Psychophysics of variable fonts: Do multiple font features interact to impact readability?

Poster Presentation: Saturday, May 17, 2025, 2:45 – 6:45 pm, Banyan Breezeway
Session: Object Recognition: Reading

Silvia Guidi1, Anna Kosovicheva1, Benjamin Wolfe1; 1University of Toronto Mississauga

When choosing a font, we have some intuitive understanding of why a particular font may feel easier to read, but what elements of a font actually affect readability? To answer this question, we used variable fonts, in which every element, such as the width or stroke contrast of each letter, can be adjusted on a continuous axis. Previously, we have shown that changes within a single axis can change saccade amplitude and reading duration thresholds (Guidi et al. VSS2024). In a new study, we examined how these axes impact readability in combination by manipulating text appearance on two axes, thin stroke and width, at three levels per axis across the full range, for a total of 9 conditions. Participants read a series of sentences in each font condition while gaze position was tracked, classifying each sentence as true or false. Sentence presentation duration was staircased and we calculated duration thresholds needed for 80% classification accuracy for each condition. Thicker thin strokes decreased duration thresholds across all width settings, while the thinnest thin strokes resulted in the highest duration thresholds (i.e., the slowest reading performance). These extreme thin strokes impacted reading speed regardless of the width of the text. Eye tracking data revealed that participants partially compensated for increased text width by increasing their saccade amplitudes. By understanding how different font elements interact with each other, we may be able to understand what parts of text presentation affect readability the most, which can then be used to help maximize reading efficiency.

Acknowledgements: This work was supported by a SSHRC Insight Grant to AK and BW.