Clinton -43% Obama 24% Gore 10%

[DemocratSinceBirth]

http://www.pollingreport.com/wh08dem.htm

[AlGore-08.com]

1. Clinton has negative 43%? Wow, that's bad!

;)

[juajen]

16. No negative numbers I could see for Clinton

Perhaps you were reading a hyphen as a negative? Am I crazy. Where's a negative in this long list of poll questions?

[SaveElmer]

2. That includes Gore...

Without it is 47-26

[Donnachaidh]

3. nice push poll. n/t

[DemocratSinceBirth]

4. Prove It

eom

[DURHAM D]

5. Huh? n/t

[MonkeyFunk]

8. How is it a push poll?

[DemocratSinceBirth]

9. People Don't Even Know What A Push Poll Is

A push poll isn't desingned to "push" a certain response...It's designed to spread false and damaging information about a candidate... A pollster will call and ask people if they would vote for "Candidate X if it they knew he was a wife-beater."

[LoZoccolo]

17. I agree; this is clearly a push poll.

They rotated the names, and that's the least of what they did. The wording of the question clearly favors Hillary Clinton.

[BootinUp]

26. dang, you might be right

:sarcasm:

[rinsd]

27. Umm do you even know what a push poll is?

A push poll is a poll where the surveyed are hit with negatives about an opponent/bill under the guise of being polled.

[zulchzulu]

6. Gallup Poll. Oct. 4-7, 2007. N=488 Democrats and Democratic leaners nationwide. MoE 5.

Small sample, large margin of error...

:puffpiece:

[DemocratSinceBirth]

7. If You Scroll Down You'll See The AP-Ipsos Poll Has Roughly The Same Results

Why is every single poll flawed in the same direction ?

Do you know that's mathematically impossible?

[zulchzulu]

15. A crappy poll methodology is my point

People should look at polls for three things immediately:

1. What was the question and how was question asked (phone/web/print/live)
2. What was the sample size?
3. What's the margin of error?

A good poll will have a good question that is perhaps not just land line telephones during the day. The sample size should really be 1,000 people or higher. The margin of error should be 3% or smaller.

[DemocratSinceBirth]

19. What Are The Chances That Every Single Poll Is Flawed In One Direction?

A statistics professor would tell you that's virtually impossible...He would tell you if they were flawed the results would be random and all over the place...

That's why you combine all the polls...The more polls you have the more robust your results...

That's Statistics 101...

[rocktivity]

10. DING DING DING! Zulchzulu, you're our grand prize winner!

...N=488 Democrats...MoE 5.

The sample's too small (should be closer to 1000), the margin of error is too big (should be closer to 3%). And they know it.

On edit: The Quinnipiac poll, the one with 717 respondents and a 3.7% MOE, has Hillary leading by only 11 points. I rest my case.

:headbang:
rocknation

[DemocratSinceBirth]

12. Why Is Every Single Poll Flawed In The Same Direction?

That's mathematically impossible...If they were flawed the results would be random and you would see some polls with Ms. Clinton ahead and some polls with her behind, but every single poll shows her with leads of twenty to thirty percent...


Anybody that has taken a community college statistics course could decipher that one...

[SaveElmer]

14. Exactly...

They are trying to get us to believe that every random sample is mistakenly weighted in the same direction...

The odds of that are nearly zero...

[SaveElmer]

11. What are the odds then...

Of say 10 polls with a similar margin of error, showing almost the exact same result...

[DemocratSinceBirth]

13. Somewhere Around Zero

Or, for all intents and purposes, zero...

[RaleighNCDUer]

21. And what are the odds of all these HillShills

saying the same things about the same polls?

It would be EASY to fix the polling, particularly with such small samples - calling only land-lines (loses youth); calling only at a certain time of day (concentrates retired); calling only in certain area codes (limit diversity).

Why is it that these polls stand in opposition to everybody I talk to? Why am I not seeing "Hillary" bumper stickers everywhere I go? Six months ago the netroots were talking about how the MSM was trying to make Hillary "inevitable" and y'all were saying that was nonsense -- today, the MSM IS calling her inevitable - I've heard those words at least a half dozen times in the last week.

