Exploring the Potential of Using Spatial Audio to Improve Web Accessibility for Screen Reader Users
- Author(s): Wang, Tao
- Advisor(s): Redmiles, David
- et al.
The web has the potential to greatly improve the quality of life for people with visual impairments. Unfortunately, there are still many web accessibility issues that prevent this vision from being realized. In this research, I explore the potential of using spatial audio feedback to improve screen reader users’ experiences browsing web pages. There are three main components to this research. They represent three key elements in problem solving: identifying causes, designing solutions, and evaluating outcomes.
First, I used a text analysis study to determine the specific communication barriers between sighted and blind web users. By analyzing written instructions produced by 48 sighted people, I identified that the use of spatial terms, among other things, was a main challenge for screen reader users. The result led to a proposal of using spatial audio cues to convey layout information to screen reader users.
Second, to learn the most effective spatial audio design patterns, I designed two lab experiments to uncover what spatial properties contribute to positive recognition of spatial audio cues. Based on more than 4000 data points collected from 18 sighted participants, I obtained an intimate understanding of how users perceived and interacted with audio cues spatialized in the horizontal audio space.
Finally, informed by both studies, I developed a proof-of-concept screen reader prototype that provides additional spatial audio feedback to convey web page layout information. To evaluate the design, I conducted a user study with 20 blind participants. Participants completed layout-related tasks during study sessions. Based on a range of data collected, including task performances, interviews, surveys, and keystroke logs, I learned that participants responded positively to the design. They believed that spatial audio feedback enabled them to deal with spatial terms and the newly acquired layout information could help them handle unexpected accessibility incidents more effectively. Participants also provided feedback on the prototype’s usability and envisioned how similar technologies could assist them in other scenarios. Inspired by the results, I reflect on implications for assistive technology designs and accessibility trainings. I also chart the research agenda in this undeveloped area and discuss promising future research topics.