RE: What is Data Model FOR?

From: Mark Winegar <mwinegar_at_mtmc_dot_edu>
Date: Thu Apr 29 2004 - 15:46:35 CDT

-----Original Message-----
From: Arthur Keller [mailto:arthur_at_kellers_dot_org]
Sent: Thursday, April 29, 2004 3:29 PM
Subject: RE: What is Data Model FOR?

At 3:07 PM -0500 4/29/04, Mark Winegar wrote:
>Well, let's consider the nature of a spoiled ballot. Isn't it a ballot
>that is invalid because a choice has been made that presents an
>exception to the rule. We can capture the exception through validity
>testing before the data is entered and allow the elector to make a
>valid choice thereby saving the ballot. I think that's a good thing.

Certainly we will do that, and DRE vendors make the same claim. Our
concept is that voters *can* make mistakes and discover them when the
review the printed ballot. We *have* to handle that case.

Why wait that long? Why not capture it at the point of entry and provide
for a correction?

>I was thinking it would be more valid in that it faithfully count to
>contests acurately and it would also represent those ballots that
>somehow didn't get into the ballot box.

That's one of the criticisms we get from DRE vendors.

Isn't that all the more reason to use and compare multiple canvassing

>Which brings forward another
>question, "How do we prevent hardcopy ballots from be lost?"

May I suggest you review the discussions we've had already on these

I'd like to. Where can I access them?

Best regards,

>-----Original Message-----
>From: Arthur Keller []
>Sent: Thursday, April 29, 2004 2:00 PM
>Subject: RE: [voting-project] What is Data Model FOR?
>The computer count is NOT more accurate than the paper count. The
>computer count includes spoiled paper ballots.
>Please think it through and make suggestions on HOW it should be done.

>We already know that it needs to be done.
>Otherwise, we're just spinning our wheels and getting nowhere.
>Best regards,
>At 1:28 PM -0500 4/29/04, Mark Winegar wrote:
>>Yes, there has to be a better process of managing the paper ballots
>>because it appears that the computer count is more accurate than the
>>paper count. In what ways can the paper count be improved?
>>Local reconciliation probably isn't a bad thing.
>>-----Original Message-----
>>From: Arthur Keller []
>>Sent: Thursday, April 29, 2004 11:13 AM
>>Subject: RE: [voting-project] What is Data Model FOR?
>>At 10:32 AM -0500 4/29/04, Mark Winegar wrote:
>>>Better accuracy is often achieved by eliminating the humor error
>>>In both scenarios discussed below there are opportunities for the
>>>error factor to contaminate the canvassing. We can improve the whole
>>>process by having two distinct canvassing processes; one based on the

>>>voter's computer input and the other based on paper ballots. I
>>>the comparison of these distinct data sets will help detect, and
>>>prevent, tampering.
>>>Mark Winegar
>>Actually, our approach is to reconcile the paper and electronic as
>>early as feasible after the polls close. Once you've done that,
>>distinct canvassing systems compounds the discrepancies. There can be

>>discrepancies between paper and electronic, such as spoiled ballots.
>>Or ballots never deposited in the ballot box. Unless you reconcile as

>>locally as feasible, it's hard to effectively reconcile in larger
>>contexts. And when you expect there to be discrepancies (spoiled
>>ballots), you have to drill down anyway.
>>Best regards,
>>Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA 94303-4507
>>+1(650)424-0202, fax +1(650)424-0424
>Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA 94303-4507 tel
>+1(650)424-0202, fax +1(650)424-0424

Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507 tel
+1(650)424-0202, fax +1(650)424-0424
= The content of this message, with the exception of any external 
= quotations under fair use, are released to the Public Domain    
Received on Fri Apr 30 23:17:25 2004

This archive was generated by hypermail 2.1.8 : Fri Apr 30 2004 - 23:17:29 CDT