Many people with motor impairments still find it difficult to use desktop computers despite earnest efforts in accessibility research. The literature continues to show that specialized assistive technologies are often abandoned or never acquired due to cost, complexity, configuration, maintenance, unavailability, and the need for training. Instead, commodity devices like everyday mice and trackballs are widely used by many people with motor impairments, but current graphical user interfaces are not well matched to the abilities of these users. In this project, rather than attempting the tired approach of adapting special users to standard point-and-click interfaces, the PI's goal is to radically change user interfaces to be better suited to users, by removing the two major obstacles to successful human-computer interaction for people with motor impairments: the need to point to confined areas, and the need to click within them. The PI will accomplish these objectives with goal crossing, an interaction technique in which people only need to move their mouse across a threshold without acquiring an area or executing a stable click. Although goal crossing has been studied and employed in pen-based interfaces, it has not been employed in mouse-based interfaces or for improving computer access. Goal crossing in mouse-based interfaces is challenging because of the occlusion problem, where an unwanted goal lies in front of a desired one (although the mouse button could be used to disambiguate, people with motor impairments cannot reliably drag). Nevertheless, a preliminary study conducted by the PI found that goal crossing has higher throughput, is faster, and exhibits more path and submovement accuracy than area pointing for people with motor impairments. The PI will leverage these performance findings in the current project, in the creation of accessible goal crossing user interfaces. He will do this through the following four steps: Development and evaluation of crossing widgets (this step will employ a highly iterative process to create interactive prototypes and testing them with people with motor impairments); Creation and evaluation of exemplar crossing applications (this step will build three exemplar applications for text editing, drawing, and photos that will explore crossing widget layout, grouping, and interaction); Development of a programmers' toolbox for crossing widgets (this step will create and deploy a crossing widget toolbox to enable developers to create crossing-based applications); Creation of the SUPPLE-X automatic user interface generator (an automatic user interface generator based on SUPPLE++ that is capable of creating ability-based goal crossing interfaces by measuring users' functional capabilities). This research will be conducted as a collaboration between the Information School, the Department of Computer Science and Engineering, and the Department of Rehabilitation Medicine at the University of Washington.
Broader Impacts: This work will lead to greater awareness of how software can be tailored to meet the needs of users with motor impairments, and to increased accessibility to information technology for this user community. The research will establish a new paradigm of accessible computing that may be transferable to other platforms, such as mobile touch screens, information kiosks, and continuous voice-based mouse control. Project outcomes will include deployed applications and programming tools that can be downloaded free of charge.
Overall, this project was concerned chiefly with making computers more usable and accessible for people with physical disabilities, especially disabilities that pose challenges to motor functioning, making mice and keyboards hard or impossible to use. Besides numerous scholarly findings and publications of value to scientists and engineers, this project has produced practical downloadable software that anyone can find on the Web and download for free for use on Windows systems. For example, our Angle Mouse software (available at http://depts.washington.edu/aimgroup/proj/angle/) makes the mouse cursor more accurate for people with poor fine motor control. Our Pointing Magnifier software (available at http://depts.washington.edu/aimgroup/proj/ptgmag/) magnifies targets on the computer screen, making them easier for people with motor impairments to click on. Both of these adaptations were reviewed by PC World and received 5 out of 5 ratings. (See www.pcworld.com/article/225289/improved_mouse_control_for_users_with_disabilities.html) Also, both adaptations were picked up by the ATHENA online inventory of Free Assistive Technology Software, hosted by the University of Athens. This inventory lists hundreds of software aids for improving accessibility. Only a handful are given high ratings, and both the Angle Mouse and Pointing Magnifier received 5 out of 5 stars. (See http://access.uoa.gr/ATHENA/eng/applications/all) We have received numerous emails from the public as to the helpfulness of our free assistive technology aids for improving mouse pointing on a computer. In addition, the work from this grant served to start a new course at the University of Washington in Seattle, WA. The course, called "Input & Interaction," has been offered now for four years, and focuses on creating technology solutions for people with disabilties to make computers and devices easier to operate and use. Other outcomes that will benefit the public are new algorithms and techniques for improving the accessibility of text entry on touch screens. Although this software is not downloadable, our work has shown how to make touch-screen keyboards more accurate, including while people are walking with small touch-screen devices. This technology will likely make its way into touch screens of the future to improve touch accuracy while typing on soft keyboards. We have released a project web site for any interested visitors at http://depts.washington.edu/agcproj/. There, onlookers can find project rationale, publications, personnel, videos, software, and related links.
