CHS: Small: User Interfaces for Improving Collaboration Between Blind and Sighted People

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

Inaccessible graphical user interfaces hinder participation by blind and visually impaired people in work, education, and other aspects of public life. Alternative user interface technology such as screen reader software can address this problem by converting inaccessible visual content into synthesized speech, but can introduce other issues when used in a shared setting; blind users are unable to follow along when their sighted partners use a graphical user interface, while most sighted users lack the training to use a screen reader. This gap makes it difficult for blind and sighted peers to collaborate on shared tasks, as each user may have a significantly different view of the task and may have difficulty following their partner's actions. The PI's goals in this research are to understand the barriers to collaboration between users of different visual abilities, and to create and evaluate new tools to facilitate such collaborations. Enabling blind and sighted people to work collaboratively will significantly improve educational and employment outcomes for millions of people. The PI will work with partners, the AccessComputing Alliance and the Colorado Center for the Blind, to directly involve individuals with disabilities as part of the research team, and will release an open source toolkit for creating software to support blind/sighted collaboration. Educational activities will include summer workshops with blind and sighted students to explore how students with mixed abilities can learn to co-design and co-develop more accessible software together. Little prior research has explored how blind and sighted people collaborate synchronously using computers. The PI's preliminary work has identified several significant obstacles to synchronous collaboration between blind and sighted individuals using existing tools: maintaining joint attention without visual feedback, lack of awareness of actions taken by others, and crosstalk between a screen reader's speech output and out-of-band conversations between activity partners. In this project, the PI will develop and test a set of metrics and reference tasks for evaluating collaborative activities between a sighted partner using a graphical user interface and a blind partner using a screen reader. The PI will develop tools to support equal and effective blind/sighted collaboration through sharing information about each user's activities with the group. This approach will connect users' computing devices (and related assistive technology) over a network, so that each user can maintain awareness of their partner's activities when interacting with shared resources. The PI will prototype various means of integrating this information into existing user interfaces, and implement a toolkit which enables software developers to support blind/sighted collaboration in their applications. The PI will validate these tools through user testing with mixed groups of blind and sighted users, and will demonstrate the robustness of the approach by applying it to synchronous co-located collaboration, asynchronous online collaboration, and micro-tasking using crowd workers.