Re: How dark is a mark?

From: Arthur Keller <voting_at_kellers_dot_org>
Date: Thu Dec 20 2007 - 00:30:29 CST

At 12:41 AM -0500 12/20/07, Rick Carback wrote:
>Showing an image of the ballot proves nothing and is not an adequate
>form of voter verification. Showing a summary screen is useful
>however as a voter verification.
>
>
>I don't think you understood what showing the image means. You
>aren't simply showing the scan, but an interpreted representation of
>it. Specifically, the software takes the scan and indicates what it
>interpreted as an overlay on a pre-rendered image of a blank ballot
>( e.g. big green check marks, circles, etc). It can also indicate
>what marks it saw that it didn't like (overvotes, etc) with big red
>X's or some other mechanism.

Yes, I didn't understand what you meant by "showing the image." How
would you indicate undervotes?

>The advantage is that it provides a redundant verification
>mechanism. The voter looks for the name on the screen, that the
>position matches, and (if possible) hears the name played back to
>them. If the ballot is somewhat transparent, the voter could hold
>the ballot over the screen as a visual overlay. Other graphical
>elements might also be useful ( e.g. flashing the name/race as it is
>being played on the screen).

Yes, those are useful.

>The scanner might provide auditory feedback about scans and their
>status. The scanner might also require a button being pressed for
>each ballot scanned, with the button being accessible to an election
>official and not to any voter. Doug, would this fix the problems
>you've identified?
>
>It might be something, but there are potential problems. You might
>be able to jam multiple printed ballots into the scanner at the same
>time (the voter still has private physical access). The scanners
>totals will then be correct, but the paper trail won't. Also, an
>election official might press the button even when the voter doesn't
>accept the ballot. In my view, giving the voter the ability to press
>that "I accept/reject" button is important. It signifies their
>agreement with what was presented to them as being correct.

First, I was thinking that pushing the button enables someone to feed
a ballot into the scanner.

Second, ballots may be printed on both sides, so both sides need to
be scanned. Ideally, that is done without flipping over the ballot,
but rather by having a top and bottom scanning mechanism.

Third, my idea includes having a tracking number (NOT known to the
voter) printed on the ballot when the ballot is accepted by the
voter. This tracking number is linked to the electronic records.
The tracking number is randomized to facilitate voter privacy. It
now occurs to me that the tracking number should be printed on both
sides of the ballot.

Fourth, perhaps shining a light through the ballot is a way to detect
whether multiple ballots are fed in at once. It's the technique
that's used to check that two twenty-dollar bills are not stuck
together in an ATM (which would result in dispensing the wrong amount
of cash). The ballot weight could enable some (small amount) light
to go through.

Fifth, we need a way to handle ballots that take multiple sheets.

Best regards,
Arthur

>
>
>Best regards,
>Arthur
>
>
>At 3:50 PM -0500 12/19/07, Richard Carback wrote:
>>Re. the security. I was told they were feeding it into the scanner in
>>public, but the review would be private. I got the impression that you
>>would scan, take your ballot into the review screen, press yes or no and
>>then drop your ballot into a separate, public, ballot box. While I have
>>my ideas about it, i'm not quite ready to pass judgment on the system's
>>security... As you've said, it is not as thought through as you'd need
>>to really evaluate it.
>>
>>The image might be useful because the voter could take the ballot and
>>compare side by side -- some graphical highlighting of the choices would
>>also be interesting to see. Also, an audio playback, too..
>>
>>-R
>>
>>Douglas W. Jones wrote:
>>> On Dec 19, 2007, at 11:17 AM, Rick Carback wrote:
> >>
>>>
>>>> Perhaps I have misheard, but isn't one of the ideas behind the
>>>> scanner supposed to be that it present a screen showing what the
>>>> scanner read on the ballot?
>>>>
>>>
>>> It's clear that scanners can auto-calibrate, recognizing background
>>> illumination levels of the paper, recognizing unmarked voting
>>> targets, and recognizing targets that are marked. This idea is an
>>> old one that needs to be used more frequently.
>>>
>>> But, scanners that present a screen showing what the scanner
>>> read on the ballot, as I understand the idea, doesn't involve showing
>>> an image of the ballot, but rather, a summary screen saying, perhaps,
>>> "Fred Smith for President". This has generally been seen as
>>> having the same value as the confirmation page in a DRE voting
>>> dialogue. It may have some value, but it does not eliminate the need
>>> to check or audit the scanning, since we know that only a minority
>>> of voters will notice inaccuracies in the reading of their ballots.
>>>
>>> And there is a problem. For security reasons, you want ballot
>>> scanners to be out in the open, in plain sight of everyone. You
>>> don't want the scanner in a booth where someone could feed multiple
>>> ballots into it or tamper with the machine. This issue has not been
>>> thought through, as nearly as I can tell.
>>>
>>> Doug Jones
>> > <mailto:jones@cs.uiowa.edu>jones@cs.uiowa.edu
>
>--
>-------------------------------------------------------------------------------
>Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA 94303-4507
>tel +1(650)424-0202, fax +1(650)424-0424
>_______________________________________________
>
>OVC-discuss mailing list
><mailto:OVC-discuss@listman.sonic.net>OVC-discuss@listman.sonic.net
><http://lists.sonic.net/mailman/listinfo/ovc-discuss>
>http://lists.sonic.net/mailman/listinfo/ovc-discuss
>By sending email to the OVC-discuss list, you thereby agree to
>release the content of your posts to the Public Domain--with the
>exception of copyrighted material quoted according to fair use,
>including publicly archiving at
><http://gnosis.python-hosting.com/voting-project/>http://gnosis.python-hosting.com/voting-project/
>
>
>
>_______________________________________________
>OVC-discuss mailing list
>OVC-discuss@listman.sonic.net
>http://lists.sonic.net/mailman/listinfo/ovc-discuss
>By sending email to the OVC-discuss list, you thereby agree to
>release the content of your posts to the Public Domain--with the
>exception of copyrighted material quoted according to fair use,
>including publicly archiving at
>http://gnosis.python-hosting.com/voting-project/

-- 
-------------------------------------------------------------------------------
Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507
tel +1(650)424-0202, fax +1(650)424-0424

_______________________________________________
OVC-discuss mailing list
OVC-discuss@listman.sonic.net
http://lists.sonic.net/mailman/listinfo/ovc-discuss
By sending email to the OVC-discuss list, you thereby agree to release the content of your posts to the Public Domain--with the exception of copyrighted material quoted according to fair use, including publicly archiving at http://gnosis.python-hosting.com/voting-project/
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Mon Dec 31 23:17:07 2007

This archive was generated by hypermail 2.1.8 : Mon Dec 31 2007 - 23:17:10 CST