months after she was couldn't win by the numbers.
It is mainly, I think, a horse race thing. Since how boring is it to find different say McCain is consistently 5-15 points down in the polls and cannot crack 45%?
A lot can happen between now and November, but anyone who follows the polls knows if the election were held today, it would be President Obama, probably with a Margin between Bush 88 and Cinton 96.
Any halfway objective person on this board, whatever his/her ideology, knows this.
This has been a pet peeve of mine for a while. People have a horrible misunderstanding of the idea of margin of error. When you say the margin of error is +-3%, all you are saying is that the polling results show that the numbers would be within 3% of those in the sample 95% of the time.
Now, if you have a polling result that shows Obama beating McCain by a small percentage that is within the MOE, all that means is that the confidence in that lead is less than 95%. It does NOT mean that the race is "statistically tied." Even a 1% lead in the polling has some significance, even if all it means is that the poll thinks that there is a 60% chance Obama has the lead.
I really wish someone would take the time to explain this simple concept to those who report on things.
If you actually read the whole thing rather than just Nate's emphasis you'd see they use the exact phrases recommended by the pollster's handbook:
"...held a 46-43 percent edge..."
"...holds a slight advantage..."
"...Obama ahead of McCain..."
and the commentary clearly provides useful context:
"...confirms...national polls released this month..."
"...not...a lead in a statistical sense..."
"...good month for Obama..."
CNN's only failing seems to be an overeager leading paragraph, most likely re-written by the editor to get more attention.
But I suppose reading the whole article is too much to ask for those eager to expose the anti-Obama media conspiracy.
Dead heat = ratings
Obama ahead = boring news
Therefore, the race must be tied, or else how do you sell ad space?
Gotta love the MSM integrity.
Anonymous at 8:56:
So if an article says "the defendant probably killed his wife" in the first paragraph but says "there's no evidence the defendant killed his wife" in the second, I am to assume that the gist of the article is that the defendant is innocent?
Per the "pollster's handbook," there is nothing - nothing at all - to suggest that the CNN poll shows a "statistical dead heat." So why say it at all?
Someone correct me if I am wrong, but as I recall the margin of error varies with the confidence interval. A poll that has a 3% MoE with a 95% CI will have a lower MoE if you are willing to drop the CI. Since this is not science the 95% convention need not always apply.
In other words, if you are willing to be only 80% certain, and not 95%, then Obama is clearly leading in all the polls, over and beyond the MoE.
Off topic, but http://politicalwire.com/archives/2008/07/01/the_political_futures.html is fairly interesting. The futures markets show numbers right around 538's.
I agree with AxelDC
Throughout June Obama has been opening up a lead that is widening, but the media often talk about ties, close race, etc. They still call states where Obama is clearly leading 'tossups' ...but it isn't a pro McCain bias, it is a pro-horse race bias. There is a lull of political news until the conventions...if Obama had a huge lead, who would want to watch Hardball or the Situation Room?
Totally off topic-
After a day spent meeting patients and reading charts, it's extremely difficult to adjust to the MSM acronym relating to the media. At least, I'm assuming you're referencing the media.
Trevor Howard -
Your sublime grasp of logical analogy has defeated me. I see now that CNN's use of idiom is equivalent to reporting that black is white. Did you clerk for one of the Supremes?
Now I must away to my CNN job, where I am late for my McCain talking points conference call.
In today's political landscape, a "true" 5-point lead would be far from "narrow". If you assume 150 million potential voters or so that the sample is representative of (which seems reasonable to me - is it?), that's a difference of 7-8 million more people who prefer Obama.
Kevin - by that logic, 100,000 people is a huge number by any human scale, but it translates into less than 0.1%.
For that matter, an 8 million vote plurality is nothing compared to the roughly 80 million people who will be eligible but decline to vote. Everything is relative. That's why we use percentages and not raw numbers, to put it in context.
I'm a little disappointed by the advice of the National Council on Public Polls. Why encourage vague phrases like "holds an edge" when more precise phrases are available?
I'd love to see articles talk about polls in terms of probability that a candidate has the lead. They would say things like, "According to our latest poll, we estimate that there is an 86% chance that Obama is leading John McCain."
It would completely change the nature of the public dialogue on polls. Gone would be widely-misused margins of error. Instead, people would say "Did you see the latest USA Today poll? It says that there is a 70% chance that Obama is winning, down from an 85% chance two weeks ago."
Unfortunately I think the MSM is too comfortable with the traditional ways of reporting to make this change. But that's what young upstarts like Nate are for. It's amazing what happens to an industry when the barrier to entry is dramatically lowered. Vive l'internet.
I recall having posted this earlier - but I am not 100% sure it was on this site.
the 95% confidence interrval ia at +/- 1.96 (I seem to remember - in any event it's just under 2.00) standard deviations around the sample estimate, assuming the the figure we're estimating is not way away from 50%.
But the probability is 2:1 that the true value lies within one standard deviation from the sample value. So if the sample error is 4%, the odds are 2:1 the the true figure is within 2% of the estimate. Within one St. dev., that I can accept as close to a tie.
Forgive me if I'm wrong, but aren't they technically in a dead heat? With a MoE of 3.5%, that means that there is a 95% chance that Obama currently has between 46.5% and 53.5% of support and McCain has between 41.5% and 48.5% of support. It's possible that Obama has 46.5% and McCain has 48.5% right? Since we can't rule that out with any sort of certainty, it's a statistical dead heat.
Could someone explain that a little more for me if I'm way off?
Nate,
You let me down here. I agree with the general point of your post. Statistical dead heat is meaningless.
But you still act like the confidence interval is some magic number, so that you can say that the candidate is leading if the difference is more than twice the m.o.e., but not if it is less.
You know better. The most likely answer (Asuming no bias) is that the lead is exactly what the poll says it is. Other answers are less and less likely as they diverge from this. The m.o.e. is an arbitrary number chosen for convenience.
I know you're just speaking the standard shorthand, but it doesn't clarify things for the laymen in the audience.
This comment posted at Open Left: http://www.openleft.com/showQuickHit.do?quickHitId=2066
In the comments section a poster asserts that the CNN article is correct because the margin allows for an overlap. or something. Can someone explain this, refute it, or clarify it?
QT
TO NATE: It would be helpful if, in your copious spare time (ha ha), you could take the RCP poll summary from 2004 and apply your methodology (you don't need the 538 regression) to produce a "2004 Election Supertracker."
And send it to Keith O and Chuck T and CNN, and everyone else who seems to be making up myths about previous elections. A lot of data entry, I know, but it would be very instructive, it seems to me.
