Reading Experiences and Interest in Reading-Assistance Tools Among Deaf and Hard-of-Hearing Computing Professionals
Oliver Alonzo, Lisa Elliot, Becca Dingman, Matt Huenerfauth · 2020 · Proceedings of the 22nd International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2020) · doi:10.1145/3373625.3416992
Summary
This study investigates whether Deaf and Hard-of-Hearing (DHH) computing professionals are interested in Automatic Text Simplification (ATS) tools to assist with reading in their professional work. ATS software rewrites text to improve readability using lexical simplification (replacing complex words), syntactic simplification (restructuring sentences), or both. While prior research has explored ATS for various groups including people with aphasia, dyslexia, and low literacy, very little work had examined interest among DHH users specifically, or within a professional computing context. The researchers used a mixed-method approach: pilot interviews with 12 DHH participants informed an online survey (N=32), followed by in-depth interviews with 5 survey respondents. The survey covered reading habits, experiences with complicated text, interest in ATS tools, and design considerations around autonomy and social acceptability. Participants were DHH individuals with experience in computing or information technology, recruited through social media, alumni networks, and tech company accessibility groups. The study was motivated by the recognition that computing professionals must continuously read technical documentation to stay current, while prior research has found median reading levels among DHH high school graduates at approximately fourth-grade level, and DHH workers are significantly underrepresented in computing (only 0.8% of Stack Overflow users).
Key findings
The majority of participants read frequently — at least once a week — with most reading occurring on electronic screens for work and academic purposes. There was a statistically significant difference between time spent reading on-screen versus on paper (p < 0.01), with 25 respondents spending over 30 minutes daily reading on screens. When encountering complicated text, the most common workarounds were looking up words in a dictionary (N=25) and searching for alternative websites on the same topic (N=21) — strategies that closely parallel what lexical and syntactic ATS approaches provide. Interest in ATS tools was strongly positive: the median response to interest was "strongly agree," with 25 of 32 participants responding at least "somewhat agree." Interest was highest for academic, medical, legal, and work-related reading. A significant Spearman correlation (rho=0.5, p=0.0034) linked sign language preference with greater tool interest. Interview participants emphasized time savings and independence as key benefits, with one stating the tool would "speed up my reading pace that is all-important." However, participants raised important design concerns: most would be upset if text were replaced before seeing the original or without their consent, highlighting the need for user autonomy. Social acceptability was mixed — some worried about embarrassment if colleagues noticed them using such a tool, while others prioritized the practical benefit.
Relevance
This paper provides essential evidence that DHH computing professionals — a specific, underserved user group — have genuine interest in reading assistance technology, filling a gap in user needs research for ATS tools. For accessibility practitioners, the findings underscore that reading accessibility extends beyond visual impairments to include the DHH community, where English is often a second language after ASL. The design implications are directly actionable: ATS tools must preserve user autonomy (showing original text, requiring consent before changes), address social acceptability concerns (unobtrusive integration), and prioritize accuracy to avoid misinterpretation. The study also reveals that existing workarounds DHH professionals use (dictionary lookups, finding simpler alternatives) already mirror what ATS technology automates, suggesting strong product-market fit. A limitation is the small sample size (N=32 survey, N=5 interviews), and the authors acknowledge that feedback may be biased toward more tech-savvy DHH individuals.
Tags: deaf and hard of hearing · automatic text simplification · reading accessibility · natural language processing · assistive technology · workplace accessibility · plain language