Data Science and Visualization
Crowdsourced Data: Accuracy, Accessibility, Authority (CDAAA) is a 3-year Early Career Research project led by Assistant Professor Victoria Van Hyning to investigate the sociotechnical barriers that Libraries, Archives and Museums (LAMs) face in making their crowdsourced transcription data derived from cultural heritage materials open and accessible to people who use screen reader software. Transcriptions have been collected by hundreds of LAMs with the goal of increasinging the accessibility of their materials, but to date there have been few studies to determine if the resulting transcriptions are accessible and usable. Through a survey and interviews with LAM and community-based practitioners who conduct crowdsourcing projects, and usability testing of LAM discovery systems with people who use screen readers, we will gather evidence to identify the barriers LAMs face to crowdsourced data integration, as well as concrete examples of successes they can adapt for their own organizations and workflows. Volunteers who dedicate time and effort to transcribe materials deserve for their work to be meaningful and sustainable. People who are Blind, have low-vision or dyslexia, or use a screen reader for another reason are currently some of the most excluded members of society when it comes to accessing cultural heritage through digital LAMs, and they deserve better access.
PI: Victoria Van Hyning
Making Data Visualization Accessible for Blind Users
Advanced educational opportunities and high-paying employment opportunities require the ability to analyze large datasets, yet data visualization approaches remain inaccessible for blind users. We are utilizing user-centered and participatory design methods (including interviews of blind orientation and mobility instructors) to understand how large data sets can be made accessible for blind users through sonification methods, and we are developing an accessible web-based tool for sonification of large-scale datasets.
PIs: Niklas Elmqvist and Jonathan Lazar
We will gather, document, and promote the appropriate use of disability data in both disability-related and non-disability-related data-driven innovation and technologies to ensure that they are usable by and applicable to people with disabilities. Goals of this project include: to increase the availability to the community of data that is collected from populations for the purpose of accessibility research and engineering, to highlight the potential limitations of these data and how they differ from mass data, to promote the effective and ethical use of accessibility data, to accelerate data-driven technological innovations for universally accessible information, and to raise awareness about machine learning bias for people with disabilities.
PI: Hernisa Kacorri
Understanding Users & their Needs
Our recent study on personal health technologies reveals the accessibility challenges faced by the blind and low-vision (BLV) people when using these tools. In a survey of 156 BLV individuals, we explored their motivations, practices, and barriers in collecting and using personal health data. Our findings show that despite facing obstacles in all phases of self-tracking, many BLV respondents actively engage in tracking health data, such as exercise, weight, sleep, and food. This research underscores the importance of addressing accessibility challenges and fostering inclusive design opportunities to ensure equitable access to health tracking technologies for all, including the BLV community.
PI: Eun Kyoung Choe
While established accessibility principles help technologists create interfaces that can be used by people with many different kinds of disabilities, our understanding of accessible technology design for people with dementia is still in its infancy. We are conducting interviews, observations, and co-design workshops with people at different stages of dementia as well as practitioners who work closely with people with dementia to contribute to knowledge in this area. Watch a presentation about one study, which identifies ways that sensory changes affect technology use.
PI: Amanda Lazar
Researchers can greatly benefit from conducting user-centered design with people with dementia, but may face issues such as recruitment and ensuring research activities are accessible and non-exploitative. Through a range of projects, including intergenerational hackathons, remote research methods, and participatory action research, we are developing new ways to design with people with dementia.
PI: Amanda Lazar
Digital technologies have the potential to help people with Down syndrome in education, employment, career development, and independent living. In the past 15 years, our team has conducted a series of studies that investigated both the general computer usage of people with DS and specific tasks or applications such as workplace-related tasks, input techniques, and authentication. Our more recent recent projects focus on improving healthcare services and health data accessibility for people with DS.
Project team: Jonathan Lazar, Jinjuan Heidi Feng, Rachel Wood (Doctoral student); Previously included: Libby Kumin, Ant Ozok, Brian Skotko, and Brian Wentz.
Autistic people experience sensory and social information differently from neurotypical people. This leads to frequent miscommunications and discrimination in majority-neurotypical workplaces. The FIT project aims to understand the precise causes and consequences of autistic-neurotypical misalignments in workplace conversations. Our long-term goal is to build a video-calling platform that promotes mutual understanding between autistic and non-autistic people.
PI: Yi Ting Huang
Tools for Users
PDF documents are often considered to be hard to make accessible (much harder than web content or ebooks), and content creators often report that is due to a lack of tools to assist with making PDF files accessible. Our current work, in collaboration with Adobe Research, focuses on creating new software tools to assist content creators in evaluating and remediating their PDF documents for accessibility, specifically in areas such as reading order, table markup, and headings.
PI: Jonathan Lazar
Exploring “teachable interfaces” as improved approach for intelligent self-adapting interfaces that also puts interfaces back under user control
We will test (through an in-depth example) the potential of “teachable interfaces” for more effective and accessible human-in-the loop personalization of innovative technologies that use machine learning. We will demonstrate how the concept of teachable interfaces can be effectively incorporated into a real-world assistive technology, explore challenges that users with disabilities may have in conceptualizing teachability, analyze user strategies for incorporating variation in their training examples and how they relate to system performance, and understand what design parameters, sensing modalities, and interactions are most influential for both system accuracy and user experience.
PI: Hernisa Kacorri
Toucha11y: A prototype that allows Blind users to use existing inaccessible touchscreen kiosks
Despite their growing popularity, many public kiosks with touchscreens are inaccessible to blind people. In this research, we developed Toucha11y, a working prototype that allows Blind users to use existing inaccessible touchscreen kiosks independently and with little effort. Toucha11y consists of a mechanical bot that can be instrumented to an arbitrary touchscreen kiosk by a Blind user and a companion app on their smartphone. The bot, once attached to a touchscreen, will recognize its content, retrieve the corresponding information from a database, and render it on the user’s smartphone. As a result, a Blind person can use the smartphone’s built-in accessibility features to access content and make selections. The mechanical bot will detect and activate the corresponding touchscreen interface. We present the system design of Toucha11y along with a series of technical evaluations. Through a user study, we found out that Toucha11y could help blind users operate inaccessible touchscreen devices. A demo video of the system and a more detailed project description is available.
PI: Huaishu Peng
Technical Standards & Tools for Developers
Kiosks come in many different forms, including: public kiosks with unlimited use and no personal information; kiosks which require personal information but allow for unlimited use (e.g. ATMs); and kiosks which require personal information and allow limited use (voting machines). The Trace Center is currently working on creating a set of unified guidelines for kiosk accessibility, bringing together requirements and guidelines from different countries and usage domains. These unified guidelines will have a combination of hardware strategies that must be applied to make kiosks accessible, while also having flexibility depending on the type of kiosk and the context of use. Read more about the Kiosk Accessibility Guidelines in an Interactions article. And watch a webinar on kiosk accessibility.
PIs: Gregg Vanderheiden, J. Bern Jordan, and Jonathan Lazar