An Empirical Investigation of Ways in Which Some of the Problems Encountered by Some Dyslexics May Be Alleviated Using Computer Techniques
Peter Gregor, Alan F. Newell · 2000 · Proceedings of the Fourth International ACM Conference on Assistive Technologies (Assets '00) · doi:10.1145/354324.354347
Summary
This paper describes the development and user evaluation of a highly configurable word processing environment designed to help dyslexic users read and produce text more effectively. The researchers at the University of Dundee identified nine common problems dyslexics encounter: number and letter recognition issues, letter reversals, word recognition difficulties, number/letter/word recollection problems, spelling and punctuation errors, fixation problems, word additions and omissions, and poor comprehension. Rather than pursuing a diagnostic approach, they adopted a pragmatic, user-centred methodology — working iteratively with dyslexic users to develop software that lets each individual customize their reading and writing environment to suit their specific needs. The system was built as a Microsoft Word add-on featuring two components: a "Preferences" program allowing users to experiment with and save display settings (colours, fonts, spacing, text size), and a "Reading/Editing" program that applies those preferences. Two rounds of user evaluation were conducted with dyslexic students aged 15-30, using qualitative methods including questionnaires, think-aloud techniques, and interviews to gather feedback throughout the iterative development process.
Key findings
All dyslexic subjects tested were able to find personalised display configurations they found easier to read than default settings, though individual preferences varied enormously — highlighting that no single "dyslexia-friendly" setting exists. The most surprising finding was the preference for low colour contrast between text and background, such as brown text on muddy green, contradicting the assumption that high contrast aids readability. Character reversals were addressed through multiple strategies: using different fonts, colours, or sizes for commonly confused characters (b/d, p/q), and providing fixation aids within the text. The column width option worked well for those with fixation problems, and the reversals feature showed promise both for avoiding reversals and helping fixation. Users consistently discovered useful features they had not previously known about, such as text colour options already available in standard word processors but not presented in a way that encouraged experimentation. A text-to-speech option was also included for additional reading support. The second prototype evaluation with seven dyslexic users confirmed that spacing adjustments (character, word, and line spacing) and font size increases were widely beneficial.
Relevance
This research makes a compelling case that screen customization to personal preferences can meaningfully improve reading ability for dyslexic users — and that the diversity of individual needs means one-size-fits-all solutions are inadequate. The findings have broad implications beyond dyslexia: the authors note that anyone prone to reading difficulties (elderly users, people with mild visual impairments, or those experiencing screen fatigue) could benefit from similar configurability. From a design perspective, the paper demonstrates the value of making existing accessibility features more discoverable and easier to experiment with, rather than always building entirely new tools. The user-centred, iterative methodology — deliberately choosing qualitative feedback over formal experiments in early stages — offers a practical model for accessibility research where the user population is highly heterogeneous. The work foreshadowed modern browser and OS reading mode features, and its core insight that personalisation trumps standardisation remains highly relevant to digital accessibility practice today.
Tags: dyslexia · reading accessibility · text customization · user-centred design · word processing · cognitive accessibility · personalization