The Help America Vote Act (HAVA) was passed in 2002 as a response to some of the issues that arose during the 2000 United States presidential election. One of the goals of HAVA was to ensure that individuals with disabilities can vote privately and independently. Since HAVA’s inception, there have been several attempts to make voting accessible, but very little real progress has been demonstrated. For example, most polling places make an official available to help those with visual disabilities to write in a candidate’s name. To our knowledge, there is no voting system today that allows blind voters to write in a candidate’s name independently and privately.
An accessible proof-of-concept system, Prime III, incorporates a universal design allowing those with disabilities, including visual and motor impairments, to vote and to write in a candidate’s name privately and independently. Prime III uses a multimodal interface design to accommodate a greater number of people with disabilities. Voters can use touch or voice to cast their votes with Prime III.
The requirement that voting systems keep voter selections private poses a problem for voting systems that integrate speech interaction into the voting process; bystanders may have the opportunity to overhear a voter’s selection. To preserve ballot anonymity, bystanders must not be able to hear whom a voter selects for any office or the voter’s decision on any proposition. Therefore, during the voting process, voters cannot simply say the name of the candidates for whom they wish to vote.
The speech interface of Prime III implements an interaction in which the voter does not need to verbalize a candidate’s name. Each option is presented to the voter through a headset in random order; when the voter hears the desired option, he or she says “vote” or simply blows into the attached microphone. When voters do not wish to choose the current option, they do not say anything, and the system moves on to the next prompt. An example dialogue follows:
Prime III: “To vote for the Republican Party, say ‘vote.’”
Voter says nothing.
Prime III: “To vote for the Democratic Party, say ‘vote.’”
Voter responds: “Vote.”
In this example, the voter chose to vote for the Democratic Party. With this type of interface, voters make their selections by simply saying “vote,” and bystanders only hear the word “vote,” which ensures the privacy of the voter and the anonymity of the voter’s ballot.
In a similar fashion, the write-in option can be presented to the voter through speech. When there is no candidate on a ballot for whom the voter wants to vote, the voter must spell the name of the candidate they intend to write in. Prime III uses alphabet clustering (a modified binary search) to traverse the letters of the alphabet and name prediction to reduce the length of time a voter would normally take to spell a candidate’s name.
The ultimate goal of electronic voting systems today should be a single design that allows everyone to vote privately and independently. While Prime III does not realize that goal for all voters, it does provide a mechanism for many voters with disabilities to cast their ballots with greater privacy and dignity than most currently used systems.
Comments are closed.