Skip to main content

View Diary: Senate Snapshot, October 18th: GOP takeover becoming extremely unlikely (69 comments)

Comment Preferences

  •  Nate's got a point (8+ / 0-)

    I believe that the reason Nate includes the possibility of a "systematic" swing is that you can observe it in past data.  You can ask whether a full set of actual Senate results (not just win / lose, but the the full list of margins) is consistent with purely independent polling error.  If I recall correctly, the proper statistical test easily rejects independence.  And, I believe that Nate builds in a level of systematic variance that is consistent with past results.  

    The pure sampling error across polls is surely statistically independent.  But, "likely voter" models can be easily systematically off in one direction and (more important for primaries than generals) last-minute swings in the direction of undecided voters can go in one direction or the other.    

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site