----- Original Message -----
From: Thom Wysong <tgw@...>
To: Lorrie Faith Cranor <lorrie@...>
Sent: Thursday, December 14, 2000 3:44 PM
Subject: 30 voters/hour/unit for touch voting systems

John Seibel wrote:
"even in relatively simple elections (2or 3 offices), we can move a maximum
of 30 voters/hour/unit through a touch screen system"

I concur with John's estimate of "a maximum of 30 voters/hour/unit" for a
touch voting system.

I worked as an Election Official this year for the November US Election in
Arlington, Virginia. In our county, we use Shoup DRE machines. They are not
"touch screen" systems. However, they are electro-mechanical "touch" voting
systems. All of the contests and voting options are printed on a large
sheet of paper which faces voters as they enter the voting booth.
Underneath the paper are small buttons and lights which are activated when
the voter presses in designated spots on the large paper sheet. So, from a
voter's perspective, it is very similar to touch screen systems - except
voters don't need to page through different screens to see the whole ballot.

At the polling place I worked, we had a long line of people waiting to vote
from the moment we opened the doors at 6 AM, until we closed them at 7 PM.
There were so many people in line at 7 PM, that it took us another hour and
a half to process them all through. So, all five of our Shoup DRE machines
were 100% occupied for 99.9% of the time from 6:00 AM until 8:25 PM.

There were just over 2500 voters. Each machine handled almost exactly 500
voters. This comes out to 1.73 minutes per voter (865 minutes / 500 voters)
and 34.7 voters/hour/unit.

For our ballot, there were a total of 13 contests to decide on (6 offices,
2 amendments to state constitution, and 5 county referenda), up to 14
selections to be made (for School Board, 2 selections were allowed), and 34
total options to choose from. So, it was more complex and time-consuming
than only voting on "2 or 3 offices".

However, the extra time needed to vote on the extra options may have been
off-set by a small amount of time saved from not having to page through the
ballot, as is needed with regular touch-screen systems (which I prefer, for
the record). So, the experiences in Arlington's precinct #14 this year seem
to closely match John's TrueBallot experiences.

The "maximum of 30 voters/hour/unit" for touch voting systems seems to be a
valid estimate for future reference.

-Thom Wysong



=====================================================================
This message was distributed through the e-lection mailing list.
For info and archives see http://www.research.att.com/~lorrie/voting/
=====================================================================