This is an archive of the old Stones Cry Out site. For the current site, click here.

« Video footage of the Ivory-Billed Woodpecker | Main | Deliver Us From WalMart’s Critics; Give Us WalMart Alternatives »

May 01, 2005

Febble's Fancy Function or the Liddle Model that Could

Last week, Terry Neal of drew attention to a study of the 2004 exit poll data posted on the web on March 31 by the organization US Count Votes. The USCV study authors, highly credentialed professors from distinguished universities, declare that the explanation put forth by the exit pollsters, Edison-Mitofsky, is not supported by the data. They suggest instead that the data are more consistent with a "fraud in Bush strongholds" hypothesis.

Yesterday, Mystery Pollster and DailyKOS carried summaries of a working paper critical of the USCV study prepared by Elizabeth Liddle, a former USCV contributor. Elizabeth draws the following conclusion from her analysis:

...the conclusion drawn in the USCV report, that the pattern observed requires “implausible” patterns of non-response and thus leaves the “Bush strongholds have more vote-count corruption” hypothesis as “more consistent with the data”, would seem to be unjustified. The pattern instead is consistent with the E-M hypothesis of “reluctant Bush responders”, provided we postulate a large degree of variance in the degree and direction of bias across precinct types.
As Mark said, the story is a bit "complex and geeky," but worth hearing. I won't get into all the details in this post because I think you should read Elizabeth's paper first and then Mark Blumenthal's excellent summary, followed by DemFromCT's post at DailyKOS, which will fill you in. The DailyKOS post has over 300 comments and it appears to be very well received.

Big hat tip to SCO reader LotharBot who generated the random data lists I requested within a few hours of putting out the call. Your contribution helped solve a piece of this puzzle. Thanks again!

Aw heck... Since everyone else is doing it, why not:
[Full disclosure and adapted from the Mystery Pollster and DailyKOS, who have similar disclaimers: For the last two weeks, I have had the unique opportunity to watch the development of Elizabeth Liddle's work through a running email conversation between herself, Mark Blumenthal from Mystery Pollster and DemFromCT from DailyKOS. This post benefits greatly from their input, although as always, the views expressed here are my own. It was one heck of a ride! ~Rick Brady].

P.S.: Buy Elizabeth's Children's Book, Pip and the Edge of Heaven and share it with your kids.

Here's a review:

This book is the journey of a child's thoughts and discovery of love and connection, heaven and G-d. My kids are 8 and 9 (boy and girl) and they still both sit so quietly for it. They smile at the sweetness of the thoughts that Pip shares with his mother. They like the way the mother talks with Pip. It sets a nice tone for us for going to bed. I love this little book. It may make it to the books I save to read to grandkids someday.

Posted by Rick at May 1, 2005 02:36 AM

Trackback Pings


Thanks for all your work on this, Rick. You were a big part of the output, which I'm optimistic about regarding acceptance. But whether the conclusions hold up or not, the process was awesome. And as much (or as little in my case) as we all contributed, the kudos really go to Elizabeth.

Posted by: DemFromCT at May 2, 2005 07:36 PM

You are a sucker like the rest of the Febble followers. This woman has no reason to be interested in US elections. She states her motivation in focusing on election polls stemmed from the fact that the election poll analyses were distracting from the story of election fraud. HELLO, there has been no press publicity on election fraud period. How can one issue, exit polling clues to fraud, distract from another, fraudulent acts, when neither received any publicity.

She's a ringer and people who jump on her silly bandwagon are fools!

Posted by: Alan at May 7, 2005 11:12 PM

Right... Let's wait one more week and talk again, okay Alan? Thanks for stopping by.

Posted by: Rick Brady at May 8, 2005 02:20 PM

A ringer for what, Alan? Spell it out.

The concept that you have to discern the motivation in order to understand and follow the math is what's silly. OTOH, what Febble has published is available for analysis. So analyze it. It will stand and fall based on its merits.

Posted by: DemFromCT at May 8, 2005 03:55 PM