The effect of serifs and stroke contrast on low vision reading

Patients with low vision are generally recommended to use the same fonts as individuals with normal vision. However, we are yet to fully understand whether stroke width and serifs (small ornamentations at stroke endings) can increase readability. This study's purpose was to characterize the interaction between two factors (end-of-stroke and stroke width) in a well-defined and homogenous group of patients with low vision.

Font legibility was assessed by measuring word identification performance of 19 patients with low vision (autosomal dominant optic atrophy [ADOA] with a best-corrected average visual acuity 20/110) and a two-interval, forced-choice task was implemented. Word stimuli were presented with four different fonts designed to isolate the stylistic features of serif and stroke width.

Font-size threshold and sensitivity data revealed that using a single measure (i.e., font-size threshold) is insufficient for detecting significant effects but triangulation is possible when combined with signal detection theory. Specifically, low stroke contrast (smaller variation in stroke width) yielded significantly lower thresholds and higher sensitivity when a font contained serifs (331 points; d' = 1.47) relative to no serifs (345 points; d' = 1.15), E(μ_{sans, low} - μ_{serif, low}) = -14 points, 95 % Cr. I. = [-24, -5], P(δ > 0) = 0.99 and E(μ_{serif, low} - μ_{sans, low}) = 0.32, 95 % Cr. I. = [0.16, 0.49], P(δ > 0) = 0.99.

In people with low visual acuity caused by ADOA, the combination of serifs and a uniform stroke width resulted in better text legibility than other combinations of uniform/variable stroke widths and presence/absence of serifs.