Y'all are not doing the party or the country any good by trying to annoint her. I HOPE we will not have to see me proved right. Again.

[SaveElmer]

22. Ok...so what you are saying

Is that every pollster is randomly calling the same small subset of people? Do you know how utterly stupid that assertion is...

I hear the word inevitable all the time...it is coming from Hillary opponents who are shaking in their boots that it might be true...

[DemocratSinceBirth]

23. One Of The Many Polls Was Conducted By Peter Hart, For Instance

The Garin-Hart-Yang Research Group, the political division of Peter D. Hart Research Associates, is one of the most respected and successful political polling firms in the country for Democratic candidates. The firm has assisted in more than 400 political campaigns and counts among its current clients 10 members of the U.S. Senate, 16 members of the U.S. House of Representatives, and five sitting governors. Garin-Hart-Yang also has been the lead pollster on the influential Women's Monitor national polling projects for EMILY's List over the past four election cycles.

While we enjoyed much success in 2006--helping to unseat an incumbent senator, governor, and a number of House members, among many other victories--Garin-Hart-Yang has a strong record of helping Democratic candidates win under the most difficult circumstances, as well. For example, in the 2004 election cycle, we conducted the polling and provided the strategic advice that helped Governor Easley of North Carolina become the first Democratic governor from the South to be reelected since 1996. In 2002, we helped then-Congressman Rod Blagojevich win an upset victory in the Illinois Democratic primary for governor against two better-known and more experienced opponents, and then we helped him get elected as Illinois' first Democratic governor since 1976. In 2000, we helped the late Governor Frank O'Bannon win a strong reelection victory while George W. Bush was carrying Indiana by 16 points. In 1998, we were the pollsters for little-known Democratic state legislator Jim Hodges' upset victory over Republican Governor David Beasley of South Carolina. And in 1994, we helped the late Governor Lawton Chiles beat back a tough challenge from Jeb Bush in a strong Republican year.

http://www.hartresearch.com/about/political.html


Is he an MSM stooge?

[LoZoccolo]

18. 5% margin of error? So?

It's closer to 4%, but whatever...none of these three candidates is within 5% of the other ones anyways.

[robcon]

20. Someone who doesn't like the results can nitpick all they want.

The polls all agree that Clinton has a 20%+ lead at this time.

[zulchzulu]

24. It gets down to poll methodology

If it's a tighter race (it will be), would a poll with a sample of 448 people and a MOE of 5% be acceptable?

I don't care about the poll results, I care about the poll methods.

As for national polls at this time, I'd suggest you visit the Gephardt Presidential Library in St. Louis. Oh wait, that frontrunner in national polls in October 2003 wasn't elected...hmmm...

[DemocratSinceBirth]

28. Dean Had 16% The Lieb Had 12 And Gephardt Had 10% At This Time In 03

http://www.pollingreport.com/wh04dem.htm


But you gave it the old college try...



:applause:

[Kurt_and_Hunter]

30. "Acceptable" for what? Every sane person in the world would bet money based on such a poll versus

a random selection.

Sometimes I wish "Margin of error" was replaced by "certainty factor" or some other phrase. The usual interpretation (and I am not attributing this error to you) is that a 5% MOE means that every line in the poll might be 5 points higher or five points lower... bad news for Dodd to find that he is just as likely to be at negative 4% as anything else.

The difference between 3% MOE and 5% MOE is not staggering. You can set the MOE at 3% in both polls by adjusting the level of certainty, or keep the certainty level the same and thereby adjust the MOE.

So if the one was MOE 3% (95%) the other (worse) poll would be MOE 3% (90%, or whatever it works out to)

People see 90% and 95% as roughly equivalent... very high probabilities. But they see MOE 5% and MOE 3% as substantially different, even though the two measures are exactly the same.

[BootinUp]

25. Without Gore: 47% Clinton, 26% Obama, 11% Edwards, 4% Richardson

same link as in the op, scroll down.

[emilyg]

29. Yes.

:applause:

[Turn CO Blue]

31. Polls don't mean anything in caucus states

since it's a negotiation instead of an election.