TownPartner
Case Study
There
are a number of risks of leaving usability out of the design equation.
Not only are the performance measures noticeably worse, but preference
ratings are also much lower. Perhaps most importantly, the task
failure rates are much higher- this means that in the real world,
users would not successfully complete a given task. In some cases
participants think they have completed a task when in fact, they
have not, while in other instances they give up knowing they have
not completed the task. In either case, the consequences are the
same-customer dissatisfaction. Again, it's extremely important to
have baseline numbers for comparison purposes.
It's
an established fact - in testing web sites that have been designed
without any human factors involvement, participants take longer,
commit more errors, experience more problems and require more assists.
Also, perception ratings for these sites are significantly lower.
No
involvement - usability test results
| Measure |
TP-
May 2001 |
PGE |
| Task
Times |
10:42 |
3:50 |
| Errors |
3.51 |
1.41 |
| Problems |
3.01 |
0.67 |
| Assists |
0:70 |
0.18 |
| Perception |
6.86 |
8.52 |
| Task
Failure Rate |
45% |
4% |
A recent
usability test resulted in 120 recommendations for improvement to
the design. Even with limited human factors involvement, the results
of the usability test were significantly better.
Limited
involvement - usability test results
| Measure |
TP
- April 2002 |
PGE |
| Task
Times |
5:31 |
3:50 |
| Errors |
1.21 |
1.41 |
| Problems |
1.92 |
0.67 |
| Assists |
0:29 |
0.18 |
| Perception |
7.57 |
8.52 |
| Task
Failure Rate |
28%
|
4% |
©
2002 Usability Concepts, LLC
|
