Database Reference
In-Depth Information
Ending the Test
Sometimes there isn't time to complete the last task of the session, especially if it's long or the users
are having difficulty. In that case, I'll ask the Computer to demonstrate the remaining steps so that the
users can get closure on what they were doing. As a bonus, we may still identify another problem or
two if the users are surprised by something and comment on it.
I usually reserve the last 10 minutes of the test session for discussion (especially if observers are in the
room) of any interesting issues that came up during the test. Although I might do this in any type of
usability test, in paper prototype tests there is a greater likelihood that the observers will have questions
for the users. For example, this is an appropriate time to ask users if the tasks were typical of their work
or to follow up on their comments. Sometimes one of the designers has revised a troublesome screen
and wants to ask if it's better. With the tasks behind us, it doesn't matter as much if someone asks a
leading question, and it's okay to answer the users' questions directly. As a facilitator, I'll keep quiet if
the observers and users are having a fruitful discussion.
I do keep my eye on the clock, however. It's important to let the users go when we said we would, even
if the discussion is interesting. Usually I'll give a "2-minute warning" when we're running out of time,
which allows for another question or two. If the users and observers are engaged in discussion, when
we're out of time I will stand up as a cue that the session is over, and people get the hint. It's usually a
good idea to escort users back to the lobby so that you can be reasonably sure that they've left,
especially if you plan to hold a debriefing meeting immediately after the test. (It's an equally good idea
to only say things about users that you wouldn't mind them overhearing, just in case a user returns to
the test room to retrieve a forgotten umbrella.)
Search WWH ::
Custom Search