Cognition: Memory, Attention, and Learning - Foundations for Designing User-Centered Systems - page 156

of programming, programming constructs, language syntax, programming
interface, online tutorial interface.
What are the results, and what do they suggest for the design of online
tutorials?
5.3 With respect to a web site for a university department (or your company's),
create a list of tasks and information that a user of a university department's
web site would be able to retrieve from memory and a similar list that users
might not initially be able to recall, but would recognize. An example of the
first could include the main phone number, and an example of the second
could include special resources or accomplishments of the department.
5.4 Recall of memories is more difficult than recognition. As an example of this,
write down the 50 US states. You can adapt this task to your own country, for
example, listing the UK counties, the Bundesländer of Germany, or the
number of member states in the European Union.
This effect also works for computer interfaces. You could attempt to name
the menu bar items for a menu driven interface that you use often.
5.5 Usability of security remains an ongoing active research area. Memory
influences the use of computer security. There are several ways that this can
happen. One way is in the choice and memorizing strategies for passwords.
Read https://www.cs.cmu.edu/*help/security/choosing_passwords.html or
find a similar study on how people choose passwords. If you have time, write
down your previous passwords, or ask people you know to fill out a small
survey on what type of password they have, or have had. What do the results
say about human memory and about computer interface design for passwords?
5.6 The first step of this exercise is to prepare a task for users to perform. This task
should take people about 2 min on their first attempt, and they should be
successful nearly all of the time. Students have used such tasks as card sorting,
shoe tying, word processing, paper airplane construction, typing a paragraph,
and web navigation. You should prepare instructions, as you will be having
someone else perform the task. If the task is longer than 2 min it gets hard to
run the necessary number of trials. If it is less than 2 min it may get hard to
time as the users get faster. Anywhere from 70 s to 210 s is likely to work
easily, and greater or lesser amounts will work to a certain extent. Online tasks
may be more appropriate for your course, but what's interesting is that it does
not matter what task you choose or whether it is online or not.
To get a good look at the learning curve, you should run 2-5 users on your
task, and the subjects should each perform the task at least 15 times. Because of
the logarithmic scale, 100 times is twice as good as 10, a 1000 times is three
times as good. You may choose to run 5 subjects on the task 15 times (at most, 5
subjects 9 2min9 15 times = 150 min, so you may want to have help), or
you might choose to run 2 users over a longer series of trials.
You should record the time taken to perform the task each time, and note
whether the user made any minor errors, significant errors, or catastrophic