Skip to main content

View Diary: Do states have 'house effects' when it comes to polling? (68 comments)

Comment Preferences

  •  This feels like... (1+ / 0-)
    Recommended by:
    Shockwave
    One final, and somewhat interesting, finding: The most consistent finding in the study was the fact that margins of victory, no matter who won the state, were almost always underestimated by the pollsters in 2012.
    some version of regression to the mean at play here, but I can't put my finger on it.
    •  Wouldn't it be AWAY from the mean? (1+ / 0-)
      Recommended by:
      MichaelNY

      Since what you see is a drift to the winner?

      My own unscientific take is that the potential winner's voters LOVE to go vote for a winner, and the potential loser's voters will take any excuse to not vote .... since psychologically they don't want to be seen as backing and "investing" in a loser.

    •  Not quite (1+ / 0-)
      Recommended by:
      Zack from the SFV

      regression to the mean applies when one selects subjects (people, polls, whatever) because of their extreme nature.

      In the classic example, the tallest men have sons who are shorter than they are (although still taller than average).

      So, in polling, the idea would be something like "the poll most favorable to Obama overestimates his vote" but even that isn't quite it.  Really, it would be more like: The states that gave the highest proportion to Democrats in 2012 will give them slightly less (although still more than the average) in 2016.

      •  The reason the tallest men have shorter sons (2+ / 0-)
        Recommended by:
        plf515, MichaelNY

        in my completely unscientific opinion, is that all the really tall dudes I know marry the shortest women they can find.

        And as a tall chick, it drove me up the wall.  This regression to the mean was caused by their mate selection, no doubt because they didn't want their sons to be quite as tall (although how conscious this decision was I have no idea).  Of course I'd love to see some actual data to back up my theory.

        Seriously, of all the many 6'3+ men I knew, all but one married a woman under 5'4, and many under 5'2.

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site