Civil Engineering Reference
In-Depth Information
Egan realized that the fastest and slowest users were not always representative of the
population, so he removed the top and bottom 25 % of the individuals from the
comparison. Then he compared the difference between the 75th percentile
TABLE 7.2
Results from Egan's Studies
Task
Max/Min
Q
3
/Q
1
Text editing
5:1
2:1
Information search
9:1
2:1
Programming
22:1
2:1
Three tasks were compared: text editing, information search, and programming. The ratio of
maximum to minimum perfor-mance and the ratio of the 75th to the 25th quartile performance are
given.
and 25th percentile. The performance ratio for all three tasks was then about 2:1, or 100%
increase in performance from 25th to 75th percentile user (see Table 7.2).
Egan went further to investigate exactly what cognitive factors could be responsible
for the differences in HCI performance. Several cognitive skills were tested using
standard testing procedures (Ekstrom et al, 1976). Typically, persons who performed well
in text editing scored higher on spatial memory than those who performed poorly. The
differences in information search and programming ability were explained by differences
in spatial memory as well as logical reasoning ability.
Spatial memory can be tested by asking participants to memorize a map of a city.
Participants first study the map for 5 minutes. Then the map is taken away, and they are
asked questions such as: Where is the church located? Where is McDonald's? and so
forth.
The significance of spatial memory is that it helps people to navigate. Users who can
remember how to navigate to a function on the computer will definitely save time. But
many of us are poor at memorizing the location of different functions. In MS Word, do
you know how to sort words according to the alphabet? Look under the pull-down menu
for Table.
Poor spatial memory can be a debilitating handicap. The question then emerges: What
can we do as designers of HCI to help people with low spatial memory perform better on
computers? One way would be to make all files and functions visible on the screen.
Thereby there is no need for navigation. This was indeed the idea behind direct
manipulation, the concept for the Xerox Star and the Apple Lisa. To avoid navigation
problems it is possible to design a direct manipulation interface, so that users do not have
to depend on their spatial memories.
Differences in cognitive performance are much greater than differences in manual
performance. Typically among workers on a factory assembly line there is a maximal
difference in manual assembly performance of about 25% (Konz and Johnson, 2004).
The cognitive differences are greater than 100% and are much more consequential.
People with low cognitive performance can, as we explained, be helped by a different
style of interface design.
