|
A study of fonts designed for screen display,
Boyarski, D., Neuwirth, C., Forlizzi, J., and Regli, S.H., CHI 98 Conference
Proceedings, 87-94 (1998).
|
It is very difficult to improve reading performance. Historically, one
way to improve performance has been to create newer, clearer fonts. Most
fonts being read on computer monitors were designed to be read from paper.
This study evaluated two fonts that were specifically designed for use
on computer monitors. The new fonts were Georgia and Verdana.
Examples
- This is an example
of 10 point Times New Roman font
- This is an example of
10 point Verdana font
- This
is an example of 10 point Verdana Italics font
After designing the fonts, they conducted a study to examine people's
speed of reading with the new fonts. Participants in the studies were
people from Carnegie Mellon University, including faculty, staff and graduate
students who ranged in age from 20 to 53.
All text was set at 10 points with 13 points leading and an average line
length of 10 words. Subjects read the text on a 17 inch screen with a
resolution of 640 x 480 pixels.
They compared the following:
- Georgia (serif) vs. Times Roman (serif)
- Georgia (serif) vs. Verdana (sans serif)
- Verdana vs. Verdana Italic
They found no reliable performance differences in reading speed.
Next they evaluated aliased vs. anti-aliased fonts. Anti-aliased fonts
use levels of color in selected pixels to eliminate the appearance of
jagged edges (the "jaggies"). In these studies, all text was
displayed at 16 points (a character must be at least 14 points for the
anti-aliasing to show).
They compared the following fonts:
- Adobe anti-aliased vs. Microsoft bitmap
- Microsoft anti-aliased vs. Microsoft bitmap
- Adobe anti-aliased vs. Microsoft anti-aliased
Again, they found no performance differences in reading speed.
The screen fonts that have been used successfully for the past few years
seem to work as well as newly designed fonts. There seems to be no reliable
differences between serif and sans-serif fonts, or between aliased and
anti-aliased fonts. If reading speed is to be improved, it appears that
it will not come from making changes to the fonts.
Improvement in reading speed when using computers will require more drastic
alternatives.
|
|
Reading paradigms: From lab to cyberspace?
Aaronsn, D. and Colet, E., Behavior Research Methods, Instruments and
Computers, 29 (2), 250-255 (1997).
|
Aaronson and Colet (1997) have proposed several ways to improve reading
speed. They described some methods that have been developed in computer
laboratories over the past few years to help in controlling text displays
for reading research.
One method had subjects pace themselves through word sequences. Each
time they pressed a key on the keyboard a new word would show in the center
of the screen (overwriting the existing word). In this way, the subjects
could pace themselves through word sequences. Reading speed was improved
over the speed of reading from a document because virtually all eye movements
(saccades) were eliminated.
A closely related, but alternative method, was to allow subjects to have
each key press show a new word, but to have some or all preceding words
remain on the screen. In this way a sentence or paragraph gradually appears,
which allowed users to glance back at previously read material.
One of the most promising methods is called Rapid Serial Visual Presentation
(RSVP). It was first used in the mid-1960s for rapidly displaying individual
words, one at a time, in the center of a monitor's screen. Each new word
replaced the old word. Users set the rate that the computer presented
the words. This approach has demonstrated a capacity to substantially
improve reading speed.
In the 1999 User Interface Update course (see below), the RSVP method
is used to illustrate how quickly reading performance can be improved.
In one class the average reading speed from a paper document was measured
to be 342 words-per-minute, with a range from 143 to 540 words-per-minute.
After determining the basic reading rate, the class members read material
presented on the screen, one word at a time, at 600, then 800, then 1,000,
then 1,300, and finally at 1,600 words-per-minute. After each set of reading
material, the students answered multiple choice questions about the text.
The top reading speeds were as follows for measured comprehension scores
of 75% or higher:
- 1,600 wpm – 3 people
- 1,300 wpm – 8 people
- 1,000 wpm – 3 people
- 800 wpm – 0 (nobody)
- 600 wpm – 2 people
The average for the class was 1212 wpm, which is about 3.5 times faster
than reading in the traditional way. There is no question that the computer
can help improve reading performance; but it must be done in non-traditional
ways.
|
