WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: Browser font sizes

for

From: Terence de Giere
Date: Mar 8, 2002 9:35PM

To Nick Wilson --

My comments on font sizes were general, and you are right that it is not
hard data. At Human Factors International, the statement "Most users do
not change browser defaults" appears in their Web course materials. I do
not know what data they used to make that statement, but it would be
true if more than 50% of users did not change defaults.

My own personal observations, from looking at settings and just plain
looking at computer monitors in various offices I have worked in
indicates that "Most users do not change browser defaults". A personal
observation is a kind of evidence but it is not precisely qualified or
quantified because it was not tablulated and there was no experimental
design set up in advance. But one does not need to do a carefully
controlled study to get a ballpark estimate. Certainly less than 50% of
users change the defaults. For my own observations I think the figure is
something around 4-5% or less users change the browser font settings.
The same is about true for resolution settings in Windows, although I
have a suspicion the number of users using "large fonts" for the Windows
OS settings may be more than 5%, but a loose observation certainly can't
pin that figure down. A few hours in a large office specifically looking
at computer settings might provide a better survey. It would not take a
peer reviewed paper to nail down a rough figure for practical use.

Human Factors International recommends that the on screen size of font
be approximately 1/8 inch high to be readable for a typical user. I do
not know the data source for this recommendation. Microsoft recommends a
font size of 16 points for more elderly users in their accessibility
guidelines. When a point size is specified, it is actually difficult to
tell how big it will appear on screen because the pixels per inch
setting of the system may not correspond to the physical size of the
monitor display. A monitor setting of 640x480 pixels can be used on a 12
inch monitor or a 21 inch monitor, so the potential physical font size
displayed may vary by almost a factor of two. That's an extreme. But I
have observed two users using 640x480 on 17 inch monitors, which results
in fonts displaying somewhat larger than a precise WYSIWYG setting. One
did not know there were other settings. The other tried them, and found
the higher resolution settings produced an annoying flicker, and 640x480
was the least annoying. This user apparently did not know there were
also advanced refresh rate settings that might have solved the perceived
flicker problem.

I set my computers to Windows at 96dpi, but the physical display results
in a resolution of over 110 dpi, resulting in fonts appearing smaller
than normal. Most people who look at my screen say they can't read it. I
got into this habit with my first PC where 600x800 was an ideal setting
but produced terrible flicker no matter what settings I used, but
1024x768 was fine. I got used to the small fonts etc.

The accessibility principle is not how many people make a change, but if
a user needs a larger font, can they change it? Newer browsers allow
easier overrides to the designer's sizing, but some of the older
browsers prevent this, or if allowed, force the user to make changes to
settings buried in the menus most people don't seem to know are there.
From the designer's perspective, the question is what happens to the
page design, and the usability of the design if the user changes the
font size. It is obviously most practical to aim for the most common
settings.

The 80/20 rule, derived from the economist Pareto, if applied to the Web
would mean that we get the biggest bang for the buck and effort if we
designed Web pages for Internet Explorer on Windows for normal users
using default Internet Explorer settings. This seems to be what we have
done. But this leaves out most disabled populations using the Web.

Now in my family there are two users. One changes font sizes all the
time and the other does not. Exactly half the users change the font
size. Unfortunately this is not statistically significant to an
acceptable degree. If anyone has knowledge of a study that has some hard
data on this subject of users changing font sizes we would all
appreciate knowing about it.

Regards,

Terence de Giere
<EMAIL REMOVED>



----
To subscribe, unsubscribe, or view list archives,
visit http://www.webaim.org/discussion/