Results of ARG's daily tracking poll of NH

[CMT]

American Research Group will be posting daily tracking polls on how things are going in NH. Here are the results from December 26-28:

Dean: 37%
Kerry 19%
Clark 12%
Lieberman 6%
Gephardt 4%
Edwards 3%
Kucinich 1%
Braun/Sharpton: Under 1%
Undecided 18%

http://www.americanresearchgroup.com/nhpoll/demtrack/

[Padraig18]

1. Clark may finish 2nd...

He appears to be sneaking up on Sen. Kerry...

[bearfartinthewoods]

11. i think we are seeing what the dean camp has been seeing

in their internal polls. i knew it.

[Karmadillo]

2. Dean led by 32 points in Dec 3rd ARG poll. Down to 18 points now.

Not that 18 isn't a big lead, but it will be interesting to see if the margin narrows further.

[NewYorkerfromMass]

4. I see a postive result for Kerry no matter what

his loss to Dean will not be significant.

[LuminousX]

5. Narrowing

It is the natural process when the eve of the election comes closer. The important thing to track is the rate at which the gap narrows and where it is standing the day before the vote.

[jmaier]

3. Well this will be fun to watch

It seems to re-confirm the NH conventional wisdom.

Dean wins
Kerry/Clark battle for second
Everyone else is probably not going to show

I did note that the results for Dean are less than the more commanding 40% to 45% showing that he was polling earlier in Dec and late Nov. I guess it's to be expected as we get closer to the finale.

[tsipple]

6. Anyone Know How to Read Polls Any More?

First of all, look at the dates: day after Christmas and the following weekend. You really can't pick a worse time to phone people. On the 26th lots of people are returning medium sweaters for size large, for example.

Second, there's that old friend, margin of error. The margin of error in this poll is plus/minus 4 points. So Dean may have 37%. Or he may have 41%, or 33%. Might even be lower or higher -- we're just 95% confident he's 33 to 41%. Or maybe not, because we'd be 95% confident if the figure was centered on 50%.

Anyway, you catch the drift. These things have imprecision, and you can only draw conclusions within those limits. So let's not hyperventilate here.

Those dates are awful, though. Don't put too much stock in this one. (Even if it does show my guy doing well, I'm deeply skeptical of a 12/26-12/28 poll.)

[ignatiusr]

9. 4% MOE is not that bad

Regardless of how you look at it, it's a drop for Dean. Hard to say whether it's a slight or a major drop. It could be anywhere from a 4 pt. to a 12 pt. drop. Clark's the only person showing significant movement, other than Dean, who is showing negative movement.

Tsipple, I admit I don't know much about the intricacies of polling, but how does the fact that it was from the 26th to the 28th mean anything? I mean, it's not as if the stress of returning Christmas presents would force voters to choose someone other than their candidate when asked. And 600 people is 600 people. If they weren't able to reach someone, they reached someone else. That's still in line with the spirit of random sampling, isn't it? Or am I missing something?

[tsipple]

22. Well, to pick something obvious...

...Jewish voters may not be out returning Christmas presents on December 26th. :7

But basically the problem with holiday polling is that you get some wacky, unpredictable, and potential sample bias. It's because you have strange things going on with who's at home and who isn't.

To pick another example, young people are probably out visiting with older relatives. So you're less likely to get younger voters that weekend, I would think. (They wouldn't have reached me for that reason.)

It's a lousy time to run a poll, IMHO (and in the opinion of many pollsters). You want a nice, quiet, non-holiday weekday (or weekdays). So I'm anticipating those numbers for polling taken, say, Tuesday, January 6th onward.

[Karmadillo]

7. Kick

n/t

[rucky]

8. Are Kerry & Clark are splitting the votes vs. Dean?

No telling just from the numbers, but the gap does beg the question.

Then there's the 18% undecided that will only close the gap if all of them switch to the same candidate.

[adadem]

17. You're probably correct

I think a good showing by Kerry in IA may boost his numbers in NH. Would be quite happy with Clark or Kerry. Am surprised Lieberman is doing as well as he is. Would think Kucinich, my emotional fav, would be doing better.

[clarknyc]

18. No.

Clark and undecided are splitting the Dean slip. Kerry is treading water, along with Gephardt and Lieberman.

[adadem]

19. How can the candidate

who is clearly polling second in both IA and NH be "treading water"?
Nonsense.

[flpoljunkie]

10. Thanks for posting this link, CMT.

Dean appears to have peaked earlier this month. Good news for Kerry, I would think.

[bearfartinthewoods]

13. good news for the party AND the country....

[CMT]

20. the interpetation is in the eye of the beholder

Dean still has a 2-1 edge which if it continues will mean a very solid victory. I expect it will narrow in the last weeks but anything over ten points for whoever wins the primary is a very solid victory.

Note too that Iowa will help influence what happens in NH to a certain extent.

[bearfartinthewoods]

21. 10 points is a good margin unless you were supposed to win by 20..

as long as the primaries remain valid and competitive, i'm happy.

[meow mix]

12. here was a comparison

http://rob.dailykos.com/story/2003/12/29/182759/57

Dean 37 (45)
Kerry 19 (20)
Undecided 18 (15)
Clark 12 (8)
Lieberman 6 (6)
Gephardt 4 (3)
Edwards 3 (2)
Kucinich 1 (1)
Braun 0 (0)
Sharpton 0 (0)

dean lost substantial points - hmmmm

[clarknyc]

14. Go read the analysis at DailyKos

Comparing a daily tracking poll with a regular poll is like apples to oranges: both are round fruit, but not the same thing.

However, if these numbers stand up over a few days, then it would be fair to say that Dean has slipped a bit, Clark and undecided have gained a bit, and the Washington Democrats are merely treading water.

[meow mix]

15. i did =)

but just thought id mention it - cause it was there, i guess =)

[clarknyc]

16. I meant

"you" in the generic sense, not you specifically. I wasn't clear as I could have been. Of course you saw the analysis --- your post links to that site! Thanks for the link, btw...