Is Obama running ahead of Kerry? Compare the red LOESS from 2008 withe the red LOESS from 2004. You could do a nice chart with that, it seems to me. Same with Gore-Bush 2000, which everyone will grant was a close election and of course a case where the EV result was inverted from the popular vote. If Gore-Bush was a "statistical dead heat," then how can Obama-McCain also be one?
(Sorry for double posting this, but I want to be sure you see it.)
Thank you QT, that poster posted exactly what I was thinking of from my days in Political Research Methods class in college.
So CNN is, well, correct and this is a non-issue.
Hey Nate! Did you change the Electoral Vote pie chart to median (as some of us -- meaning yours truly -- recommended)?
Or is the mean just happen by chance to fall at 312.0, same as the mode and (probably) the median too?
Anon @10:26
"Dead heat" implies a 50-50 chance, or at lest something close to it. But if the measured margin is 3.5% and the MOE is 4.0%, the odds that the leading candidate is really ahead are about 9:1. (It's about 19:1 at the MOE, and 2:1 at 1/2 the MOE.)If you want to take the underdog at even money I'm game.
I have several monitors on my desk and one always shows this site. Regardless of what CNN says, this is the one site for keeping track of the polls and the projected winner. The super tracker shows how the respective campaigns are fairing.
I wish that there was an auto-refresh feature when a new posting is made.
By the way, I love MSNBC and especially KO.
Cheers. -Steve in Austin
To Anonymous @ 8:56p and 9:46p (who would seem to be one and the same):
You certainly appear to be the master of the snarky remark, out to score points where there are none to be made.
No one in this thread (or any other on 538.com) has suggested an anti-Obama media conspiracy. If you would bother to read the posts (as you chastise Nate to read the entire article he referenced), the concensus in this thread is that the media plays up the horse race, because controversy equals ratings equals money.
Your attack on Trevor Howard's comment, while (half) witty, clearly shows your lack of intellectual rigor, rather than any failing on Trevor's part. Belittling a contrary argument, while fun, does not impress anyone who regularly posts here. Trevor was suggesting (as others here also have suggested) that CNN's story was incorrect factually, since we can clearly state, for a poll with an MoE of +/- 3.5%, where Obama has a five point lead over McCain, that we have a confidence interval of approximately 90% that Obama has a lead based on that poll. When you have 90% confidence in a result, it is wrong to state that the result has no meaning.
While the stated poll result is the most likely single occurrence, it is far more likely that the stated result is not exactly correct. For instance, it is far more likely that the true result is at least 1% different than that it is within 1% of the estimate.
If you need any reminder that polls have limitations, see this prediction of a Kerry landslide from November 1, 2004.
CNN is not just hyping the horse race, this truly is a horse race that could go in any direction in the next week, let alone the next 4 months.
@ Queen TiYe:
The rason that when the MOE is 4% and the two candidates are, say, 6% apartment it is not within the samplng error is that once you have specified the score for one choise, the other score is not independent and the same math formulas do not apply. (That's way oversimplified but it's the basics.) You would have to calculate the odds for the second candidate given conditional upon the distribution of the first candidate.
To the posters who think this is really a "dead heat" because of the margin of error, that's not quite true. Here's why:
For simplicity, let's consider the 2-person vote share, and assume Obama has 52.6% and McCain has 47.4%.
Imagine the probability that a given voter supports Obama is given by p, and the probability that the voter supports McCain is (1-p). The poll is way to estimate p. It gives us a probability distribution for p, with mean 52.6% and variance given by:
(.526)*(.474) / 906
The standard deviation, which is the square root of the variance, is therefore about 1.65%. The margin of error is usually taken to be two standard deviations, which in this case would be about 3.3%. (The difference with CNN's value of 3.5% might be due to rounding errors and/or CNN excluding people who didn't answer.)
We can roughly assume that the probability distribution is Gaussian (a bell curve). We know that for such a distribution, there is a 95% chance that you are within two standard deviations of the mean. So in this case, there is a 95% chance that p is between 49.3% and 55.9%. Because the lower end of this range is less than 50%, CNN claims it is a "dead heat".
But what we really want to know is: what is the probability that p is greater than 50%? This can be easily calculated, and it is about 88%.
So as above posters have said, what the CNN poll really tells us is that there is nearly a 90% chance that Obama is in the lead. It might be a stretch to call that a "dead heat."
sorry if the end is uniuntelligble - my cutting and pasting were not fully successful - in about 1-2 weeks I'll have my new glasses and will be able to proof better.
counsellorben - your fellow posters explicitly stated that CNN and other media poll reports have downplayed Obama's leads. The point of Nate's post was to chastise CNN for their sensationalism at the expense of statistical accuracy and Obama's image. From outside the bubble, there is clearly a perception on 538 that Obama gets a bad media rap, as there is from McCain supporters and Clinton supporters on their sites.
As to whether CNN is technically "correct" in calling this a dead heat, since "dead heat" is an idiomatic expression with no definitive meaning, it's pointless to take CNN to task based on statistical probabilities that pollsters do not even publish. "Dead heat" is generally taken to mean "too close to call", and a 5-point race in June is certainly just that. Sorry if that Newsweek poll turned out to be a blip, but CNN is perfectly justified in describing this as a toss-up without any insinuation of sensationalism or bias.
But if you are so concerned about intellectual rigor and fact-checking, and I applaud you if you truly are, why don't you start by turning your keen eye for detail to Obama's own website, press releases, issue papers and speeches and help them set an example of factual certitude for those sloppy, ill-informed journalists?
If you love 538 and want to make your own odds and test the election as you see it you will love this site http://election-projection.net/interactive.html
Anonymous @10:26 & 10:32
Yes, the distributions overlap. But that is not how one does statistics... The distribution you are looking for is a joint distribution. In other words, what is the probability of Obama - McCain < 0, is the integral as x runs from 0 to 100 of the probability of Obama gaining X% of the vote, multiplied by the probability of McCain gaining a vote Y% where Y>X. If one assume normal distributions (kind of OK if the race is around 50%), then the result is a standard normal at sqrt(diff in polls)/SD of poll. The standard deviation of a poll with a sample size of 906 is ~1.66, so that number is about 1.35. Std normal CDF of that is 91%. i.e. there is a 91% chance that Obama will beat McCain, today, based on that poll (if the national vote was the thing). That might not be 95%, but if I was betting money, I would not put it on McCain... i.e. it's not a dead heat.
"If you love 538 and want to make your own odds and test the election as you see it you will love this site http://election-projection.net/interactive.html?"
Yes, this is fun. Thanks for the link.
