The Politics of Paper Prototyping - Paper Prototyping: The Fast and Easy Way to Define and Refine User Interfaces - page 220

the selected features.
"Because the end users were internal, a number of them were involved in the design process but
the manufacturer hadn't budgeted any time to conduct usability tests. Fortunately, I teach
usability courses at a local university, so I had my class conduct a usability test on a software
prototype. Only one user participated, but he uncovered a handful of potentially serious
problems—it was hard to know where to start, there was a lack of feedback, and an important
radio button had confusing options. I estimate that the prototype had taken about 100 hours to
develop, and perhaps another 20 hours were needed to fix the problems before the first version
of the interface was deployed. Eventually, a rev 2 version was developed, which fixed additional
problems and added features.
"About a year later, I was teaching another usability workshop and I asked my class create and
test a paper prototype of the original rev 1 interface. The class conducted the testing in small
groups (with a student playing the role of user) and then the groups compared their findings.
Even though these 'users' had a different background than the real user population, they still
were able to identify most of the same problems found by the real user, including all of the
serious ones. In addition, the students had several suggestions for improving the interface that
(unbeknownst to them) had actually been incorporated into the rev 2 release.
"Had the paper test been done prior to programming, the first version of the product would have
been much stronger, and the second version might have been unnecessary or could have
focused on enhancing the feature set."
M. David Orr
(< [e-mail] >)