« A Personal Note | Main | Why & How Pollsters Weight, Part II »

October 06, 2004

Veep Debate

A quick take on the post debate polling. Andrew Sullivan (who kindly introduced many of you to this blog – thank you Andrew) gets it basically right in his quick polling summary this morning. Let me add a few thoughts.

My first impression, based largely on the ABC survey results, is that among the debate was essentially a "draw,” in that it reinforced existing opinions of debate viewers. The CBS survey suggests a better performance for Edwards among uncertain voters, but the release leaves out a few details that might give us greater confidence in the result. If Gallup surveyed poll watchers last night, I can’t find it online.

As the ABC release notes prominently, they showed an overall "win” for Cheney largely because Republicans were more likely than Democrats to tune in to the debate (a finding that indirectly supports the notion that Cheney’s continuing presence on the ticket helps motivate the GOP base). Slightly more of those initially supporting the Bush/Cheney ticket (80%) thought their man won last night than did those supporting Kerry/Edwards (69%). However, that difference was smaller than the gap between Kerry (89%) and Bush (709%) among their respective supporters last week and the vote preferences of the debate viewers hardly budged. Note to The Note: As of this posting, the full ABC release is not yet available here (though it should be soon) but is referenced here.

The CBS study again used a projectable Internet methodology (a real random sample as explained here) to interview "uncertain voters” who found Edwards more persuasive than Cheney. However, the CBS release does not report the pre-debate or partisan leanings of their sample except to say that before the debate, "most of these uncommitted voters were undecided in their overall evaluations of either Edwards or Cheney” (emphasis added). I’m not sure what "most” means. As such, it is hard to rule out the possibility that Edward's better numbers resulted from having more Democrats or Kerry supporters in the sample. Last week, the CBS survey of uncommitted debate watchers favored Kerry before that debate by a 31% to 19% margin, with half the sample completely undecided. Although the sample this time was slightly smaller (n=179) than the one last week (n=209) but they do not say if they changed their selection criteria for this debate.

UPDATE: Democracy Corps, the Democratic polling entity of Stan Greenberg and James Carville, just posted results from their own debate reaction survey using the same projectable Internet methodology as CBS. Based on a quick skim of the cover page, but the findings seem consistent with those above: a perceived draw among all viewers, perhaps a slight Edwards advantage among very small subsamples of persuadable voters.

Related Entries - Debates

Posted by Mark Blumenthal on October 6, 2004 at 08:12 AM in Debates | Permalink

Comments

Interesting post, thanks for the insight -- one question; you say that the Cheney/Edwards difference in supporters who think their candidate won (80%/69%) is less than the difference for Kerry/Bush (89%/79%) -- how is this true? I would have thought you just took a straight difference in which case 80 - 69 > 89 - 79; clearly the calculation is more complicated. What is it?

Posted by: Jeremy Osner | Oct 6, 2004 9:39:49 AM

My biggest concern is how many respondants did Konwledge Networks need to contact to arrive at 179?

There previous work included 1300 or so which was 20%. That computes to 6500 respondants - people who actually7 answered enough questions to be defined as an "persuadable".

It seems implausable.

Posted by: Eric | Oct 6, 2004 9:44:54 AM

Jeremy --

Good catch! Not complicated, just a typo. The 79% for Bush should have been 70%. That's what I get for blogging from my daughter's hospital room on four hours sleep. Thanks for pointing it out. I've corrected it above.

Posted by: Mark Blumenthal | Oct 6, 2004 12:48:17 PM

A key point is: did the VP debate move independent and unsure voters' perceptions regarding the main thrusts of the two camps' strategies? Were they influenced to associate the Reps with dishonesty/lies/ lack of credibility (which seems to be the Dems theme)? And Dems with flip flops and unsafe leaders (which seems to be Reps theme)? Do any of the polls shed light on these points?

Posted by: Glenn | Oct 6, 2004 1:31:22 PM

"As the ABC release notes prominently, they showed an overall “win” for Cheney largely because Republicans were more likely than Democrats to tune in to the debate (a finding that indirectly supports the notion that Cheney’s continuing presence on the ticket helps motivate the GOP base)."

Could this be partly the result of generally Democratic New York and Minnesota viewers tuning in to the ballgame instead?

Posted by: Josh | Oct 6, 2004 3:10:20 PM

The comments to this entry are closed.