Advanced Auditory Menus: Design and Evaluation of Auditory Scroll Bars
Pavani Yalla, Bruce N. Walker · 2008 · Proceedings of the 10th International ACM SIGACCESS Conference on Computers and Accessibility (Assets '08) · doi:10.1145/1414471.1414492
Summary
This paper from the Sonification Lab at Georgia Institute of Technology investigates how to design auditory scrollbars — non-speech sound cues that convey contextual information (menu size, current position) while navigating auditory menus. Visual menus use scrollbars, icons, lines, and shortcuts to communicate structure beyond just the menu items themselves, but auditory menus typically offer only text-to-speech of item names with no equivalent contextual feedback. The authors systematically designed four auditory scrollbar types by adapting visual menu concepts: single-tone (one pitch per item, changing proportionally to position), double-tone (adds a second reference tone for the menu boundary), alphabetical grouping (pitch changes at each new letter of the alphabet), and proportional grouping (pitch changes every fixed number of items). Three studies were conducted across sighted and visually impaired participants, progressively refining the designs based on user feedback. The research used a simulated mobile phone address book with 50 names, combining text-to-speech for item names with pure tones for scrollbar feedback.
Key findings
Across all three studies, pitch polarity (whether pitch increases or decreases while scrolling down) did not significantly affect performance or preference — participants were evenly split and generally preferred whichever they heard first. This finding contrasts with some prior sonification research, possibly because address book content lacks inherent spatial or ordinal mapping to pitch. The proportional grouping scrollbar consistently outperformed other designs. In Study 2, participants estimated menu sizes with only 10.1% mean error using proportional grouping versus 47.4% for alphabetical grouping (statistically significant, p = .001). In Study 3 with visually impaired participants using a real Nokia N95 phone, proportional grouping again yielded the lowest error (5.7%, SD = 2.0) compared to single-tone (51.9%) and double-tone (75.3%), with a significant ANOVA result (F(2,14) = 4.00, p = .086 by Pairwise comparisons p < .05). Notably, visually impaired participants who had lost vision later in life and were already familiar with visual scrollbar concepts were more enthusiastic about auditory scrollbars than those blind from birth, suggesting that visual analogies in auditory design particularly benefit users with prior visual experience.
Relevance
This research provides practical design guidance for making menu-based interfaces accessible through audio. The finding that proportional grouping is most effective has direct implications for designing auditory interfaces on mobile devices, smart speakers, and any system where visual display is unavailable or inaccessible. For accessibility practitioners, the study demonstrates the value of universal design principles — the auditory enhancements benefit both sighted and visually impaired users. The observation that users with different vision histories respond differently to auditory scrollbars highlights the importance of recognizing diversity within the visually impaired user population when designing accessible interfaces. The work also underscores that accessibility is not just about providing content (via TTS) but also about conveying the structural and contextual information that sighted users take for granted.
Tags: auditory interface · auditory menu · sonification · non-speech sounds · universal design · visual impairment · mobile accessibility · earcon