Dyslexia and web accessibility: synergies and challenges
Luz Rello · 2015 · Proceedings of the 12th International Web for All Conference (W4A) · doi:10.1145/2745555.2746655
Summary
This position paper reviews the key challenges of studying dyslexia in the context of web accessibility, drawing on Rello's extensive body of research including her doctoral work on the DysWebxia text accessibility model. The paper addresses three interconnected problems: measuring dyslexia's prevalence in the population, overcoming limitations in current research methodologies, and integrating dyslexia-specific guidance into web accessibility guidelines. Rello explains why measuring dyslexia's impact is so difficult — its manifestations vary across languages (only appearing in languages with writing systems), diagnosis criteria differ between countries, and it frequently co-occurs with conditions like dysgraphia, attention deficit disorder, and visual stress syndrome (Meares-Irlen syndrome). The paper challenges common assumptions about readability and comprehension for people with dyslexia, presenting the counterintuitive finding that in one experiment, text with a 16% word error rate actually produced higher comprehension in dyslexic readers than in non-dyslexic readers, even though both groups read more slowly. This demonstrates that readability and comprehension — often treated as interchangeable — function differently for people with dyslexia and require distinct approaches. Rello also questions the widely recommended solution of text customization, citing her finding that only half of participants (with and without dyslexia) could correctly identify which text settings led them to faster reading.
Key findings
The paper identifies a critical pattern across dyslexia research: text presentation conditions that benefit readers with dyslexia also benefit readers without dyslexia, but the effects are significantly larger for dyslexic readers. This was consistent across 12 experiments and 11 sub-experiments using eye tracking. Larger font sizes and sans serif typefaces (such as Arial) are the two most consistently supported recommendations across empirical studies. Rello proposes these as concrete additions to WCAG: font sizes of around 18 points for a 17-inch screen, and specific recommendations for sans serif, non-italic typefaces. The paper highlights that current lab-based studies with small participant groups produce inconsistent, non-comparable results — data from dyslexic participants never showed normal distributions, unlike control groups. Rello argues for large-scale web-based studies using crowdsourcing or school partnerships to overcome these limitations, which could enable statistical machine learning approaches and help cluster reading behaviors currently dismissed as individual differences.
Relevance
This paper is valuable for accessibility practitioners because it makes a pragmatic case for two simple, evidence-based design choices — larger font sizes and sans serif typefaces — that improve readability for everyone while disproportionately benefiting readers with dyslexia. The argument that dyslexic readers are "more sensitive" rather than fundamentally different from typical readers reframes cognitive accessibility as a universal design concern rather than a niche accommodation. The paper also raises important questions about WCAG's coverage of cognitive disabilities, noting that dyslexia's classification varies between a learning disability, a reading disability, and a disorder depending on the source — making it unclear where it fits within WCAG 2.0's framework. For web developers, the takeaway is straightforward: use relative font sizing with generous minimums and default to clean sans serif typefaces.
Tags: dyslexia · cognitive accessibility · readability · typography · web accessibility guidelines · text customization · font size · typeface
Standards referenced: WCAG 2.0