It is important that the users, tasks and environment used for the test
are representative of the intended context of use.
Select the most important tasks and user groups to be tested (e.g. the
most frequent or the most critical).
Select users who are representative of each user group. 3-5 users are
sufficient to identify problems. 8 or more users of each type are required
for reliable measures.
Produce a task scenario and input data and write instructions for the
user (tell the user what to achieve, not how to do it).
Plan sessions allowing time for giving instructions, running the test,
answering a questionnaire, and a post-test interview.
Invite developers to observe the sessions if possible. An alternative is
to videotape the sessions, and show developers edited clips of the usability
problems.
Two administrators are normally required to share the activities of
instructing and interviewing the user, operating video equipment (if used),
noting problems, and speaking to any observers.
If possible use one room for testing, linked by video to another room
for observation.
If usability measures are required, observe the user without making any
comments.
If measures are not required, prompt the user to explain their
interpretation of the contents of each screen and their reason for making
choices.
