Case Studies - Paper Prototyping: The Fast and Easy Way to Define and Refine User Interfaces - page 28

spent time fiddling with the customization controls. The 3D navigation proved cumbersome, with users
voicing objections such as, "Why do I have to walk my guy into the classroom? Why can't I just take the
class?" It also wasn't clear what viewpoint the avatar should show. Should the interface appear as
though the user was looking through the avatar's eyes, or should the user have an outside perspective
where they could see their own avatar along with everyone else's? How would the classroom " seating "
be handled? (With a spatial layout, students in the back of the classroom wouldn't be visible to those in
front, and students in the front might block others' view.) None of these questions had easy answers,
and they all distracted users from the learning experience.
The bottom line was that the cutting-edge 3D functionality got in the way of the real purpose of the
application—enabling large companies to provide training to geographically dispersed employees. The
designers realized that they were jeopardizing their entire business model with an interface that felt too
much like a video game.
Based on the feedback, Centra dropped the 3D functionality, which shortened their development
schedule. Ronnie Thomson, Director of Engineering, explains, "Time to market was our major concern,
and the 3D functionality was our biggest challenge from a technical perspective. Our original plan was
to implement it in C++. When we dropped the 3D, we were able to use Java instead, which was a
simpler development environment. The end result was that we released rev 1.0 sooner than if we'd
stayed with the 3D and C++."
Social Considerations
The paper prototype tests were conducted with everyone—users, facilitator, Computer, and
observers—sitting in the same room, so obviously the users realized that people in the test setting
could hear them. But the users were able to extrapolate the test setting to real life and identify some of
the social concerns that they would have if they were to use this interface alone at their desks.
Because of bandwidth constraints, it wasn't possible to support a full-audio environment. The instructor
had a microphone and could "pass" a second microphone to one student at a time. In testing the
prototype, we found that the test participants wanted a clear indication of when their microphone was
on. (In a similar paper prototype test of a video-conferencing system, one of my colleagues found a
situation in which a test participant mistakenly believed that the person on the other end of the pretend
system couldn't hear him. He made a joking remark that would have been embarrassing had it
happened in real life.) In both of these cases, the designers modified the interface to make the
microphone indication more salient.
I learned the live microphone lesson the hard way. During a break from teaching a large class, I
went to the ladies' room without turning off my cordless mic! I really wish someone would have
found this problem with a paper prototype and implemented a warning buzzer that goes off once
the mic moves more than a certain distance from its receiver.
When the 3D functionality was removed, the designers realized they still had to give students a way to
see who else was in the class. They added what they called the "people panel"—a listing of the class
by name, including visual indication of which person has the microphone ( Figure 2.4 ).