Abstract
Individual differences in reading needs have mostly been addressed with accessibility features, specially-designed fonts, and dedicated reading tools. However, it is remarkable that given the prevalence of digital reading on personalized devices, text formatting is predominantly one-size-fits-all. Past work on text legibility and format readability has shown that certain adjustments to fonts, sizes, and spacing can benefit specific demographics (e.g., readers with dyslexia, low vision, etc.), and increasing evidence is emerging on the benefits of individualizing text formats to all readers, regardless of age, experience, and learning differences. The challenge remains how to help readers discover the formats that work best for them, without having to explore an intractably large space of text adjustments. Our approach is to bundle a set of text settings together into a small set of reading themes that readers can select from. We developed our reading themes using an iterative participatory design with crowdsourced participants. On each iteration, we hand a reduced set of text settings to participants, and have them adjust the default settings provided. Then, we invoke an automatic visual clustering algorithm on hundreds of resulting text formats to produce a small set of distinct reading experiences. This approach converges after four iterations with a final set of three reading themes - which we term Compact, Open, and Relaxed - corresponding to increasing character, word, and line spacing, and distinct font choices. Our reading themes are a result of pilot studies with 271 participants and 4 designers, and four iterations of our design process with 485 participants, sampled uniformly across age bins ranging from 18 - 87 years, 59% female, and 49% readers with dyslexia, to represent diverse reading needs. We further demonstrate the benefits of reading themes on reading comfort, speed, and comprehension - to show that reading experiences should not be one-size-fits-all.