USA TODAY/ Gallup Poll Still Shows Kerry up among Registered Voters

[Zynx]

I saw that on the bottom ticker with Kerry/Edwards up 50 to 47, but it was reversed with their "likely voters".

[NMDemDist2]

1. remember folks that most folks under 30 don't have home phones

to get polled anyway

[Frodo]

4. Doesn't matter much... they don't vote either.

I suspect the weakness in the poll (one among others anyway) is in how they determine "likely" voters. If this year is different (and it sure SEEMS like it is), the prior assumptions of WHO is "likely" to vote may be way off.

I don't think the polls showing big bounces (but polling predominantly Democrats) are any more realistic.

Let's just watch the totality of polling over the next week or two.

[Zynx]

5. The "likely voter" sample once showed Kerry up by 12%.

That was bullshit. Kerry wasn't up by 12 in February.

[Frodo]

7. I suspect that polls overplay each movement

and are designed to.

A poll swinging from a 5pt Kerry lead to a 10pt Kerry lead probably (JMHO) means a move from a 2% win to a 3% win.

I suspect that the "true" swing in voters (were the election held today) would be little more than a 2-3% victory by either side.

[fearnobush]

2. That poll is also total BUSHIT. They never covered the complete

polling peroid and they did not mix the results like all other polls do. I'm sure this one has a sampling skew of 60 reg repukes to 40 dem. Wait for the other polls to come out. This is pure CNN GOP ass sucking. I suggest it is time to protest the CNN offices directly. We cannot have 2 FOX news networks out there.

[Zynx]

3. I am going to bet serious money it was a sample of 35% Republicans

34% Dems and 31% independents. It should be 37% Dems, 31% Reps, and 32% Independents.

[Frodo]

6. Where's that from?

What data puts the electorate at 37/31/32?

[Zynx]

8. The 2000 exit poll.

It's something like that.

EDIT: Actually it's 39/35/27 so I was off. I think what skewed those numbers is that independent turn out was lower.

[Frodo]

9. Tough to use one year's exit polling

What, for instance were 2002's numbers?

Why was there a difference? And what basis shall we "call" what the mix will be this time around?



I prefer to look at the internal numbers (but hardly anyone lets you see them). How did Kerry do among Democrats (was there a surge in how strongly people felt about going out to vote)? How did Bush do among Republicans? Who leads among Independants?

THEN let US decide who we think will show up. That's how campaigns do it ("we need at least 3% better turnout from our base than last year" etc).

[Zynx]

10. Kerry always, in every poll, has won among independents.

[tom22]

11. I see that Gallup has

some doubts about the results themselves. They are continuing polling tonight. I believe they release their results on Tuesdays. We shall see.

[Zynx]

12. Yeah. Two days of polling is not responsible.

[Frodo]

13. NONE of the polling has been "responsible"

Everyone wants to get the number out NOW. By next week, the "bounce" will have already been settled on and reported.

But there hasn't been a single decent poll out yet.


BTW - it isn't true that Kerry has consistently led among independents. He's doing better than a challenger has a right to expect, but it's hardly been a constant.

[Zynx]

14. When's the last poll that he hasn't? It's been a while.

[sandnsea]

15. Show me

Every single poll I've seen has Kerry ahead among Independents, for a very long time now. And young people, women, minorities and seniors. White men are the fucking problem, particularly ages 35 - 49. And I wish some of you white men around here could help the rest of us understand why. What do your white buddies SEE in the arrogant little bastard? Truly, I am totally baffled.

[tandt5044]

20. Can you provide

a link with those numbers ??

[DemocratSinceBirth]

21. Party ID Fluctuates.... That's The Challenge

However that doesn't explain the fact that on 11-27-2000 Gall(o)p had the race 52% -39% in favor of Bush while Zogby had it 45% -43% in favor of Gore.....

[Feanorcurufinwe]

16. 'Likely voters' means they decide which poll respondants to count.

All they have to do is adjust their method for determining who is a 'likely voter' to come up with any result they want.

[Zynx]

17. Indeed. Gallup's numbers are always the most volatile.

The likely voter sample is not reliable.

[fearnobush]

18. Yes, that is why... way back in Jan. They spacifically used their

Registered voter samples to show Bush some 20 pts ahead of dean, yet their Likely Voter sample had Dean behind by only 5 pts. Remember those days?

[Zynx]

22. Gallup showed Bush up by margins that were unrealistic at that time.

Then somehow his numbers collapsed twenty points? I don't believe that either. I believe they fell, but I don't believe they fell by as much as Gallup said.

[DemocratSinceBirth]

23. There Can Not Be A Fifteen Percent Difference

between likely and registered voters unless you're using a model that assumes an election day turnout in the low forties or high thirties....


Not gonna happen...

[IndianaGreen]

19. I think the race is already locked in with few undecideds

Getting the vote out is going to be very critical this year, and that means getting people that haven't voted in the past to vote this year.

[Zynx]

24. Probably true, but Bush's fundementals are weak.

Mid-40s approval rating, 55-40 wrong direction numbers or worse, general malaise about the country all add up to trouble in the final couple weeks of this thing.

[apnu]

25. I just saw the front page from USA TODAY and...

... it shows Kerry up 50%, Bush 46%, Nader 2%. Whereas the pre-convention poll shows at best a dead heat with Kerry having a very slight margin ahead of the Chimp.

Kerry has a lead that is at the upper margin of the sampling "fudge factor"

How is this not a bounce? Does anybody believe that Kerry was gonna bounce 20+ points in a nation that is on the brink of political civil war? (I mean that in a cultural sense, not in violence)

[shivaji]

26. Has anyone polled the computer programmers from Diebold?

That is where the election might be decided. When I lived
in Chicago, it was not who voted, it was who counted the
votes.

[Inland]

27. Can't explain it, but Zogby and Newsweek have Kerry up BIG

See the results at www.pollingreport.com. Z has Kerry up by eight among likely and NW has Kerry up by seven among registered.