Pollster.com

Articles and Analysis

 

NJ: Corzine 40, Christie 37(NYTimes 10/9-14)



New York Times
10/9-14/09; 987 adults, 3% margin of error
867 registered voters (no margin of error provided)
475 likely voters, 5% margin of error
Mode: Live telephone interviews
(Times: story, results)

New Jersey

Job Approval / Disapproval (all adults)
Pres. Obama: 62 / 25 (chart)
Gov. Corzine: 33 / 49 (chart)
Sen. Menendez: 32 / 22 (chart)
Sen. Lautenberg: 38 / 22 (chart)

Favorable / Unfavorable (registered voters)
Jon Corzine (D): 30 / 46 (chart)
Chris Christie (R): 19 / 37
Chris Daggett (i): 13 / 10

2009 Governor (likely voters) (chart)
40% Corzine, 37% Christie, 14% Daggett
Update: Including leaners - 40% Corzine, 39% Christie, 14% Daggett
46% Corzine, 43% Christie, 3% Daggett (vol)

 

Comments
havoc:

dont believe the the hype Corzine is toast.

____________________

tjampel:

I don't believe the favorables for Christie but then again, the faves for Corzine look horrific as well. No time to look at cross tabs this morning. Two candidates no one likes....and an idependent that is likable; tossup to me but the dem advantage is so large in NJ I stick with my original prediction of Corzine +2

____________________

platanoman:

Corzine will win this one

____________________

Stillow:

hehe, the most liberal partisan paper in the country shows Corzine leading....say it isn't so............

____________________

Scott:

The most Democratic-biased media pollster has Corzine up by 3 (or 1 depending on the question) basically means Christie is ahead by 3 or 4 points.

____________________

Stillow:

Putting aside the NYT clear bias, this poll is of adults....its not even registered voters, just adults. I will stick with the LV polls...or at least registered voters.

____________________

jmartin4s:

you know how there is that one kid in a class that everyone wants to STFU. That kid would be you stillow

____________________

Stillow:

Thanks for showing everyone I won the debate.

____________________

platanoman:

Clear bias? Only idiots are saying this are biased rightwing hacks like Stillow

____________________

Stillow:

LOL...thanks again for proving I am right. Once the name calling begins you know you have hit a nerve. Everyone outside the liberal circle of repetition knows the NYT is a liberal paper.

____________________

John:

The race numbers are likely voters, the favourablity numbers are RV, and the approval numbers are all adults. Seems the race is pretty much tied at the moment....

____________________

Xenobion:

I do like how bias is now associated with organization rather than method. I mean unless you say there is a bias in the method, but of course most people are not saying this.

Fox News polls are actually good in method, so good that most of the time Fox News uses Rasmussen rather than their own polling agency when reporting. Now that's bias :p

____________________

Stillow:

Bias is now institutionalized into much of our edia. There's little doubt that msnbc is an arm of the obama administration...and major media outlets like ny times or fox has bias one way or another. Our media is becoming more and more hyperpartisan every year. That will rub off in there polling. Its inevitible.

____________________

Scooby The Doo:

Christie had a 14 point lead just a few weeks ago and now he's trailing by 3. Doesn't matter who the pollster is honestly, it's about the average of all the polls that tells the story of where race is headed. The overall average is less than one point between them and Corzine is gaining steadily. Remember that a tie usually favors the incumbent no matter how unpopular he is. I predict Corzine survives.

____________________

Xenobion:

Polling is statistical science. There are three ways of seeing bias. Either in the question, the assumptions, or messing with the results. Yes you can use statistics to get what you want or show what you want, but if the questions, method, and data are there that's something you really can't lie about.

The only possible bias I see in these cross tabs are due to leading questions about Obama asked first, not switching Dagget, Corizine, and Christie in order (They made Corizine first always), and that's pretty much it. Throw in statistical error and you should have no qualms.

____________________

Stillow:

Pollsters engage in bias by selecting certain regions to call...regions that wil favor there point of view. Like calling certain area codes that have shown support or opposition to soething you like. In addition there is no gaurantee the tabs they put out do not get altered.....couple keystrokes and you can make it look like your party id is 35/35/30 when you actuallydialed and talke dto 40/30/30.

cbs for example is so partisan they fabricate news....that is why i dimiss every single cbs poll as nothing but a herd of crap the moment its released....

____________________

Cyril Washbrook:

I wholeheartedly endorse Xenobion's analysis. You should never impugn a poll purely because you think the organisation that sponsored the poll is "biased". That includes Daily Kos, Fox News and any other organisation. You have to look at the methodology within the poll. Stillow likes to bash the CBS polls, but here's the crucial thing: CBS publishes all their questions and crosstabs, meaning that we can analyse the survey's methodology and take it into account. In fact, it's not uncommon for commenters on Pollster to re-calculate the survey result with different party ID: if a poll shows something like 35-21 in the Dems' favour, I've seen people recalculate them with, say, 33-28 (or whatever).

Look at the crosstabs, Stillow: don't go for the lazy cheap shots about "oh, it's the 'liberal media', the poll is biased". Not only is it poor polling analysis, it's also poor statistics analysis, given that "biased" in the context of population sampling means something quite different to how you're using the term.

I also note that we're seeing the same routine we saw prior to the election with Scott's comment above: saying that a 3-point lead for Corzine in a NYT poll means that Christie's ahead by 3 points. Back before the election, I saw a lot of comments saying "it's a dead heat, but the Bradley effect means that McCain's ahead by 3". Then when Obama moved ahead, it suddenly became "Obama needs to lead by 5 points to counter the Bradley effect". Then when polls showed him extending his lead even further, it became "Obama needs to lead by 10 points to counter the Bradley effect". It's the same old routine, and regardless of what the "real" margin is, it's poor polling analysis.

I'm supporting Daggett by the way, and while I would probably choose Corzine if forced to pick between him and Christie on pain of death, I'm not supporting Corzine.

____________________

Rasmus Pianowski:

This kind of recalculation doesn't lead to valid results though. If the PartyID in a sample is off that much from a 'reasonable assumption', then the whole sample of individual interviews is corrupted and not representative of the electorate.

A simple recalculation doesn't help much.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR