Re: Question re. Op Scan error rates vs. Punch Card error rates

From: Douglas W. Jones <jones_at_cs_dot_uiowa_dot_edu>
Date: Tue Aug 16 2005 - 10:04:27 CDT

On Aug 15, 2005, at 10:34 PM, Kathy Dopp wrote:

> Thank you very much to those who gave me the answers I needed re.
> Diebold's redundancy.
>
> I also need help with this:
>
> Utah's Lt. Governor also claims that
>
> "The average optical scan system error rate was 5.69%, far worse than
> the average punchcard error rate of 3.83%"

It looks like he got his figures from Florida 2000, for example, from:
     http://www.usccr.gov/pubs/vote2000/report/ch8.htm
     Chapter 8: The Machinery of Elections
Look at Chart 8-1, about 1/3 of the way down that web page. He's
quoting the number for Optical (central), while HAVA asks us for
Optical (precinct), where the Florida 2000 data shows the rate is
under 1%.

That 6% rate is typical of central-count optical scan systems when
they are administered in a rather nasty way. The folks in
Arlington Virginia were seeing those kinds of rates on their
absentee ballots prior to instituting a new procedure.

The new procedure dropped the undervote to under 1% on the top
of the ticket. The procedural change was to have the ballot
counter kick back all overvotes and blank ballots for examination
by the canvassing board. This has been done all along in
Miami-Dade County. When you do this, you find that most of the
kicked-back ballots are easy to interpret and the rate goes way
down.

Precinct-count optical scan ballots (where the scanner is in the
precinct) must, under HAVA, be programmed to kick back blank and
overvoted ballots so the voter can correct them. When this is
done, the precinct-count systems have under a 1% undervote on the
top of the ticket.

Of course, you want the voter to fix their ballot, not the canvassing
board, but with absentee ballots, you have little choice.

                Doug Jones
                jones@cs.uiowa.edu

_______________________________________________
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Wed Aug 31 23:17:26 2005

This archive was generated by hypermail 2.1.8 : Thu Sep 15 2005 - 11:44:12 CDT