February 29, 2008
Leve on Pollster Report Cards, Part III
I am playing catch up after a day spent mostly in airports and offline. So I pass along this item after only a brief review: SurveyUSA's Jay Leve has posted yet more on their new and evolving pollster report cards. It follows up on our previous exchange and my column posted earlier today.
The gist of this post is that, in at least one example -- polls conducted prior to this year's Iowa caucuses -- the "most accurate" poll was not necessarily the last one. Anyway, our always attentive readers will be able to give Leve's post more attention than me today, so please have at it.
Meanwhile, you may want to click through to Leve's post if for no other reason than he kindly posted a photo that, in all the craziness of the last few months, I never managed to post here: It shows Charles Franklin and me receiving the AAPOR Warren Mitofsky Innovators Award last May (that's Franklin on the left, me on the right and Mia Mather, Warren Mitofsky's widow, in the center).
By Mark Blumenthal on February 29, 2008 12:01 PM | Permalink
I rather think that Leve missed the point of Jon Krosnick's comment. Krosnick basically was saying that it's legitimate to evaluate final polls on whether they are close to the election returns; I didn't think he was hinting that in general the latest "final polls" are the most accurate.
With respect to the Iowa Republican primary: polls in September 2007, not so predictive. Doh. (Huckabee was generally in fourth place then.) It appears from the Pollster.com charts that Huckabee's surge had slowed by mid-December, so it may well be that after that point, timing didn't matter much.
I don't mean to imply that usually timing would matter a lot. It's an empirical question, and I think it's great that Leve is exploring it. I don't think we can learn very much from a single election, but if anyone out there thought that the latest poll is always the rightest, one counterexample suffices.
Posted on February 29, 2008 9:46 PM