Skip to main content

View Diary: CNN Snap Poll - Admits Oversampling Republicans (136 comments)

Comment Preferences

  •  there were some problems (3+ / 0-)
    Recommended by:
    GOPGO2H3LL, elwior, HudsonValleyMark

    with some of the purported problems with last week's polls, but cnn is admitting it themselves. and why would they over-sample republicans?

    The cold passion for truth hunts in no pack. -Robinson Jeffers

    by Laurence Lewis on Thu Oct 11, 2012 at 09:48:50 PM PDT

    •  many here don't know what "oversample" means (1+ / 0-)
      Recommended by:

      And others are just jumping to conclusions.

      As you know, oversampling refers to a sampling design in which the probability of selection is higher for some kinds of individuals than others. Oversampling is legitimate if the data are handled appropriately. (For instance, some polls oversample people in racial and ethnic minorities -- to produce more accurate estimates for those minorities -- and then use weights to produce representative estimates for the population.) But here there is no evidence that we are talking about "oversampling" in the first place.

      To rephrase your question: Why would CNN design their sample to overrepresent Republicans? No reason, really, despite another wave of poll birther CT that says otherwise.

      Does CNN have 'too many Republicans' in its sample? Probably. We don't know the partisan composition of the debate viewership, but it's very plausible that the Obama supporters in the previous poll were harder to reach in this "snap" poll than the Romney supporters were.

      It's also very plausible (and not contradictory) that more Romney supporters than Obama supporters sat around watching the debate in the first place. It's at least possible that most viewers thought that Ryan won and that Biden clobbered Ryan among the viewers who might actually change their voting decisions (not necessarily self-identified "undecideds"). Right now we have some evidence for the latter (from CBS), very little for the former.

      Basically, this snap poll is another non-event. The one apples-to-apples comparison, just like last time around, is the before-and-after favorables, which show little net change. As for whatever CNN said about this poll on the air, geez, how many people were watching CNN after the debate? Just how hard are we gonna look for something to be annoyed about?

      Election protection: there's an app for that!
      Better Know Your Voting System with the Verifier!

      by HudsonValleyMark on Fri Oct 12, 2012 at 03:52:53 AM PDT

      [ Parent ]

    •  Thanks for remembering, Laurence (2+ / 0-)

      I'm so tickled that Markos recommended this one, which is the exact counterpart to the "unskewing" arguments of weeks ago, and doesn't even get into the problem of bad sampling by demographics that I'll bet underlies this wacky result.

      CNN's motivation, which motivated my own story, remains the same, though.  Ready to jump on the "don't let Candy Crowley moderate" bandwagon yet?  I've heard there's a good moderator who might be available, some gal named Raddatz.

      Pro-Occupy Democratic Candidate for California State Senate, District 29 & Occupy OC Civic Liaison.

      "I love this goddamn country, and we're going to take it back." -- Saul Alinsky

      by Seneca Doane on Fri Oct 12, 2012 at 05:34:29 AM PDT

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site