This journal paper presents a large-scale crowdsourced reading study that compares the readability of 16 different fonts. Readers were able to read 35% faster in their best fonts compared to their worst fonts, and which font was best depended on the individual reader, pointing to the need to customize font choices for different readers.

This paper presents FontMART, a font recommender capable of making personalized recommendations for faster fonts to read with. In a study conducted with 252 crowdworkers, we found that the font recommendation from FontMART helped readers speed up by 14-25 words per minute, without hindering comprehension. This encouraging evidence provides motivation for adding our personalized font recommendation to future user interfaces.

This abstract, presented as a poster at the Vision Sciences Society Annual Meeting, demonstrated the reading benefits of using a digital reading ruler, for dyslexic readers and the general population more broadly. Three different digital reading rulers were investigated, showing that different readers benefit from different reading ruler designs.

This long-form journal paper is both a survey and a guide to running readability studies. We discuss the components that come together to form reading experiments, including the typographic considerations, reading populations, experimental hardware (including eye trackers and brain imagining tech), metrics (speed, comprehension, preference, etc.), and analysis techniques.  Authors of this paper include typographers, educators, vision scientists, neuroscientists, and computer scientists.