The One-Key Challenge: Searching for a Fast One-Key Text Entry Method
I. Scott MacKenzie · 2009 · Proceedings of the 11th International ACM SIGACCESS Conference on Computers and Accessibility (Assets '09) · doi:10.1145/1639642.1639660
Summary
This paper presents SAK (Scanning Ambiguous Keyboard), a novel one-key text entry method designed for people who can only operate a single switch input. The design combines two established techniques — scanning keyboards (where items are highlighted sequentially and selected with a single key press) and ambiguous keyboards (where multiple letters share the same key, with disambiguation through dictionary lookup) — into a unified system. SAK uses three virtual letter keys with alphabetically arranged letters (ABCDEFGH, IJKLMNOP, QRSTUVWXYZ) plus a SPACE key. The keys are highlighted in sequence, and the user presses the single physical key when the desired key receives focus. Because each virtual key maps to multiple letters, the system maintains a candidate word list based on the key sequence entered so far, using a 9,025-word dictionary. When SPACE is selected, scanning transfers to a word-selection region showing candidate words ordered by frequency. A key innovation is "double selection" — if the highlighted key bears both the desired letter and the next letter, two selections can occur in a single scanning interval, reducing input time. The paper also reviews the broader landscape of text entry with few keys, summarising nine published user studies and comparing their entry speeds, methodologies, and limitations.
Key findings
In an evaluation with 12 participants (non-disabled university students), average entry speed reached 5.11 wpm across all trials with 99% accuracy, or 7.03 wpm considering error-free trials only, after just five blocks of practice (less than one hour total). The scanning interval decreased from 1100ms to 700ms across blocks. A "timer restart on selection" modification was implemented based on participant feedback — restarting the scanning timer after each selection gave users more time and enabled multiple selections per interval. One participant performed extended trials (10 blocks) with this modification and reached 9.28 wpm by the final block with only 0.81% error rate. This exceeded the previous fastest one-key entry speed in the literature (8.4 wpm by Koester and Levine's system). The SPC (scan steps per character) metric showed SAK achieved 1.834 for the test phrases, meaning less than 2 scan steps per character on average. Analysis of the extended trials revealed 98 double selections (17.5%), 17 triple selections (3.0%), and 3 quadruple selections among 558 total letter selections, confirming that multi-selection opportunities are meaningfully frequent.
Relevance
This research is directly relevant to the significant population of people with severe physical disabilities who are limited to single-switch input — including those who can only use a head movement, eye blink, sip-and-puff, or similar binary input. Text entry speed is a critical bottleneck for communication and computer access with single-switch systems, and SAK's approach of combining scanning with ambiguous key disambiguation represents a meaningful speed improvement over traditional row-column scanning keyboards. The entry speeds achieved (up to 9.28 wpm) are still far slower than conventional typing but represent substantial progress for one-key input. The comprehensive review of prior work provides a valuable reference for practitioners selecting or developing text entry solutions for switch users. The design principles — minimising scan steps through ambiguity plus dictionary prediction, and enabling multiple selections per scan interval — are generalisable to other scanning interface designs beyond text entry.
Tags: text entry · switch access · scanning keyboard · ambiguous keyboard · motor disability · assistive technology · input methods · human-computer interaction