Skip to main content

View Diary: Rigged Elections for Romney (Updated With Ohio 2008) (100 comments)

Comment Preferences

  •  Thank you again very much (0+ / 0-)

    You are a good egg.

    There's two takeaways from this plot IMO.

    1) There's something funky about the data. The Y-axix has solid blue, but the data cloud slopes away from the X-axis. It is hard to imagine so many 100% Democratic precincts and no 100% GOP examples. Here we need to look at specific matchups of the data to specific precincts to find out what is going on with these data.

    2) If there wee no bias we'd expect to see a square distribution. That is roughly the distribution we see here,after one accounts for the Y-axis having roughly twice the resolution per inch of the X-axis. That is, what would be a square appears on this graph as a horizontal rectangle aka a 'stretched' square. There is some deviation from the ideal but not much. So whatever share/size effect there is, if any, does not seem strong.

    Personally, I would want to clear up the discrepency between the blue Y-axis and the clean X-axis before I used thee data to make any conclusions whatsoever. One cannot trust the results unless one trusts the raw data.

    Thank you for your responsiveness and willingness to incorporate alternative explanations.


Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site