Dean leads Kerry 37%-16% in latest New Hampshire tracking poll (12/29-31)

[goobergunch]

NH Democratic Tracking  
Dec 29-31

Braun      0% 
Clark     13% 
Dean      37% 
Edwards    4% 
Gephardt   6% 
Kerry     16% 
Kucinich   2% 
Lieberman  6% 
Sharpton   0% 
Undecided 16%

http://americanresearchgroup.com/nhpoll/nhpoll.shtml

[Scott Lee]

1. John Kerry, we hardly knew ye

Pack it up. Say buh-bye, and see if the bank will let you have your house back.

[CMT]

2. solid Dean lead, Kerry may be overtaken by Clark

if the trend continues. The first day of the tracking poll, Kerry had 19% and over the last four days it has declined to 16%. Dean has stayed steady at 37% and Clark has gone from 12 to 13%. Gep over this time has also gained a bit of ground moving from 4% to 6%.

[mouse7]

3. This result in would seriously shorten the primaries

Dean in first with Kerry in second, but with Dean doubling Kerry's support, but with Clark still beaten by Kerry.

Dean's support slipped a bit last week but stabilized this week.

[Raya]

4. Clark up 5% ? Look like Big-Media Senario Come to Pass: Dean or Clark, No

one else need apply. No Kucinich, No Kerry, No Lieberman, No Gephart.

I think we are all being manipulated. My Dean friends most of all.

[chaumont58]

7. "We're being manipulated" What does your statement mean?

Whose manipulating whom? The poll takers? The poll respondents?

[Hep]

5. Polls don't mean anything unless youre talking about Dean dropping

Didn't you get the memo?

[HFishbine]

6. +/- 4% Margin of Error

There hasn't been any statiscally meaningful movement in the tracking poll yet.

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

[Donny247]

8. Clarification

I keep hearing people say that the margin of error makes movement in polls like these insignificant. This isn't the case. The MOE of +-4% simply means that you can say with 95% confidence that the results are within this range. This DOES NOT mean that it is just as likely to be +-4% as it is to be spot on. For instance, if you lower the confidence level to, say 75%, you could predict within 1 or 2 points. Statistical results are shaped like a bell curve, and to have a MOE with 95% confidence, you have to go pretty far out on the curve. The results are still VERY likely (60-70%) to be within a point of the survey findings.