Not only have people correctly pointed out about the 90% accuracy-but when you can look at multiple polls with near identical leads the lead is more likely to be true since it ha been collaborated. If it were truly that close of a race and it were neck-and-neck why the constant +5 polling? If the pro-McCain folks want to look at moe and say its a tie, it becomes more and more illogical as they have to assume the same error in polling for each new poll.
Gabriel@9:25,
Uh oh, there's gonna be some confusion here....
Please note, all, that Gabriel and I (JGabriel) are two distinct and different people/entities.
.
Great call with http://election-projection.net/interactive.html
I just put in all the odds that Nate has on the site and put them in the simulation
I had it do 100,600 simulations with the numbers to see how close it would be to Nate's projections.
I got:
Obama win %-90.5, McCain win %- 9.0, Tie%- 0.5
Expected EV- Obama- 310.5, McCain- 227.5
I wish it gave you PV
Nate has 67.4/32.6 for odds, giving ties in the simulation for the site to Obama it gives 91/9 (same odds as the poll conversation in this post coincidentally)- so that is a big difference. Nate's is clearly much more realistic in my opinion because of his system, although 91-9 sounds pretty awesome to me. I feel as if Obama has already got enough states solidly under his column to be sure to win unless he has a major gaffe but we shall see
I love the site Nate
I forgot to mention how the EV for 538 is 312.0/226.0 and the projection gave 310.5/227.5- so the projection on pure raw percentage to win that Nate gave is pretty close to the EV count to his numbers after his simulations that add in a lot more sophisticated measures for accuracy.
It is interesting that the win percentage becomes so much better with just the pure numbers- which to me means that McCain wins more in the 538 projections because it uses trends that aid McCain in many cases- but as it stands now this looks pretty safe for Obama
To the person who said that "Statistical dead heat" is a meaningless idion, it measn, as you indicated, that the results cannot be distinguish. what appears to be your iweakness in probability and stat hides from you the fact that yes, ideed, these swcore can bew distingusihed and that the distinctionis real.
One more point about the mishegoss about the “statistical dead heat” BS. We are generally now dealing with multiple polls. Let’s assume that we have four polls for a given state close enough in time to be combined. Let’s further assume that each has a MOE of 5% (that means they have the same sample size) and that in each candidate A leads by 4%. According to the CNN (and others) school of data analysis, we have four dead heats. Ooh, is this a close one.
In reality we have candidate A leading by 4% with a MOE of 2.5%. MOEs are inversely proportional to the square root of the sample size, so if you quadruple the sample size you cut the MOE in half.)
Obviously in real life the results would not be so cut and dried, but the point remains.
Yeah, the press is desparate to somehow spin everything in McCain's favor. It's unclear why. Perhaps they think if they report that Obama is ahead, they'll be accused of partiality, so they'll just keep pretending like it's a toss-up.
Morans.
"The story should mention that there is a small possibility that Candidate B is ahead of Candidate A."
This doesn't sound quite right to me. Candidate B is either ahead or not ahead - there's no probability involved. The "small probability" is that the sample (the poll) does not reflect the reality, i.e. that if you were to repeat the poll 100 times, how many times would you get a result indicating B>A when in fact A>B.
Remembering Isabel Lugo's post about adding numbers with error margins:
Isn't it the case that if we have many polls reporting similar leads, that the margin of error of the consensus result is less? That is, if two polls ask sufficiently similar questions, can't we add together the sample sizes and find a smaller MoE for the combined result?
Frankly,
I think I'm fine with them doing this. I'm not sure I want our people overconfident. One sidenote of the finance decision is that Obama needs the net donors to come through for him. If they think its a blowout, they won't give. McCain could actually catch up. If they think its close, they will be supermotivated to give and GOTV. McCain gets his $85m regardless.
CNN Headline, 4/20/04: 'Poll: Bush has lead on Kerry'
The results? Bush 51, Kerry 46 -- MOE +/- 4%.
The '04 story is schizophrenic with the lead paragraph stating "President Bush increased his lead over Sen. John Kerry in a CNN/USA Today/Gallup Poll" and the 6th graf stating "Bush and Kerry remain locked in a dead heat more than six months before the November election."
I'd say CNN has a problem with consistency in their editing standards but that would imply they have standards in the first place. A 5-point spread is a 5-point spread but somehow the "story" is not.
Queen Tiye is right except for one detail: some time during election night, Wolf Blitzer's head will explode.
Anonymous @ 11:06p:
From Wikipedia: dead heat - is a common expression used in horse racing to describe a tie.
Keeping with the horse metaphors, I will beat a dead horse by reiterating that there is 90% confidence that the CNN poll in question shows that Obama is leading now. A lead is not a tie. However, it is June, and the horse race has a long way to go.
I agree with you that many partisans always feel that the media has a bias against their candidate, and that is a truth in politics. As for my vetting a candidate's website, I expect that if I were to take the time to vet both McCain's and Obama's websites, I would find many things that are non-factual. However, I am sure that it would be a full-time project, and I do not have the time for it until, oh, sometime in 2011.
Finally, kudos to judas_priest for bringing Yiddish to this site!
Anonymous @ 9:46:
Yes, I did clerk for one of the Supremes and, let me tell you, everything you've heard about Diana Ross is true.
As for the rest of your post, counsellorben did such an excellent job restating my point - and you are still so unlikely to grasp it - that I see no reason to state it a third time.
Anonymous @ 10:45:
"If you need any reminder that polls have limitations, see this prediction of a Kerry landslide from November 1, 2004."
The webmaster of electoral-vote.com made the same mistake the media has been making - that is, presenting one poll as indicative of the state of the race.
Over at Election Projection, the projection missed only Iowa and Wisconsin; at RealClearPolitics, the projection missed only Wisconsin. Both sites used multiple state polls to make their projections.
You're all off track on this: "dead heat" is what keeps McCain warm at night.
The CNN headline is misleading and is of course intended to provoke interest in the article which itself is perfectly accurate. The reality is that a headline which says "Twentieth poll in two weeks puts Obama ahead by 5" would hardly be eyecatching.
This is not some great conspiracy, it is just lazy journalism. In some ways, I think it greatly helps Obama for the race to be perceived as close. The greatest danger is that Obama's campaign and his supporters get complacent. They will still need to fight every hour of every day to stave off the filth that will be thrown their way over the next 4 months.
Individual polls matter little. The average of ALL the polls shows Obama ahead by somewhere between 3 and 6 points. It's a good place to be but it's far from over.
Bush-Kerry was a dead heat. Bush won by about 2.5%, which was well within the margin of error of most polls.
