Skip to main content

View Diary: So, about that CNN Snap Poll (290 comments)

Comment Preferences

  •  Excuse me (2+ / 0-)
    Recommended by:
    biscobosco, wishingwell

    How hard is it to get a geographically diverse sample?  The area codes are a very big clue.

    •  It isn't hard to get a geographically . . . (1+ / 0-)
      Recommended by:

      diverse sample.

      But it may be that they conducted a poll and had a high non-response rate in other regions.  Whereas in the south, there was more interest in the debate, people not only watched the event, but were more willing to share an opinion.

      If the goal is to measure reactions in a particular region, then you focus on that population.  If the goal is to get a snapshot of the national audience, you poll the entire nation.  It may turn out that the national audience is heavily concentrated in one particular region -- it isn't concentrated across regions.  In fact, that's what the snap poll seems to suggest.

      The response rates in regions outside of the south may have been too low to provide a statistically reliable picture at the regional level.  But it may still be statistically representative of the national audience.

      If they had polled 1,200 people instead of 600 nationally, perhaps they would would have had enough data to make a judgment about views in particular regions.  

      •  that's what you're going with? (3+ / 0-)
        Recommended by:
        biscobosco, wishingwell, wu ming

        They actually called an equal number of phone numbers in all areas of the country, and only the southerners answered their phones?

        •  yeah pretty bogus (0+ / 0-)

          "We have always known that heedless self-interest was bad morals; we know now that it is bad economics" - F.D.R.

          by biscobosco on Thu Oct 04, 2012 at 12:27:46 AM PDT

          [ Parent ]

        •  Not just answer the phone (2+ / 0-)
          Recommended by:
          NotGeorgeWill, wishingwell

          But be willing to share their opinions about the presidential race to a pollster and also to be willing to be called again after the debate. I think, after reading through the posts, ORC may have inadventently introduced a bias based on southern hospitality, if you believe in such things.

          •  Yes as just today a friend told me his daughter (0+ / 0-)

            is very liberal , was against fracking , but said the Texans were her best and most polite customers and left big tips too. They were all wingnuts who would sit and bash Obama and love themselves some fracking and big oil but they were extremely polite and kind to everyone in the restaurant and treated the staff like gold.

            Follow PA Keystone Liberals on Twitter: @KeystoneLibs

            by wishingwell on Thu Oct 04, 2012 at 12:51:19 AM PDT

            [ Parent ]

        •  No, that's your misinterpretation . . . (0+ / 0-)

          of what I am saying.

          Actually I see the total number of polled respondents was actually less than 400.

          What this means is that you could have 20 people from a given minority group in a given region providing a response.  However, the response pool is too small to provide a reliable statistical measure.

          I am saying that out of a relatively small sample, the poll seems to suggest that they had enough responses in the South amongst white respondents to provide sub-category numbers that are statistically meaningful.  Even in those cases the margin of error is something like +/- 9 points.  

          •  That's a little generous (0+ / 0-)

            Judging by the margins of error, the sample was about 90% white to 10% 'not-white'. It's not that they didn't have enough people of any given minority to get data from: they didn't have enough of all minorities combined to get decent data!

            (It was also roughly 80% 50-and-over, which is also pretty insane.)

        •  Yes that would be strange as granted the older (0+ / 0-)

          crowd that would be my parents friends and siblings always answer their phone. I never get their machines.  
          We are mid 5oish and always check the caller ID first.

          Then there is my son, nephew who are in their 30s and seldom answer their phones and only have cell phones.

          But there are people like my aunt in PA and my aunt in OH would always answer their phone and watched the debate.

          It is too strange to think that even if they only surveyed all people over 50, only southerners would answer their phone. They sure missed all my relatives and most of the rest of yours too....

          Follow PA Keystone Liberals on Twitter: @KeystoneLibs

          by wishingwell on Thu Oct 04, 2012 at 12:49:48 AM PDT

          [ Parent ]

        •  but the regions aren't equal (0+ / 0-)

          There doesn't even have to be non-response bias to explain the result.

          Even in a perfectly representative sample of 430 U.S. adults, broken into four regions of which the South is substantially the largest, there's a good chance that the other three will end up with fewer than 100 respondents each.

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

          by HudsonValleyMark on Thu Oct 04, 2012 at 03:38:21 AM PDT

          [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site