Reading is a daily activity that many of us can do effortlessly.
The ability to read fluently depends on many factors including
early sensory influences, eye movement control, high-leve l cognitive
and linguistic factors. The main motivation of this study
stemmed from our interest in seeking a simple method to modify
text characteri stics that would be beneficial to people with visual
impairment . As such, we used a method to measure reading speed
that minimizes the requiremen t to make reading eye movements
(RSVP), and we also minimize the influences of cognitive and linguistic
factors by using a within-subject comparison, that is, each
observer was tested with all the condition s and our primary interest
was to determine how reading speed changes with the different
testing condition s (print size, testing eccentrici ty and strokewidth
boldness). We showed that by reducing or increasing the boldness
of letter-strokes (referred to as weight in typography ), reading
speed at the fovea is not affected by a wide range of boldness until
the letter strokes become very thin (0.27 the standard strokewidth)
or very thick (3.04 the standard stroke-wi dth). At 10
eccentricity in the periphery, reading speed is even more tolerant
of thin letter-str okes, as reading speed for the thinnest strokewidth
condition is not different from the reading speed for the
standard boldness. However, in the peripher y, reading speed falls
off from the optimal reading speed at a boldness of 1.89 the standard,
compared with 3.04 at the fovea, suggestin g that peripheral
reading is less tolerant to bold print. Considering the subjective
preferences for reading boldface print indicated by people with visual
impairment , our results confirm that subjective preferences
do not imply performance, and any evaluations of devices or developments
to improve visual function should consider performanc e
measureme nts instead of relying on subjective preferences per se.
Ac