The Sunday before the election, most talking heads refused to predict who would win, because the margin was too narrow to gauge accurately.
Obama, however, has about a 6 point lead, so he is clearly ahead at this point. All comparison to all other elections provide context, but no predictive value.
I hope people start making up their minds for Obama or McCain soon. The closer the race, the easier it is to rig an election where electronic voting or electronic vote tabulation takes place. Hence, tight races could leave much room for foul play, and history has shown which party that usually favors.
If we're just going to add 3.5% to McCain's number and subtract 3.5% from Obama's then yeah, McCain's leading 48.5-46.5.
But if you're going to do that, you can also do the opposite and say Obama's leading 53.5-41.5.
There's nothing very scientific about either of those calculations and they're about equally (un)likely to occur. But you won't see many people spinning this poll as a "statistical 12 point lead" for Obama.
Thanks for jumping on this. I thought that CNN analysis was absurd! Unfortunately (or perhaps fortunately for CNN -- they love a horse race), most people don't know enough about polling to question the idiocy of their "analysis."
Hey, I know it's off topic, but is there a way to have the percentage win (the stuff over on the left) with a graph showing changes since Obama won the nomination?
Coupled with the polls I think that would be really useful, and even shed some light on the dead heat brou-ha-ha. I'd love to see a running graph of how Indiana, say, has moved.
Any hope for this?
I'll try and post again, since my original post was deleted by the user. I'm not surprised given my "right-leaning" with the hard "left-leaning" of this site, though. Anyway, I was just pointing out that even exit polling (done AFTER the '04 election) showed Kerry had won some of the states he really lost and that he had a 2-4% lead overall nationally. Yeah, we see how that turned out. Obama should be very afraid of his chances, because he's up against a battered and bruised GOP that has numerous issues working against it and very liberal polling showing him with only a slight edge. He will lose by a considerable margin to John McCain come November. Mark my word.
If Obama(or even McCain) had been consistently ahead by about 10 points and had been for months and was still ahead by ten points, but yesterday it had been eleven, CNN's pretty girls and boys would still be reading off the teleprompter "The race tightens...more after this" in the hope that the viewer didn't have a remote control immediately to hand.
So much easier to have a horse race than do any real reporting about the issues reflected by the poll numbers.
anonymous @ 8:07
Define "very liberal polling." If the pollster asks "If the election were held today, would you vote for Barack Obama, the only man capable of saving our country from a robot invasion, or John McCain, the guy who shoves old women down stairs and eats babies?" is that very liberal polling?
Does Fox News do "very liberal polling"? Even their last poll showed Obama up by four points.
Others have noted the most obvious "bias" in CNN's (and other media's) headline writing. They all, especially the cable news networks, have a vested interest in maintaining the picture of a close race, shifting momentum, fading frontrunners, emerging underdogs, etc. If reality can be stretched to support such headlines, they'll do it.
I'll only add a personal observation from having taught statistical sampling and questionnaire design at a couple of major universities some years ago.
Most of my students were social science undergraduates and graduate students. Not an especially mathematically inclined group. But since I'd shared that perspective as an undergraduate, I knew where they were coming from and was more successful than most in getting through to them.
There were, however, two groups that were especially resistant to the inferential statistical methods: law students and journalism majors.
I concluded that each group suffered from the same weakness: their inclination and training inclined them to make generalizations based on antecdotal evidence. In the case of law students it appeared to be an over-reliance on precedent. My students were often willing to cite a case of a medieval dispute involving the ownership of a cow as evidence in making a case, but strongly resisted the notion that 95% certainty in a sample result was sufficient to guide a decision.
In the case of journalists, the preference for an antecdote was tied to the need to "tell a story." A convenience sample of four voters interviewed at length with suitable quotes was far more persuasive to them than a random sample of a thousand voters.
In general, I was impressed by the fact that these perspectives seemed uncorrelated with intelligence. Law students were among the brightest I taught. And journalism undergraduates, while not the best, were frequently bright kids. (Espcially compared to those majoring in "education.")
Their resistance wasn't based on an inability to understand statistical inference as much as an unwillingness to depart from single cases to broader samples.
Anonymous @ 8:08a said "Obama should be very afraid of his chances, because he's up against a battered and bruised GOP that has numerous issues working against it and very liberal polling showing him with only a slight edge."
"Liberal" polling? Public polls may reflect a bias of the pollster or the organization commissioning the poll, but that bias is as likely to be to the right as to the left. Are the polls commissioned by Fox (Not Really) News "liberal" polls? (Though I see someone else made this exact same point while I was typing away)
If you can offer hard evidence of a particular bias among the universe of polls Nate uses, then please do so, as that will only help to improve this site, the purpose of which is to present an unbiased model of this election cycle.
I admit to my liberal/progressive bias, but I am here because the discourse generally is fact based.
If you are "right-leaning," then you are likely to believe in the wisdom of markets. In 2004, the predictive markets favored Bush for all but a few days. Currently, all of the predictive markets favor Obama, with contract prices representing figures very close to Nate's "Win Percentage" tracker.
You are correct that things may change, but are you willing to bet your own money on this election?
Points taken. I can't speak to specific "liberal" and "left-leaning" bias in the polling except that I know you have to have a true representative sample for accurate numbers. If your sampled group has more liberal representative members than the actual population, and you have a result showing a 5-10 point Obama edge, that's pretty much a useless result. I can no more prove that "left-leaning" people have been over-sampled than you (or anyone) can prove that they haven't. It'll show up on Election Day when the proof is in the pudding. I just know (and this is biased too, but it's a stat, nonetheless) that at least 75% of hard-line Democrats I'm familiar with will NOT be voting for Barack Obama. Even if the national figure is much below that, can he really overcome that kind of a loss with "new" votes? No one had a counter to my argument that exit polling showed Kerry winning in '04. How is that explained?
jsh1120:
As a mathematically inclined lawyer (now software developer/lawyer), I must take offense! Fie on you, sir. <~grin~>
Seriously, lawyers are trained as advocates for our clients, and to take a position and defend it unto death. Also, as a general rule, most lawyers have strong analytical skills and relatively weak mathematical skills, a fact which I exploited often when I was a full-time practicing lawyer.
None of which has anythign to do with this thread. Pardon the interruption.
The " National Council On Public Polls" link is incorrect. I assume you meant it to be this ( http://www.ncpp.org/node/4/#12 ).
Counsellorben said:
"jsh1120:
As a mathematically inclined lawyer (now software developer/lawyer), I must take offense! Fie on you, sir. <~grin~>
Seriously, lawyers are trained as advocates for our clients, and to take a position and defend it unto death. Also, as a general rule, most lawyers have strong analytical skills and relatively weak mathematical skills, a fact which I exploited often when I was a full-time practicing lawyer.
