## A3. Cognitive Ergonomics — Visual Crowding and Typography **Core finding:** Spacing is the active ingredient in typographic accessibility — not specialised letterforms. OpenDyslexic does not outperform standard sans-serif fonts. Individual variation is enormous; personalisation matters more than any single font choice. **Spacing evidence:** - **Zorzi et al. 2012** (*Proceedings of the National Academy of Sciences*): 74 Italian and 20 French dyslexic children. Extra-large letter spacing (increased ~2.5pt) **doubled reading accuracy and increased reading speed by over 20%** in dyslexic children, with no effect on controls. Mechanism: reduced visual crowding. - **Galliussi et al. 2020** (*Annals of Dyslexia*): Critical nuance — **increasing letter spacing without proportionally increasing word spacing actually DECREASES reading speed** because word boundaries become ambiguous. Letter and word spacing must be coordinated. - **Joo et al. 2018** (*Cortex*): Measured individual visual crowding profiles. Only a **subgroup with elevated crowding** benefited from increased spacing — others did not. This confirms personalisation is essential. **Font evidence (against specialised "dyslexia fonts"):** - **Rello & Baeza-Yates 2016** (*ACM TACCESS*): Most comprehensive eye-tracking study — **97 participants (48 with dyslexia), 12 fonts**. OpenDyslexic did **not** outperform standard sans-serif fonts like Arial, Helvetica, or Verdana. Sans-serif, monospaced, and roman (upright) fonts significantly outperformed serif, proportional, and italic alternatives. **Italic text significantly impaired reading.** - **Kuster et al. 2018** (*Annals of Dyslexia*): 170 children with dyslexia read no faster or more accurately in Dyslexie font than in Arial. Majority preferred Arial. - **Wery & Diliberto 2017** (*Annals of Dyslexia*): Confirmed no improvement with OpenDyslexic across multiple reading tasks. - **Wallace et al. 2022** (*ACM Transactions on CHI*): 16 fonts across hundreds of participants. Potential speed gains of **up to 35%** when comparing an individual's fastest vs. slowest font. No single font optimal for everyone. Font preference did not predict reading speed. **ADHD-specific:** - **Stern & Shalev 2013** (*Research in Developmental Disabilities*): ADHD adolescents showed differential benefits from spacing and screen presentation. All participants performed better on computer than paper. - **Cooreman & Beier 2024** (*SSSR Conference*): Larger x-height fractions increase processing speed at the perceptual level — particularly relevant for ADHD users with reduced processing speed. **Colour contrast:** - **Rello 2012** (*W3C Symposium*): People with dyslexia read fastest with lower-contrast warm pairs like **black on crème** — not black on white. Only 13.64% of dyslexic readers preferred black-on-white vs. 32.67% of controls. **Implication for Kon:** Default to a clean sans-serif with large x-height (Atkinson Hyperlegible or Lexend) with coordinated letter, word, and line spacing controls. Offer warm off-white background options (crème, not white). Never use italic for extended reading. OpenDyslexic should be available as an option but not recommended — spacing is the intervention, not letterform. Most importantly: allow full typographic personalisation, because no single configuration is optimal for all neurodivergent users.