Users with Motor Disabilities Adapting to Ordinary Keyboards, and Vice Versa

Shari Trewin, Dept. of Artificial Intelligence, Univ. of Edinburgh.

Abstract

Many computer users with motor disabilities choose to use keyboards for input, but find the physical movements required difficult. These difficulties can cause frequent input errors such as repeated characters, or extra characters. Such errors can often be minimised or eliminated by employing existing software configuration facilities. For example, Repeat Keys is a widely available configuration facility which allows users to control the repeating behaviour of keys. While such facilities are now integrated into the majority of modern operating systems, there is little published data describing their effect on keyboard usability. For some people, keyboards are unusable without these facilities, but the facilities themselves can be difficult to use. This talk will introduce some of the common input errors due to physical difficulty in manipulating keyboards, and the facilities currently available to compensate for these difficulties.

Unfortunately, the configuration facilities are often underused, due to a lack of awareness of their existence, and the percieved or actual difficulty of employing them, particularly on shared machines, where one person's ideal configuration may conflict with that of another.

Automated support for keyboard configuration could help to solve this problem. As an initial step in this direction, we have developed a dynamic model of four important areas of keyboard skill, capable of making recommendations of appropriate configurations. The model operates by observing a user typing free English text.

Evaluation of the model involved twenty keyboard users with motor disabilities and ten with no disability affecting their typing. The results, which will be presented, show that automatic configuration recommendation is a feasible proposition. If time allows, potential applications of this model will be briefly discussed.

For more information about this talk please contact shari@dai.ed.ac.uk