Poll Watch: New Cook Report Nat’l Primary Poll

Trackback URL for this post:
http://race42012.com/2007/04/30/poll-watch-new-cook-report-natl-primary-poll/trackback/

13 Responses to “Poll Watch: New Cook Report Nat’l Primary Poll”

1. Sean Says:
   April 30th, 2007 at 5:17 pm

   “In less than one month, Giuliani’s lead over McCain dropped from 17 points to just 7. Mitt nearly doubled his support, but is still far back from Rudy and J-Mac, and none of the second- or third-tier candidates are making any noise whatsoever.”

   AKA movement well within the margin of error . . .

2. David B Says:
   April 30th, 2007 at 5:24 pm

   Hopefully an outlier regarding Rudy.

   I have been on record saying Rudy would poll around 25% if/after Fred announces.

3. Matt Says:
   April 30th, 2007 at 5:40 pm

   Sean,

   Not only isn’t a 6 point drop, in poll with around 450 people (1000 for both parties, likely 450 for the GOP) “well within the margin of error”, it isn’t within the margin of error at all. Based on their data from previous polls, I’d guess this poll has a margin of error of about 4.5-4.9%. That puts both Rudy’s and, to a lesser extent, Romney’s, movement outside the margin of error, and McCain’s just outside of it. And considering every poll I’ve seen has shown almost identical upward movement for Romney since March, your statement strains credibility even outside the common, scientific definitions of “margin of error”.

4. Matt Says:
   April 30th, 2007 at 5:42 pm

   I meant that McCain’s was “just inside” the margin of error of course.

5. Sean Says:
   April 30th, 2007 at 5:50 pm

   Matt,

   If I understand your argument correctly, then you are making a common mistake re error margins. The error margins apply to both polls conducted. Therefore, we can say we are 95% confident that Giuliani’s result in the first poll is between 29.5% and 38.5% (using +/- 4.5% as my baseline, just for simplicity’s sake, though, given the results in the high 20s/low 30s, its probably more like +/- 4.2%). We can be 95% confident that Giuliani’s result in the second poll is between 32.5% and 23.5%. Because the two ranges overlap, we can’t be confident that there has been movement.

6. Matt Says:
   April 30th, 2007 at 5:59 pm

   Fair enough Sean. It’s been awhile since my statistics class. I still your distinction is an immaterial one. Using that criteria of margin of error, virtually every movement seen in polls, from month to month, has “no statistically significant movement”. Given that we’re in the prediction business here, that’s not terribly helpful. Far more helpful, and I confess I don’t know how to do this, is to determine the likelihood that any given change represents a statistically significant change. Of course, I suspect that’s not nearly as simple as margin of error calculations.

7. Sean Says:
   April 30th, 2007 at 6:09 pm

   Probably wouldn’t have said anything if you hadn’t used the “strains credibility” bait. Like I said, common mistake.

   And you’re right — most polls do show movement within the error margin. That’s why the better measure is what you alluded to in the second half of your post. Looking at all polls together, there just hasn’t been much movement against Giuliani recently — the most recent movement came right before the first poll above was conducted, when people started talking about FDT entering. Since then, the race has been pretty darned stable. (Though Romney seems to have improved, and for reasons that I don’t fully understand, results that skew toward 100 or 0 have smaller error margins, so the movement is probably “real”).

8. Adam zuckerman Says:
   April 30th, 2007 at 6:24 pm

   I predict Fred Thompson will soon drop in the polls further soon after he announces….he has no momentum even now….everyone loves the second-string quarterback, until he gets into the game…

9. Tommy Says:
   April 30th, 2007 at 6:38 pm

   Adam, he has a very large support in regions of the country. He was endorsed yesterday by Reagan indsiders, he was endorsed by the Texas Legislature over the weekend, Mark Levin has all but actually endorsed him. He’s speaking at the LCRC on May 4. Where do you get the idea he has no momentum. After he announces, I can assure you that MORE BIG NAMES are going to endorse him rather quickly. You can say that he won’t win, even I say Rudy’s the favorite right now, but to say he’s just going to drop right back down is rather absurd, considering that the “Flavor of the month” tag expired a month ago.

10. JayPe Says:
    April 30th, 2007 at 7:32 pm

    Giuliani has kept dropping after being Flavour of the Month, so I see no reason why Thompson wouldn’t as well…

    Romney needs to be hitting double digits every month if he’s to build momentum and keep the cash rolling in. McCain will be relived he’s back over 20%!

11. Argo Says:
    April 30th, 2007 at 7:35 pm

    Tommy,

    According to those who put their money where their mouth is on this stuff, over at Intrade, Fred Thompson peaked three weeks ago.

    Here’s a chart of FDT stock:

    http://data.tradesports.com/graphing/jsp/closingPricesForm.jsp?contractId=462622&tradeURL=https://www.intrade.com

    For comparison’s sake, here are the current prices of all prospective GOP nominees:

    Giuliani 31.7
    McCain 20.0
    Romney 15.4
    F. Thompson 14.8
    Gingrich 3.3
    Hagel 3.0
    Huckabee 2.4
    T. Thompson 0.9
    Brownback 0.6
    Tancredo 0.3
    Paul 0.2
    Hunter 0.1

12. Grant Gormley Says:
    April 30th, 2007 at 9:11 pm

    Let’s lose the general election. Look at the polls of general election matchups.

13. econ grad stud Says:
    April 30th, 2007 at 10:37 pm

    Sean you’re wrong the approipriate alpha level isn’t 0.75 (which is what you get when you apply 0.05 to both individually polls) the appropriate level is 0.05 for BOTH collectively polls.

    The odds of Giuliani having no drop in support is 0.0517.