[E-voting] Question on optical scan

[Margaret McGaley]

As I said, I'm becoming more and more convinced that OpScan is a cleaner
implementation of VVAT than DRE+printer. If I was going to try to implement
e-voting, OpScan would be my only choice.

With regards to the number recognition problem, I think stickers are the
answer. With the ballot paper, the voter would be given one sheet of stickers
with unique numbers from one to the number of candidates running. The glue on
these stickers would be designed such that they can be removed and restuck
within the first 30 seconds - 1 minute, but after that removal would rip the
paper they're stuck to. You could have one scanner in each polling station for
voters who want to reassure themselves that their ballot will scan correctly.

Another, more expensive, option would be to have some device which helped
voters cast their vote, but didn't record the vote electronically. This system
would have more user interface issues, I feel, and would seem like a waste
(you can imagine a response that said "why doesn't the printer just save the
vote"), so I'd be more comfortable advocating the stickers.

Margaret




On Wed, Jan 19, 2005 at 01:44:45PM +0000, Brian O'Byrne wrote:
> I'd like to pose a question for the group, if I may, on optical scan systems.
> 
> As I understand optical scan systems the voter places a mark representing their
> vote on a pirce of pre-printed paper. This is then scanned and the scanning
> system decides whether the vote appears to be valid. It may also present the
> voter with a copy of the vote on a screen, to verify that the scanner has
> correctly read the voter's intent.
> 
> The advantages include:
> - An undisputed paper trail. The paper is the vote and there is no confusion or
> ambiguity.
> - The possibility of using the scanned records to get a counted result quickly.
> - The possibility to show that the scanned records match the votes, either by
> random sampling or by a full paper-based recount.
> - A reduction in accidental spoiled ballots. Spoils can be recognised as such
> before the voter leaves the polling station without violating privacy.
> 
> The question is: Has anyone seen or designed a optical scan ballot for an STV
> election?
> 
> It seems easy enough to design one of these for ballots where you choose one of
> a number of options. Put a box beside each option and have the voter mark one
> box. That is simple and reliable.
> 
> It seems less easy where the voter has to express an order of preference. Do you
> have many boxes beside each candidate (mark the leftmost box for first
> preference) or do you have OCR that tries to read numbers, or some other
> system?
> I'd be concerned that OCR has a fairly high failure rate, and the many boxes
> option is confusing for voters.
> 
> Does anyone have experience of this?
> 
> Thanks,
> Brian.
> -- 
> Brian O'Byrne, Statesoft Ltd.
> Tel: +353 1 449 8151, +353 86 240 4719
> http://www.statesoft.ie/
> 
> _______________________________________________
> E-voting mailing list
> E-voting at lists.stdlib.net
> http://lists.stdlib.net/mailman/listinfo/e-voting
> http://evoting.cs.may.ie/
> 

-- 
Margaret McGaley
Margaret.McGaley at redbrick.dcu.ie
http://evoting.cs.may.ie