None of which has anythign to do with this thread. Pardon the interruption..."
I see no reason that we can't hijack the thread, Ben. It's not like we're the first to do so.
You're right, of course. And what I always found interesting about law students was the strong "analtyic" skills combined with the tendency to make generalizations based on unique, antecdotal information.
In fact, I used to have a quotation from Justice Vinson, I believe, in which he ridiculed evidence in a case argued before the Supreme Court because it ONLY met the confidence test of 95%. :)
Anonymous @ 8:58a
Re: 2004 Exit Polling
You are correct about the 2004 exit polling in the Presidential race showing a popular vote victory for Kerry, and many have raised concerns about the discrepancies, which I will not address.
Of interst is this report on efforts to explain the discrepancy. However, if you focus on Appendix C at pp. 28-29, it appears that the exit polling controversy is a tempest in a tea pot, as the only states where the exit poll winner and the counted vote winner differ are IA and NV, and the differences between the exit polls and counted results are well within the MoE for those sub-samples.
Further, even if you flip those states to Kerry, Bush still wins the Electoral College. Game, set and match.
The 2004 exit polling discrepancy is a good intellectual exercise, but has little or no relevance, since the overall discrepancy was within the exit poll's MoE. As we have bashed about at length in this thread, when you are within the MoE, you cannot take anything definitive from a poll.
"Game, set, and match"?
Boy, I love it when someone is so unctuous when trying to make a point. Almost reminds me of the Democratic nominee for President in 2008.
My point was mainly to prove that there was a problematic shift in the actual electorate versus the exit polling. Understand? If the actual election results showed Bush winning a state with 60% versus 40% for Kerry and the exit polling showed Bush won 53% versus 47%, yes, you're right in that it didn't matter as far as the outcome and Bush carried the state anyway, but it still points to a glaring flaw somewhere. And that is, that the polling done then (and possibly now?) skews one certain way. I can see how you would want to spin your way out of it like every other liberal.
Of course, whatever I gleaned from the report was done so with a grain of salt, anyway, as I saw in the Introduction portion a heavily positioned argument that the election itself could have "possibly" been hijacked by Bush, Republicans, Right-Wingers, etc. I love it when the blind naivety of hard core members of both parties won't acknowledge that their comrades are equally likely of improprieties.
Anyway, I'll return the favor...game, set, and match. I'm champion!
Great job on your site Nate. One thing I would like an educated opinion on, seems most relevant to this post which is why I'm asking here... What are your thoughts on the Keith number that was used on countdown early in the primary season? (margin of error plus % undecided to get an idea of race volatility) Obviously you can't just add the numbers, way too simplistic but is there something to combining these numbers in a way to better understand how a race stands and how useful a poll actually is? Seems intuitive that a poll showing 52 obama 48 mccain is more useful than 46 obama 42 mccain 8 undecided.
Anonymous @ 11:04a
Yes, you are the champion, that's for sure.
You are the champion of the non-scientific approach. It does not matter to you what can be said with confidence using the science of statistics, you are the champion because you can use the word "spin."
Polling is a statistics based approach, and provided that the methodology is sound, we can rely on those results with confidence. But you are the champion because the 2004 exit polls were within their margin of error (meaning once again that we cannot state with enough confidence that in fact Kerry did win the popular vote).
You are the champion because your ignorance says you are.
Congratulations, and enjoy your victory.
Why has no one commented on the fact that this CNN poll shows the Obama lead shrinking to 3 points when the two third party candidates are included? There was much discussion in an older thread that Nader might be helping the Democrat this time around but the 6% recorded for Nader here suggests otherwise.
Is there some way for Nate to project the likely decline in support for Barr and Nader as the election approaches? This seems to be a better approach than simply ignoring them for the two-man race. And, shouldn't you be looking at states other than Georgia for likely third-party impacts?
If it's within the margin of error, it's certainly a stastitical tie!!
The meaning of "statistical tie" is that the difference is not large enough, given the degrees of freedom (sample size minus 2, I think, since it's a comparison test), to state with a 95 % confidence level (significance level) that there is a difference.
If you can say that there is 70 % chance that Obama is in the lead, that is NOT significant. You could argue that 90 % is significant, but it's borderline and most statisticians prefer the 95 % measure.
But of course you can combine surveys without violating any statistical rules, in order to reduce the margin of error. Doing that, with the recent polling, Obama would be in the lead, significantly.
You're right, of course. And what I always found interesting about law students was the strong "analtyic" skills combined with the tendency to make generalizations based on unique, antecdotal information.
Well, part of that is because that's what the law is set up to do. The whole point is to draw bright-line distinctions between cases that are inherently on a continuum. So you have to learn to pick out details of individual cases that matter, i.e., decide which distinctions are difference-making. Because your job is to figure out which cases are the "edge" cases, it's a mode of thought that is uninterested in amassing data and generalizing from it.
That is, statistical analysis is really pretty foreign to the way legal reasoning is set up. Statisticians see the "edge cases" as outliers, and disregard them; lawyers, instead, are precisely interested in focusing on those edge cases. And it's not like this is a weakness or intellectual failing of law, it's just a different approach, based on different interests.
counsellorben,
Don't preach to me about your sound scientific approach. Yes, I haven't come on here quoting this, that, or the other as if it were a courtroom defense. My apologies.
Rest assured, though, that after having attained a master's degree in economics and becoming well acquainted with econometrics, statistics, etc. and also working for 2-1/2 years as a survey statistician for the US government, there is no loss of familiarity on my part of the methods employed to do sampling and draw statistically significant (or otherwise) results.
My argument is more qualitative in nature, anyway. I question results from a source in which the home page exudes a rather liberal bias and a source your cite (the report on the exit polling discrepancy) that was quick to insinuate a jinxed election due to right wing conspirators.
I'm also well aware of how one can slice and dice statistics behind the veil of objectivism yet still produce something inherently skewed to one side or another. How can I prove that you've done it? I can't. Is that what you want? Does that mean it's untrue? Hardly. All my training in (objective) quantitative analysis has only convinced me that you still have to have an objective qualitative aspect to it as well.
In the exit polling argument, you "sliced and diced" the data regarding the IA and NV vote, pointing out that those were the only two states where the edge was statistically significant difference where Bush won the election but Kerry was favored in the exit polling. But what about the states where the edge was statistically significant and showed a skewed shift to Kerry in the exit polling but Bush won anyway. Can you answer me that? Why must you slice and dice to the lowest common denominator in an attempt to evade coming to grips with what was actually happening?
