"The Economist" Poll: Kerry 45, Bush 44

[sonicx]

http://www.economist.com/media/pdf/YouGovI.pdf

btw, it says 'recruited panel.' what does that mean?

[da_chimperor]

1. The best source of information EVER

That and they have the big mac index.

[TexasBushwhacker]

6. Yeah, I wonder what the numbers are for 65+ now

This poll was done BEFORE the big announcement that Medicare premiums were going up. Age 65 and over was 56% for Bush, 41% for Kerry. Do you think it will be the same next week?

[DaveinMD]

2. good news

but I doubt we are in the lead right now. I trust Kerry's internal polls better than any private polls.

[da_chimperor]

9. My best guess on the 'recruited panel'

Is that the sampling was more careful rather than just cold-calling random people. They were trying harder to try to get a representative sample of the American population. That also may explain the low margin of error of 2%. This is just from what I remember about statistics. Anyone with a better idea is free to chime in at any time.

[papau]

13. This is the Census procedure as I recall - a sample is taken via the usual

random telephone selection process.

The sample is asked if they are willing to be sampled over a period of time (in this case until the election).

A very large number (10,000 to 40,000) yes's are obtained, and that group randomly reduced to a representive group of 10,000 or more (this is the same procedure used in stratified telephone samples)

The group is then a panel.

the panel is then sampled with 2000 to 4000 samples taken with replacement from the panel.

The only new step relative to a "National poll" is that those in the panel have agreed to be sampled through a given point in time.

Like any poll that is valid, you can not just opt in.

It tends to be a vaild indicator - esp. about trends!

It is the new, more scientific basic being installed for TV ratings by Nielson - which Fox objects to. When Fox objects to something that tends to validate it in my opinion!

:-)

[ruggerson]

3. Was this released today?

How come no coverage on the cable channels?

[Doosh]

5. Frances

[olddem43]

10. It doesn't fit their preconceptions or their wishes.

[Thurston Howe IV]

12. Why no coverage

I'm not 100% positive, but I believe the Economist is published out of England, so it wouldn't get much coverage here. The sample size is larger than most surveys, but it seems like it ended before Bush's speech. I believe it takes a day or two after the convention before the effect of the convention is fully measurable, so this survey may not reflect whatever "bounce" Bush might have got.

[UrbScotty]

14. Welcome to DU!

:-)

[saracat]

4. This appears fair and more reputable!

And not because of the 1% lead. The nature of the questions is signifigant.

[Downtown Hound]

7. Good news

And they use huge samples in their polls too. None of this let's poll 500 people and have most of them be Republicans and ex-military like certain supposedly liberal magazines.

[Florida_Geek]

8. Sample size of 2500+ says a lot nt

[ruggerson]

11. kick

[papau]

15. Kerry lost 3 points, Bush gained 1, to current 45/45 tie in last month -

seems reasonable.

:-)

[Feanorcurufinwe]

16. Will the media ask why Bush didn't get a bump?

[ruggerson]

17. kick

[swag]

18. kick for rationality

[Toronto Ron]

19. That poll was released on Thursday

i.e. prior to *'s "speech". This is the best poll around, in my opinion: Relatively large sample size, objective media outlet. It will get updated every Thursday; the last letter on the URL increments each week, i.e. next Thursday's results will be found at http://www.economist.com/media/pdf/YouGovJ.pdf .

[sonicx]

20. Kerry is killing in 18-24...

we REALLY need that demo to turn out this time.

[USMC0311]

21. Poll Good/Bad

Polls Good/Bad.