Cognition: Human–Computer Communication - Foundations for Designing User-Centered Systems - page 201

related to other knowledge structures. By considering the interaction between the
human and the computer as a sort of conversation, you can use Grice's maxims to
help optimize the interaction.
7.3 How Users Read
One of the most common activities users do is reading. They read pre-printed
matter such as manuals and the documents that they print out. Increasingly
material is read on devices screens like monitors, laptops, tablets and smartphones.
Understanding how users read will help you design interfaces that are better suited
to the way people read and work.
We know enough about how people read (e.g., Just and Carpenter 1987 ; Rayner
and Pollatsek 1989 ) to be able to develop computational models of the process (Just
and Thibadeau 1984 ). The research in this area has identified several regularities of
reading behavior that are useful for designing interfaces where written text is used.
For example, we know that people typically read at a rate of about 150-230 words
per minute when reading for comprehension, and that this rate varies with material
and task. At this rate, it should take 2-3 min to read an A4 or 8.5 00 9 11 00 page of text.
Perhaps the three most important regularities of reading behavior, though, are the
word length effect, the word frequency effect, and the difference in time between
picture naming and reading. The word length effect arises out of the fact that longer
words simply take longer to read (e.g., Rayner et al. 2011 ). Words like ''mouse'' and
''polo'' are faster to read (everything else being equal) than ''keyboard'' and
''autocratic'' because they are shorter.
The word frequency effect arises because words that are commonly used are
faster to read than words that are less commonly used (e.g., Liversedge et al. 2011 ;
Rayner 1998 ). This effect can be seen as being related to the learning curve. The
word frequency effect and the word length effect are independent, so they both
apply to the same word. So, for example, while the word ''polo'' is short, it is also
relatively rarely used and hence will take longer to read than other more frequently
used words of the same length. It will, however, be read more quickly than equally
rare words that are longer.
Reading takes less time than picture naming (Chilton 1996 ). This explains why
unfamiliar icons, which are accessed through picture naming, take longer to
process than words that are used in the same place. Using icons instead of words
for simple searching makes things slower. There is a trade-off, however, against
generality and esthetics. An icon, once learned, is easier to recognize than a word
irrespective of one's native language and literacy level. Icons are often created to
be rectangular in shape, for example, and having consistent, regular shaped objects
can make for a better, clearer design.
There are other effects as well that influence reading speed. Several known
factors that make reading more difficult are noted in Table 7.2 , and some of these
effects are illustrated in Tables 7.3 and 7.4 .