It's okay, counsellorben. I kind of like for you to have this naive optimism about the "scientific" polling done that shows Barack Obama currently with a statistically significant edge over John McCain.
I'm sure those feelings were mirrored by many when Carter held a 26 point lead over Reagan in 1980 early in the polling season and when Dukakis had an 18 point lead over Bush Sr. in 1988.
I wonder how sound the research or "statistically significant" those polls were.
Come on now, you know what they mean. I agree, 5 points, consistently across many polls makes ‘dead heat’ pretty ridiculous. And the phrase may be technically wrong, but since when do the media use precise, correct scientific terminology? But, you know that they are saying is “the lead is within the margin of error,” which is too long to include in a headline. And this is such a fun way to say it! The reason they are even using that silly term is because they’ve been berated in recent years for saying stuff like “Obama is winning by 5 points.” So consider this an improvement.
最值得信賴的搬家公司-元氣搬家，不論是大台北搬家或大桃園搬家，都能為客戶提供最佳的搬家服務。
普普風室內設計網站提供室內裝潢、家居佈置、空間設計規劃、景觀櫥窗展場資訊等服務。
提供婚禮部落格社群討論、新娘秘書、婚紗攝影及傳統結婚習俗等資訊。
晶英酒店為宜蘭住宿的首選，為宜蘭飯店之最。
有網頁設計需求時，千萬不要以價錢去挑選網頁設計公司，以免後續維護困難。
如果在創業前，能多參考在網路上的創業加盟資訊，瞭解各加盟商的優缺點，就可以降低加盟創業的風險，但在網路資訊不發達的年代，無法有這樣便利的資訊，以致我們的創業失敗，即使當時懷孕，不適合搬家，也只能賣屋搬家，上網找搬家公司，找好合適的搬家公司，買了搬家用的包裝材料，開始搬家前的打包，搬離傷心地，還只能租屋，之後在坐月子期間，因為沒有錢，無法到坐月子中心，所以只能由老公代為料理月子餐，後來買了幾次坐月子餐，這一路走來箇中的辛酸，外人很難瞭解。
想起出國留學前，為了籌措到美國留學的費用，白天打工，晚上到托福(toefl)補習班，沒課的時候又兼了幾個家教，除了準備托福(toefl)考試外，因為沒錢委託留遊學中心代辦，申請學校的相關事宜只好自己來，真的很辛苦，還好有一起努力的三劍客同學，其中一個準備到英國留學，所以準備ielts(雅思)考試，另一個同學則是到日本留學，所以到日文補習班提升jlpt的日文考試成績，他因為經濟情況佳，所以由留遊學代辦中心幫忙處理留學的事，一起為未來努力。後來留學回來，現在大家都有不錯的結果，我們因為不錯的英文能力，常有被企業外派的機會，參加企業教育訓練提升專業能力，職位隨著好的工作績效不斷跟著提升，所以即使無法出國留學，也應參加英檢，以提升英文能力，如多益 (toeic)考試，或全民英檢 (gept)等英檢考試，多益(toeic)對於到外國公司工作有幫助，而且有些外國企業會要求提供多益(toeic)的成績，而全民英檢(gept)是國內大學畢業的基本條件及就業必備。
創業時要找專業的會計師，或信譽好的會計師事務所代辦相關事宜外，還要避開所謂的小人，就必需要事前找徵信社做徵信調查，就可以避開這些人，但徵信社總給人不好的感覺，直到後來有機會認識徵信公司的人，才瞭解到現在的徵信社也講專業分工，徵信費用也是訂好的，不會是漫天開價，若能早知道，遇到大的生意合作前，就該找徵信社幫忙徵信一下，就不會搞到要上法院；這時想到同學的際遇，大學畢業那年遇上個有才華的男人，發生了關係，直到新郎結婚新娘不是她，才知自己是第四者，如果在深入交往前，委託徵信公司做一下婚前徵信，找對徵信社，就不會發生這樣的遺憾。
在公司決定要搬家，空間設計要規劃過，身為總務人員的我又忙了，因為光OA辦公家具的挑選就有好多意見，還有地板磁磚及洗手間瓷磚是否要用馬賽克或拋光磚等等，這時只好找一些室內設計公司，聽聽室內設計師的意見及他們的室內設計作品，綜合整理後，再給主管定奪，之後只要再安排配合室內裝潢及室內裝修的進度，解決我挑選室內裝潢挑選建材的困擾，還有辦公事務機的擺設，如影印機、傳真機、印表機等的動線規劃問題也一併解決，所以室內設計真是一門大學問。
看到電視上大S穿著瑪莉蓮的調整型內衣所展現出的曼妙身材，真叫人羨慕，好想也買一件塑身衣來試試，只是那所費不貲，得考慮一下，因為眼前還有創業加盟不當的負債要還，其中公教貸款利率還好，信用貸款的負擔就比較重，還好三信銀行讓我整合銀行貸款的負債，讓負擔輕些，因此還能利用信用卡的贈品—民宿優惠券加上小小的花費，來趟國內旅遊，其中花蓮民宿之旅讓我享受了恬淡的鄉間生活，離開時老闆還準備了貼心的小禮品，最為難忘的旅程，所以決定將這些回憶放在我的部落格，增加收入，好早日還掉我在三信商業銀行的貸款。
SEO服务
SEO
视频聊天
情色小说
