Poll Worker Training Blamed for Voting Snafus

Proof that plain language requirements apply especially to pollworker materials and training:

The Associated Press ran a story this week about the number of polling problems that can be attributed to poor poll worker training. The snafus include hiding electronic voting machines because the workers did not like them; handing out the pens used for touch screens to mark paper ballots and calling them "invisible ink pens." Poll workers have provided wrong information - especially about this year's spate of confusing primaries - and inadvertently disenfranchised voters.

The story notes the concerns of the League of Women Voters, whose researcher, Lloyd Leonard, notes that we are "running the most important part of our democracy on the backs of untrained, poorly paid volunteers." Some two million pollworkers form the largest one-day work force, according the Pew Center on the States. Most receive just a few hours of training.

The training itself can fall far short: some did not receive instruction about operating the voting system in the precinct they were working. Some did not have an opportunity to actually practice on the machine: election day was their first experience following training.

UPA's Voting and Usability Project members note that good instructional material doesn't replace good training. Too many pollworkers are put in situations where they aren't well prepared and where they don't have the information they need. Usability and technical support professionals both know how "creative" people can be in trying to solve problems. We can't change human nature, but we can understand the stresses and problems of running an election, and be prepared.

The constraints of running an election often mean that pollworkers will need to fall back upon their instruction manuals and other written materials. Election officials are encouraged to 1) review their instructional materials, 2) clarify them with the plainest language possible, and 3) test their instructions with poll workers using a process similar to the LEO usability testing process for ballots.

More information:

http://news.yahoo.com/s/ap/20080223/
ap_on_el_ge/pollworker_problems

Human-Centered Voting Presented

Auburn University Distinguished Professor Juan Gilbert and his research team from the Human Centered Computing Lab will join other scientists presenting innovative voting technologies at the Future of Voting Forum on Capitol Hill on March 6th from 9am – Noon. Professor Gilbert’s Interactive Voice Response voting system allows voters with varying abilities to vote equally, while simultaneously providing a voter-verifiable audit trail.

More: http://www.itif.org/index.php?id=124

GAO Says 13th Congressional Machines Test OK -- but They Fail to Usability Test

The General Accounting Office (GAO) conducted functionality tests of Florida’s 13th Congressional District race, and determined that the machines appear to be functioning properly. The 13^th Congressional District came under scrutiny in 2006 when some 18,000 undervotes resulted in scrutiny from the US House of Representatives.

After determining the source code in escrow produced the same firmware, the GAO tested the function of a sampling of machines throughout the district. They concluded that “(1) the machines properly displayed, recorded, and counted the selections for all test ballots cast during ballot testing involving 112 common ways a voter may have interacted with the system, and (2) the deliberately miscalibrated machines, though difficult to use, accurately recorded the ballot selections as displayed on screen.”

The GAO notes, however, that “the large undervote in Florida's 13th Congressional District race could have been caused by factors such as voters who intentionally undervoted, or voters who did not properly cast their ballots on the iVotronic DRE, potentially because of issues relating to interaction between voters and the ballot." (Our emphasis)

We are Disappointed

UPA’s Voting and Usability project members are dismayed that the GAO missed their opportunity to promote usability testing as a means to answer those questions, as well as the likeliest method for preventing serious usability obstacles in ballot design.

Usability testing would have helped identify the likely interaction problems real voters would encounter, including the likelihood that voters might fail to vote in the US House race. Although the GAO has an expert review that suggests this problem could have been predicted, usability testing would confirm and quantify those indications.

What’s more, we wish the GAO would promote usability testing before elections to help prevent such problems. The federal Voluntary Voting Systems Guidelines and certification can only cover the capabilities of a voting system. The ballots for each election are produced by the local election officials, and require separate, easy to implement testing, such as the testing method available through the LEO Usability Testing kit.

More about the GAO report: http://www.gao.gov/new.items/d08425t.pdf

More about LEO: http://www.upassoc.org/civiclife/voting/leo_testing.html;

More about the Voting and Usability Project: http://www.upassoc.org/civiclife/voting/index.html

More about the VVSG: http://www.eac.gov/voting%20systems/
voting-system-certification/2005-vvsg

More about Banner Blindness in Ballot Design - Jakob Nielsen, Feb. 2007
http://www.useit.com/alertbox/banner-blindness-ballot-design.html

California Ballots Still Being Counted

A combination of flaws in the Feb. 5 California primary ballots has resulted in a vote count that continues, according to the New York Times. It reports that “election officials say a combination of high turnout, technology flaws and millions of mailed-in and dropped-off ballots have led to painstakingly slow returns in some counties, with nearly 800,000 ballots remaining to be processed.”

Among the problems noted by the Times:

• In Los Angeles County, 205,000 ballots were waiting to be tallied. Some 49,000 of those are being examined because voters failed to mark a bubble indicating the party they wished to vote for, but properly marked a vote for a candidate. (Some consider that “extra bubble” to be a design flaw that was incorporated when the ballot was adapted from punch card, but unnecessary for this election.)
• Absentee ballots in Contra Costa County are being ironed so that they can be fed into the vote-counting machines.
• Damage from warehousing the ballots, weather and mishandling by voters means that some ballots have to be handled with special, time-consuming care.
• High turnouts are overwhelming the already-stressed election officials, who must juggle three major elections this year.

More: (may require registration)
http://www.nytimes.com/2008/02/17/us/politics/17vote.html
?_r=1&sq=california%20primary%20ballots&st=nyt&adxnnl=1&oref=slogin&
scp=1&adxnnlx=1203429495-hjwTTFk69FtX/kieWsuWlg

Ballots Not Counted in LA Primary Because "Extra Box" Not Marked

In addition to the normal confusion over voting in presidential primaries, the New York Times notes that some independent Los Angeles County voters did not have their vote counted because they failed to mark "an extra box" on their ballot.

http://tinyurl.com/2hp9tp
(Registration may be required.)

The article notes "One of the more significant problems occurred in Los Angeles County, where independent voters in at least 15 precincts said they were never told they had to mark an extra box on their ballots for them to be counted."