Poll: Dean widening lead over Kerry in NH

[stickdog]

Class, can anyone tell us what is wrong with the wording in this AP article?

http://story.news.yahoo.com/news?tmpl=story&cid=694&ncid=694&e=5&u=/ap/20030918/ap_on_el_pr/new_hampshire_poll

N.H. Poll: Dean Has 10-Point Lead on Kerry
Thu Sep 18

WASHINGTON - Howard Dean (news - web sites) holds a 10-point lead over John Kerry (news - web sites) among likely voters in the New Hampshire primary, according to a poll that suggests the race is tightening between the two New Englanders.

Dean, the former Vermont governor, had 31 percent in the poll by the American Research Group of Manchester, N.H., while Kerry, the Massachusetts senator, had 21 percent. The remaining candidates were in single digits; 27 percent were undecided.

In the last ARG poll, in mid-August, Dean was 7 points ahead of Kerry, 28 percent to 21 percent.

Rep. Dick Gephardt (news - web sites) of Missouri was at 8 percent, and Sen. Joe Lieberman (news - web sites) of Connecticut had 5 percent. Florida Sen. Bob Graham (news, bio, voting record), North Carolina Sen. John Edwards (news, bio, voting record) and retired Gen. Wesley Clark, who entered the race Wednesday, had 2 percent, while Ohio Rep. Dennis Kucinich (news, bio, voting record) and Carol Moseley Braun had 1 percent. Al Sharpton had 0 percent.

[blm]

1. They say in a different poll

Dean had the wider lead, and then go on to show the earlier ARG poll. It does say different. It would be awkward to someone reading it quickly.

[stickdog]

3. And blm, does comparing different polls make any sense at this point?

Especially when you have a poll by the same company that shows the lead widening just from just one month ago?

Or does it make a hell of a lot more sense simply to examine the trend lines of each poll?

As in:

http://americanresearchgroup.com/nhpoll/dem/

June
------------------------------
Kerry 28%
Dean 18%

July - Dean's margin +4%
------------------------------
Kerry 25%
Dean 19%

August - Dean's margin +13%
------------------------------
Kerry 21%
Dean 28%

September - Dean's margin +3%
------------------------------
Kerry 21%
Dean 31%

Over the last 90 days = Dean +13%, Kerry -7% = Dean's margin +20%.

Dean's favorable/unfavorable = 66%/9% = 7.3 favorable to every unfavorable

Kerry's favorable/unfavorable = 64%/13% = 4.9 favorable to every unfavorable

Clark's favorable/unfavorable = 22%/5% = 4.4 favorable to every unfavorable

And just for all of the highly vocal Kucinich backers out there

Kucinich's favorable/unfavorable = 12%/18% = 0.7 favorable to every unfavorable

[Feanorcurufinwe]

2. The article is worded just fine.

It states the result of the poll, and compares it to the same poll taken in August, as well as other polls taken more recently.

Everybody can play spinning top all they want, the only poll that matters in NH is going to be taken January 27th.

[stickdog]

4. Keep that head well buried. (NT)

[Feanorcurufinwe]

5. So you are saying the poll taken on January 27 isn't the one that counts?

[stickdog]

8. Yeah, that's what I'm saying.

Where's that scarecrow icon?

[Fabio]

6. Agreed.

[quinnox]

7. The wording is fine, the problem is a key sentence is missing

from the post. "Dean's lead over Kerry is about half what it was in a different New Hampshire poll late last month but close to the 12-point difference in another poll a week and a half ago."

[stickdog]

9. No, the problem is that

the same company (ARG) does this same exact poll in the same manner every month and it has consistently shown better results for Kerry vis a vis Dean in comparison to Zogby polls.

Hence, this ARG poll is far more accurately compared to its own August poll than to a late August Zogby poll if the purpose if to show an actual -- as opposed to chimerical -- trend.

The idea that Kerry is gaining on Dean is obviously completely spurious. My God, Dean increased his lead by 3% over the exact same poll taken in August!

How obvious do I have to make this for y'all???