Skip to main content

View Diary: Minnesota Senate Undervotes III (42 comments)

Comment Preferences

  •  Senate Undervote >> Presidential Undervote (0+ / 0-)

    I don't think there is any way the deliberate undervote for Senate is less than the deliberate undervote for President so the upper limit on what 538 calls the Correctable Error Rate is ~= 10,000.
    In a radio interview said about 1/3 of presidential undervotes are deliberate so that would put the CER for senate at ~= 6,500.
    That would put the Franken gain at ~= 309.
    As lucky suggests, your analysis assumes senate and pres undervote are independent, I don't think this is likely.
    I think the people more likely to mismark ballots are also more likely to undervote the senate race.
    You could check this correlation by seeing if the senate/president undervote ration varies according to the Franken Coleman ratio.
    If there is a correlation, I think using precinct pres. undervote multiplied by the Franken percentage would give a better result.

    Did you see the canvass audit results?
    Everyone has been focused on the potential for vote gains but I was surprised at the number of votes LOST.
    Any reason to believe these are anything but random?

    Also, I think your method underestimates the Franken gain since you are using the precinct voting rates.
    As 538 suggests, this is unlikely.

    Great job, I love this kind of thing.

    •  Autocorrelation. (0+ / 0-)

      You seem to be suggesting that there should be a positive correlation between undervoting on the senate and presidential races.

      I'm not sure about that.
      If people mismark the ballot because they don't understand how to do it correctly, then definitely there should be a positive correlation for those people.

      But if people omit voting in a particular race intentionally--because they're neutral/undecided on that race--then there should be a negative correlation for them.

      Say I was gaga for Obama, but was turned off by Franken's personality or by some sexist comedy bit that he once did. I might vote for Obama and skip the senate race.

      Or say I tend to vote Democratic but I have some misgivings about Obama because of Rev. Wright, Ayers etc. I might have voted for Franken and abstained from the voting in the presidential race.

      On the other hand, if I was inclined not to vote in both contests, I'd probably simply not show up. What would be the point of standing in line to leave both races blank?

      Show me that river...

      by luckyoldsun on Tue Nov 11, 2008 at 11:26:14 AM PST

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site