anime wholesale
animme products
翻译公司
视频聊天
情色小说
北京翻译公司
美女图片
重庆旅游
重庆旅行社
美女图片下载
美女视频
アダルト
徵信
徵信社
外遇
徵信
徵信社
外遇
情趣用品
情趣用品
免費a片
a片
免費av
色情影片
情色
情色網
色情網站
色情
成人網
成人圖片
成人影片
18成人
av
av女優
性
情慾
走光
做愛
sex
H漫
免費a片
a片
免費av
色情影片
情色
情色網
色情網站
色情
成人網
成人圖片
成人影片
18成人
av
av女優
性
情慾
走光
做愛
sex
H漫
a片
アダルト
アダルトサイト
離婚
抓姦
外遇蒐證
外遇抓姦
外遇
侵權
仿冒
應收帳款
工商徵信
情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣用品,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,情趣,美國aneros,rudeboy,英國rudeboy,英國Rocksoff,德國Fun Factory,Fun Factory,英國甜筒造型按摩座,甜筒造型按摩座,英國Rock Chic ,瑞典 Lelo ,英國Emotional Bliss,英國 E.B,荷蘭 Natural Contours,荷蘭 N C,美國 OhMiBod,美國 OMB,Naughti Nano ,音樂按摩棒,ipod按摩棒,美國 The Screaming O,美國TSO,美國TOPCO,美國Doc Johnson,美國CA Exotic,美國CEN,美國Nasstoy,美國Tonguejoy,英國Je Joue,美國Pipe Dream,美國California Exotic,美國NassToys,美國Vibropod,美國Penthouse,仿真按摩棒,矽膠按摩棒,猛男倒模,真人倒模,仿真倒模,PJUR,Zestra,適趣液,穿戴套具,日本NPG,雙頭龍,FANCARNAL,日本NIPPORI,日本GEL,日本Aqua Style,美國WET,費洛蒙,費洛蒙香水,仿真名器,av女優,打炮,做愛,性愛,口交,吹喇叭,肛交,魔女訓練大師,無線跳蛋,有線跳蛋,震動棒,震動保險套,震動套,TOY-情趣用品,情趣用品網,情趣購物網,成人用品網,情趣用品討論,成人購物網,鎖精套,鎖精環,持久環,持久套,拉珠,逼真按摩棒,名器,超名器,逼真老二,電動自慰,自慰,打手槍,仿真女郎,SM道具,SM,性感內褲,仿真按摩棒,pornograph,hunter系列,h動畫,成人動畫,成人卡通,情色動畫,情色卡通,色情動畫,色情卡通,無修正,禁斷,人妻,極悪調教,姦淫,近親相姦,顏射,盜攝,偷拍,本土自拍,素人自拍,公園露出,街道露出,野外露出,誘姦,迷姦,輪姦,凌辱,痴漢,痴女,素人娘,中出,巨乳,調教,潮吹,av,a片,成人影片,成人影音,線上影片,成人光碟,成人無碼,成人dvd,情色影音,情色影片,情色dvd,情色光碟,航空版,薄碼,色情dvd,色情影音,色情光碟,線上A片,免費A片,A片下載,成人電影,色情電影,TOKYO HOT,SKY ANGEL,一本道,SOD,S1,ALICE JAPAN,皇冠系列,老虎系列,東京熱,亞熱,武士系列,新潮館,情趣用品,情趣,情趣商品,情趣網站,跳蛋,按摩棒,充氣娃娃,自慰套,G點,性感內衣,情趣內衣,角色扮演,生日禮物,生日精品,自慰,打手槍,潮吹,高潮,後庭,情色論譠,影片下載,遊戲下載,手機鈴聲,音樂下載,開獎號碼,統一發票號碼,夜市,統一發票對獎,保險套,做愛,減肥,美容,瘦身,當舖,軟體下載,汽車,機車,手機,來電答鈴,週年慶,美食,徵信社,網頁設計,網站設計,室內設計,靈異照片,同志,聊天室,運動彩券,大樂透,威力彩,搬家公司,除蟲,偷拍,自拍,無名破解,av女優,小說,民宿,大樂透開獎號碼,大樂透中獎號碼,威力彩開獎號碼,討論區,痴漢,懷孕,美女交友,交友,日本av,日本,機票,香水,股市,股市行情, 股市分析,租房子,成人影片,免費影片,醫學美容,免費算命,算命,姓名配對,姓名學,姓名學免費,遊戲,好玩遊戲,好玩遊戲區,線上遊戲,新遊戲,漫畫,線上漫畫,動畫,成人圖片,桌布,桌布下載,電視節目表,線上電視,線上a片,線上掃毒,線上翻譯,購物車,身分證製造機,身分證產生器,手機,二手車,中古車,法拍屋,歌詞,音樂,音樂網,火車,房屋,情趣用品,情趣,情趣商品,情趣網站,跳蛋,按摩棒,充氣娃娃,自慰套, G點,性感內衣,情趣內衣,角色扮演,生日禮物,精品,禮品,自慰,打手槍,潮吹,高潮,後庭,情色論譠,影片下載,遊戲下載,手機鈴聲,音樂下載,開獎號碼,統一發票,夜市,保險套,做愛,減肥,美容,瘦身,當舖,軟體下載,汽車,機車,手機,來電答鈴,週年慶,美食,徵信社,網頁設計,網站設計,室內設計,靈異照片,同志,聊天室,運動彩券,,大樂透,威力彩,搬家公司,除蟲,偷拍,自拍,無名破解, av女優,小說,民宿,大樂透開獎號碼,大樂透中獎號碼,威力彩開獎號碼,討論區,痴漢,懷孕,美女交友,交友,日本av ,日本,機票,香水,股市,股市行情,股市分析,租房子,成人影片,免費影片,醫學美容,免費算命,算命,姓名配對,姓名學,姓名學免費,遊戲,好玩遊戲,好玩遊戲區,線上遊戲,新遊戲,漫畫,線上漫畫,動畫,成人圖片,桌布,桌布下載,電視節目表,線上電視,線上a片,線上a片,線上翻譯,購物車,身分證製造機,身分證產生器,手機,二手車,中古車,法拍屋,歌詞,音樂,音樂網,借錢,房屋,街頭籃球,找工作,旅行社,六合彩,水噹噹,台中隆鼻,果凍隆乳,改運整型,自體脂肪移植,新娘造型,婚禮顧問,下川島,常平,常平,珠海,澳門機票,香港機票,貸款,貸款,信用貸款,宜蘭民宿,花蓮民宿,未婚聯誼,網路購物,婚友,婚友社,未婚聯誼,交友,婚友,婚友社,單身聯誼,未婚聯誼,未婚聯誼, 婚友社,婚友,婚友社,單身聯誼,婚友,未婚聯誼,婚友社,未婚聯誼,單身聯誼,單身聯誼,白蟻,白蟻,除蟲,老鼠,減肥,減肥,在家工作,在家工作,水噹噹,台中隆鼻,果凍隆乳,改運整型,自體脂肪移植,新娘造型,婚禮顧問,下川島,常平,常平,珠海,澳門機票,香港機票,貸款,貸款,信用貸款,宜蘭民宿,花蓮民宿,未婚聯誼,網路購物,婚友,婚友社,未婚聯誼,交友,婚友,婚友社,單身聯誼,未婚聯誼,未婚聯誼, 