Blind People and Mobile Touch-based Text-Entry: Acknowledging the Need for Different Flavors
João Oliveira, Tiago Guerreiro, Hugo Nicolau, Joaquim Jorge, Daniel Gonçalves · 2011 · The Proceedings of the 13th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS) · doi:10.1145/2049536.2049569
Summary
This paper investigates how individual differences among blind people affect their performance with different touchscreen text-entry methods. Thirteen blind participants (ages 24-62) were evaluated using four methods on a Samsung Galaxy S: QWERTY (standard VoiceOver-style keyboard with split-tap selection), MultiTap (phone keypad-style layout with 12 medium-sized buttons), NavTouch (gesture-based navigation using directional swipes with adaptive positioning), and BrailleType (six large targets representing Braille cell dots with long-press selection). The study comprised two phases: characterising participants' individual attributes through sensory tests (pressure sensitivity via Semmes-Weinstein monofilaments, spatial acuity via Disk-Criminator), cognitive assessments (verbal IQ via WAIS-R Digit Span, spatial ability via Planche a Deux Formes), and functional evaluations (mobile phone, computer, and Braille reading/writing speed); followed by text-entry performance measurement with all four methods in a within-subject design with counterbalanced ordering.
Key findings
The four methods showed a clear speed-accuracy tradeoff: QWERTY was fastest (2.1 WPM) but most error-prone, while BrailleType was slowest (1.49 WPM) but most accurate. Individual differences had statistically significant impacts across multiple dimensions. Age of blindness onset was a major factor: congenitally blind users and those who lost sight before age 6 performed worst across all methods, particularly with QWERTY and MultiTap, but the gap was smaller with NavTouch and BrailleType. Pressure sensitivity significantly affected MultiTap performance, as the sensitive touchscreen required precise touch control for multi-tapping sequences. Spatial ability had the strongest impact on QWERTY and MultiTap (methods requiring screen exploration) but no effect on NavTouch and BrailleType (methods not requiring spatial knowledge of button positions). Verbal IQ significantly affected all methods, with lower verbal IQ associated with slower performance and more errors, particularly on MultiTap and NavTouch which demand more short-term memory and attention. Prior Braille reading experience significantly improved BrailleType speed. Case studies illustrated how the same individual attribute profile could make one method optimal and another impossible — one 62-year-old participant with poor spatial ability and pressure sensitivity could not complete QWERTY or MultiTap tasks at all but performed adequately with NavTouch and BrailleType.
Relevance
This research makes a compelling case against one-size-fits-all text-entry solutions for blind users. The finding that spatial ability, pressure sensitivity, verbal IQ, age of blindness onset, and prior experience with specific input paradigms all significantly influence which method works best for a given individual has direct implications for mobile platform design. Rather than providing a single accessible text-entry method, platforms should offer multiple options and ideally recommend methods based on user profiling. For accessibility practitioners and developers, the study provides a framework for understanding why a text-entry method that works well for one blind user may fail completely for another. The work also contributes to the broader ability-based design movement by demonstrating concrete, measurable relationships between specific sensory and cognitive capabilities and interface performance. The characterisation methodology — combining standardised clinical tests with functional evaluations — offers a replicable approach for other researchers studying individual differences in assistive technology use.
Tags: blindness and low vision · text entry · touchscreen accessibility · mobile accessibility · ability-based design · personalization · braille · input methods