Purchase this article with an account.
Jean-Baptiste Bernard, Carlos Aguilar, Eric Castet; A new font to reduce crowding. Journal of Vision 2014;14(10):784. doi: https://doi.org/10.1167/14.10.784.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Crowding refers to the deleterious effect of surrounding items on object recognition. It affects the recognition of groups of letters and is presumably an important factor limiting reading performance. In this study, we investigated the possibility of designing a new font to reduce the effect of crowding on letter recognition. The new font (named the ABC font) was designed to decrease the physical similarities of letters frequently confused during crowding. The 26 ABC letters derive from lowercase and uppercase symbols of classical fonts. They are low-complexity letters with special attributes including oblique or curved strokes that increase their specificity. We measured letter recognition performance for the ABC and Courier fonts in four subjects. The task consisted of the recognition of three crowded letters of the same font, horizontally aligned and briefly presented at 6° eccentricity in the lower visual field (1,000 trigrams per font and per subject, duration: 100-200 ms depending on subjects' performance). Center-to-center spacing was the same for both fonts (Courier standard spacing) as well as the letter stroke-width and x-height which was set to the critical print size (CPS, range: 1.09°-1.18°) for each subject (values larger than the CPS do not increase reading speed). For each subject, crowded letter recognition performance was much higher for ABC than Courier font: Errors decreased by 43.91% on average (range: 37-50%), from 0.78 to 0.43 letter errors per trigram. Frequently confused pairs (>5% confusion errors) were reduced from 25 to 9 pairs on average with ABC compared to Courier. Letter mislocation rates were similar for both fonts (0.16 mislocation errors per trial for Courier and 0.13 mislocation errors for ABC on average). These results suggest that crowded letter recognition performance can be increased by reducing physical similarities between letters.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only