Group project for 2018 CHI student design challenge
Fall 2017 (3 months)
Team co-lead, UX researcher/UX designer
Interview, Sketching, Wireframing, Prototyping,
Absentee Voting Assistant (AVA) is a smartphone app that works in conjunction with screen readers to help the visually impaired individuals vote independently at home by operating the touchscreen or by voice control.
SOLUTION AT A GLANCE
Voting tool for the visually impaired
How might we improve the accessibility of voting for the visually impaired population
Inconvenience of traveling
Traveling to the polling site is not convenient for the visually impaired
Often require outside help
Have to require outside help in reading and writing when voting
Disclose their choices
Those with vision impairment may face issues in keeping their votes private
We spent the first 2 months conducting the background research into the voting system in the United States especially regarding current policies, methods, facilities and voting procedures. We also studied the behavior of the visually impaired, particularly on how they interact with devices from computers to smartphones.
To better understand the voting procedures and to learn how the current facilities accommodate the visually impaired, we went to the polling site for a general election for our state to observe two of our interviewees voting.
Pictures of polling site in Ann Arbor
Permission was obtained for taking those pictures
We conducted interviews on five experts from different organizations and communities for guidance, including Washtenaw County Clerk and Ann Arbor Center for Independent Living (CIL).
Personalized interview protocol
Each interview was presided by 2 people: one to ask questions, and another to take notes
Recorded each interview and conducted interpretation sessions to discuss the main points with team
We conducted the competitive analysis to research the current available products that assist the blind in filling out the ballot. The aforementioned accessible voting machine gave us a starting point of how to instruct users to vote step by step. Using this as a base, we developed a set of instructions that the solution would give. Due to the lack of relative products in assisting visually impaired in voting, we researched how other products help people with vision impairment write and read, which are key abilities for filling out a ballot.
Pain points of visually impaired
People with vision impairment prefer using voice interaction, such as Voiceover on iPhone and Screen Reader on the computer.
The newly installed accessible voting machine is designed to aid all sorts of disabled people. However, we noticed that setting up took a significant amount of time due to the staff’s lack of experience using the machine which frustrated our interviewees.
People with vision impairment preferred absentee voting for its convenience.
There are “a lot of politics” that prevent lawmakers from creating more laws that improve accessibility at the polls.
We created two personas reflecting the main groups of users we targeted at: people who are completely blind and people who has vision impairments.
Sketching & Brainstorming
Sketching allowed us to visualize how the solutions would be eventually created. We started to sketch how users interact with the devices once we brainstormed some ideas.
Through the brainstorming and discussion session, we created the interface and the voice commands used in the app.
Prototyping & Iteratively polish
After validating our design concepts with our target users, their feedbacks led us to create our prototypes. We went through 2 rounds of paper prototyping, 3 rounds of medium- fidelity, and 1 round of high-fidelity. Our paper prototypes focused on concept validation.
Our medium-fidelity rounds were more focused on testing the usability of our mobile app, the information architecture, and our voice commands. And our high-fidelity prototyping phase was more concerned about the visual design.
During the medium-fidelity prototyping phase, we got the chance to test our solutions with the visually impaired community. It consisted of 2 rounds with a total of 7 participants.
To better understand how our app accommodates people with vision impairment and improve our design, we asked our testers to fill out the entire ballot. The result offered valuable insight on how blind individuals react to certain cues, but also guided us in solidifying our final solution.
Some of the most important elements of our solution came from user testing. For example, our decision to include touch-based input in addition to voice input was an implementation of suggestion.
Pictures of conducting usability testing with one of our users
Within a user-centered design process, we designed Absentee Voting Assistant (AVA), a mobile application that supports the blind and visually impaired individuals independently vote at home.
This app allows users to cast their vote by means of voice input or button input on the screen. AVA then prints out user’s choices on the original absentee ballot that he/she receives from the mail. The app bypasses the difficult task of having to mark candidate choice on a ballot by controlling the process via voice or touch.
Process of using AVA to vote
Black & white to increase readability
The entire app is in black and white to increase readability. While visually unimpressive, black and white was the only viable choice for two reasons:
Black text on a white background creates a strong contrast.
It is readable for those with color blindness as well.