婚友社,婚友,婚友社,單身聯誼,婚友,未婚聯誼,婚友社,未婚聯誼,單身聯誼,單身聯誼,白蟻,白蟻,除蟲,老鼠,減肥,減肥,在家工作,在家工作,婚友,單身聯誼,未婚聯誼,婚友,交友,交友,婚友社,婚友社,婚友社,大陸新娘,大陸新娘,越南新娘,越南新娘,外籍新娘,外籍新娘,台中坐月子中心,搬家公司,搬家公司,中和搬家,台北搬家,板橋搬家,新店搬家,線上客服,網頁設計,線上客服,網頁設計,植牙,關鍵字,關鍵字,seo,seo,網路排名,自然排序,網路排名軟體,交友,越南新娘,婚友社,外籍新娘,大陸新娘,越南新娘,交友,外籍新娘,視訊聊天,大陸新娘,婚友社,婚友,越南新娘,大陸新娘,越南新娘,視訊交友,外籍新娘,網路排名,網路排名軟體,網站排名優化大師,關鍵字排名大師,網站排名seo大師,關鍵字行銷專家,關鍵字,seo,關鍵字行銷,網頁排序,網頁排名,關鍵字大師,seo大,自然排名,網站排序,網路行銷創業,汽車借款,汽車借錢,汽車貸款,汽車貸款,拉皮,抽脂,近視雷射,隆乳,隆鼻,變性,雙眼皮,眼袋,牙齒,下巴,植牙,人工植牙,植髮,雷射美容,膠原蛋白,皮膚科,醫學美容,玻尿酸,肉毒桿菌,微晶瓷,電波拉皮,脈衝光,關鍵字,關鍵字,seo,seo,網路排名,自然排序,網路排名軟體,英語演講,托福,Toastmaster,汽車借款,汽車借款,汽車借款,汽車貸款,汽車貸款,借錢,借貸,當舖,借款,借貸,借錢,週轉,
^^ nice blog!! thanks a lot! ^^
徵信, 徵信社,徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 外遇, 抓姦, 離婚, 外遇,離婚,
徵信, 外遇, 離婚, 徵信社, 徵信, 外遇, 抓姦, 徵信社, 徵信, 徵信社, 徵信, 外遇, 徵信社, 徵信, 外遇, 抓姦, 徵信社, 征信, 征信, 徵信, 徵信社, 徵信, 徵信社, 征信, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信社, 徵信社, 徵信, 外遇, 抓姦, 徵信, 徵信社, 徵信, 徵信社,
^^ nice blog!! ^@^
徵信, 徵信, 徵信, 徵信社, 徵信社, 徵信社, 感情挽回, 婚姻挽回, 挽回婚姻, 挽回感情, 徵信, 徵信社, 徵信, 徵信, 捉姦, 徵信公司, 通姦, 通姦罪, 抓姦, 抓猴, 捉猴, 捉姦, 監聽, 調查跟蹤, 反跟蹤, 外遇問題, 徵信, 捉姦, 女人徵信, 女子徵信, 外遇問題, 女子徵信, 徵信社, 外遇, 徵信公司, 徵信網, 外遇蒐證, 抓姦, 抓猴, 捉猴, 調查跟蹤, 反跟蹤, 感情挽回, 挽回感情, 婚姻挽回, 挽回婚姻, 外遇沖開, 抓姦, 女子徵信, 外遇蒐證, 外遇, 通姦, 通姦罪, 贍養費, 徵信, 徵信社, 抓姦, 徵信社, 徵信, 徵信公司, 徵信社, 徵信, 徵信公司, 徵信社, 徵信公司, 女人徵信, 外遇
徵信, 徵信網, 徵信社, 徵信網, 外遇, 徵信, 徵信社, 抓姦, 徵信, 女人徵信, 徵信社, 女人徵信社, 外遇, 抓姦, 徵信公司, 徵信社, 徵信社, 徵信社, 徵信社, 徵信社, 徵信社, 女人徵信社, 徵信社, 徵信, 徵信社, 徵信, 女子徵信社, 女子徵信社, 女子徵信社, 女子徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社, 徵信, 徵信社,
帶著簡單的行囊,來花蓮旅遊走走,順便體驗一下各種花蓮民宿帶給你的不同感受,品嚐道地花蓮美食的特別花蓮小吃風味,也可以騎著單車來品味花蓮 ,呼吸著新鮮的花蓮空氣,所以來花蓮住宿就帶著輕鬆的心情吧!!花蓮黃頁擁有各種不同風格的花蓮民宿,歡迎您來花蓮旅遊放鬆心情體驗美麗的花蓮民宿!!
艾葳酒店經紀公司提供專業的酒店經紀, 酒店上班小姐,八大行業,酒店兼職,傳播妹,或者想要到打工兼差、打工,兼差,或者八大行業,酒店兼職,想去酒店上班, 日式酒店,制服酒店,ktv酒店,禮服店,整天穿得水水漂漂的,還是想去制服店當上班小姐,水水們如果想要擁有打工工作、晚上兼差工作、兼差打工、假日兼職、兼職工作、酒店兼差、兼差、打工兼差、日領工作、晚上兼差工作、酒店工作、酒店上班、酒店打工、兼職、兼差、兼差工作、酒店上班等,想了解酒店相關工作和特種行業內容,想兼職工作日領、假日兼職、兼差打工、或晚班兼職想擁有快速賺錢又有保障的工作嗎???又可以現領請找專業又有保障的艾葳酒店經紀公司!
艾葳酒店經紀是合法的公司工作環境高雅時尚，無業績壓力，無脫秀無喝酒壓力，高層次會員制客源，工作輕鬆，可日領、現領。
一般的酒店經紀只會在水水們第一次上班和領薪水時出現而已，對水水們的上班安全一點保障都沒有！艾葳酒店經紀公司的水水們上班時全程媽咪作陪，不需擔心！只提供最優質的酒店上班,酒店上班,酒店打工環境、上班條件給水水們。心動嗎!? 趕快來填寫你的酒店上班履歷表
水水們妳有缺現領、有兼職、缺錢卡奴的煩腦嗎?想到日本留學缺錢嗎?妳是傳播妹??想要擁有高時薪又輕鬆的夜間兼職工作,打工機會和,假日打工,假日兼職賺錢的機會嗎??想實現夢想卻又缺錢沒錢嗎!??
艾葳酒店台北酒店經紀招兵買馬!!徵專業的酒店打工,想要去酒店的水水,想要短期日領,酒店日領,禮服酒店,制服店,酒店經紀,ktv酒店,便服店,酒店工作,禮服店,酒店小姐,酒店經紀人,
等相關服務 幫您快速的實現您的夢想~!!
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店經紀,
酒店打工經紀,
制服酒店工作,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
專業酒店經紀,
合法酒店經紀,
酒店暑假打工,
酒店寒假打工,
酒店經紀人,
菲梵酒店經紀,
酒店經紀,
禮服酒店上班,
酒店小姐兼職,
便服酒店工作,
酒店打工經紀,
制服酒店經紀,
酒店經紀,
菲
梵,
Post a Comment