Skip to main content

View Diary: CNN unskews its own snap poll (147 comments)

Comment Preferences

  •  They say that they are polling people with whom (4+ / 0-)
    Recommended by:
    along, DaveP, Brooke In Seattle, Smoh

    they have spoken before.  In the earlier conversation, they asked if the respondent planned to watch the debate and whether or not it would be okay to call them afterward.

    The good we secure for ourselves is precarious and uncertain, is floating in mid-air, until it is secured for all of us and incorporated into our common life. Jane Addams

    by Alice Olson on Fri Oct 12, 2012 at 10:56:52 AM PDT

    [ Parent ]

    •  Thank you (0+ / 0-)

      for clearing that up!  I've often wondered if the whole east coast is excluded from these.

    •  right, so (0+ / 0-)

      I don't understand how this could be considered representative of all debate watchers.

      At best, it's representative of all registered voters who said, several days ago, that yes, they were going to watch the debate Thursday night.

      But things change. Some folks end up not watching, a lot more probably decide to go ahead and watch. This sample does not represent them.

      In fact I don't think this is a "poll" at all. It's a self-selected panel of debate watchers. Only some of whom agreed to be called back and requestioned.

      This methodology seems fatally flawed to me. I'd love to have a statistician describe why I'm wrong.

      •  it's pretty flawed, yes (1+ / 0-)
        Recommended by:
        along
        In fact I don't think this is a "poll" at all. It's a self-selected panel of debate watchers. Only some of whom agreed to be called back and requestioned.
        It's not what would normally be called "self-selected," since the initial sample was selected at random. But with three chances to opt out (before the first interview, at the end, and when called for the second interview), it could be pretty darn messy.

        CNN should have consistently made clear that this was a panel study -- and one, I'm guessing, with such a high attrition rate that you might consider "panel study" a euphemism.

        I don't think it's a useless study. I think it's mildly interesting to reinterview people and see if the debate really made a difference to them. But portraying it as a sample of "registered voters who watched the debate," full stop, is wishful. And their SPECIAL NOTES OF CAUTION don't really help.

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

        by HudsonValleyMark on Fri Oct 12, 2012 at 12:45:29 PM PDT

        [ Parent ]

        •  thanks for your thoughts (0+ / 0-)

          the second problem then becomes obvious: if it's a panel study, then it's crucially important to know the baseline support/lean position of each respondent. Ideally, a panel like this should be evenly divided. But in this case, all we know is their fav/unfav opinions, and those were quite heavily skewed to Ryan. Another bit of info not provided with the initial CNN report, and not presented as a caveat in ANY mention of the poll results I've read so far. I can't see how this exercise has any value at all.

          •  depends on how you look at it (1+ / 0-)
            Recommended by:
            along

            I think it's mildly interesting that in both snap polls, the favorables hardly moved. So, I'm (mildly) glad they did the polls. And at least this time CNN made clear, or clear-ish, that the sample was probably off -- so if one found the more complete report and stared at it for a while, one wouldn't be too surprised by the CBS results. That said, the reporting was pretty useless, and the written report isn't a whole lot better.

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

            by HudsonValleyMark on Fri Oct 12, 2012 at 06:44:26 PM PDT

            [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site