You are viewing an obsolete version of the DU website which is no longer supported by the Administrators. Visit The New DU.
Democratic Underground Latest Greatest Lobby Journals Search Options Help Login
Google

Reply #10: Not a proof, but raises a legitimate point [View All]

Printer-friendly format Printer-friendly format
Printer-friendly format Email this thread to a friend
Printer-friendly format Bookmark this thread
This topic is archived.
Home » Discuss » Topic Forums » Election Reform Donate to DU
andym Donating Member (1000+ posts) Send PM | Profile | Ignore Thu Mar-10-05 05:08 PM
Response to Original message
10. Not a proof, but raises a legitimate point
Edited on Thu Mar-10-05 05:28 PM by andym
This is not a proof of exit polls accuracy.

You correctly define I as intended vote. Your initial equation is reasonable.
I = R + S + F

although I would add M for mistaken votes, where a voter might have unintentionally voted for the wrong candidate as happened in FL in 2000.
I (candidate) = R + S + F + M

But Exit polling data are not equivalent to intended votes, that is, I is not equal to exit polling data. However, what is true is that exit polling data attempts to measure I. Therefore in theory, exit polling data may be a more accurate estimation of the intention of the voters than the actual results . I think this is a valid point.

But this does not cover potential problems with exit-polls. That is the statistical model underlying the exit poll and the sampling methodology have to be accurate or exit polls will contain sampling errors (different from the MOE) that may make the exit-polls less accurate than the actual vote in determining the value of I.

If you read the NEP paper and the rebuttal by the group of academics in http://uscountvotes.org/ucvAnalysis/US/USCountVotes_Re_...
you will see that the statistical model used by Mitofsky/NEP only broke down in the "within-precinct" error (WPE). Mitofsky claimed that this was likely due to the Reluctant Bush Responder. But the uscountvotes paper uses Mitofsky's own data to show that this hypothesis is unlikely.

Now this opens the door, but does not prove the possibility that I was significantly different from R due to the variables F (fraud), M (mistaken votes) and S (spoilage). The reason that it does not prove it, is that there are other possible systematic errors in the exit-polling that have to ruled out (everything from possible poor design of the questionnaires to effects of the actual poll-takers on the people filling out the form to inability to correctly obtain data for absentee voters). However, this does not detract from the fact that the door is open to investigate F, M and S and that some combination of these may have actually caused R to deviate significantly from I to cause the wrong man to have been elected President.

Now, the best possible followup work would be to identify precincts statistically where I may have deviated significantly from R and then conduct carefully scrutinized full hand recounts there.







Printer Friendly | Permalink |  | Top
 

Home » Discuss » Topic Forums » Election Reform Donate to DU

Powered by DCForum+ Version 1.1 Copyright 1997-2002 DCScripts.com
Software has been extensively modified by the DU administrators


Important Notices: By participating on this discussion board, visitors agree to abide by the rules outlined on our Rules page. Messages posted on the Democratic Underground Discussion Forums are the opinions of the individuals who post them, and do not necessarily represent the opinions of Democratic Underground, LLC.

Home  |  Discussion Forums  |  Journals |  Store  |  Donate

About DU  |  Contact Us  |  Privacy Policy

Got a message for Democratic Underground? Click here to send us a message.

© 2001 - 2011 Democratic Underground, LLC