The Politics of Paper Prototyping - Paper Prototyping: The Fast and Easy Way to Define and Refine User Interfaces - page 223

Figure 13.1: The search form used in testing a hotel reservation Web site.
This example is a fairly simple one; you can imagine how the potential for bias might increase for more
complicated tasks. Unlike when you test with a non-profile user (which your team members will readily
notice and complain about), bias due to tasks can lurk undetected beneath the surface of your findings.
Because it's virtually impossible to eliminate bias, you should at least look at your tasks and question
whether there is a bias that might interfere with the most important things that you're trying to test.
Bias: Unrealistic Test Setting
Real people don't do their work in a usability lab. In the user's natural environment, there a myriad of
factors that may influence how they use the interface: lighting, interruptions, other software,
performance measures in their jobs that they're rewarded for, and so on. As mentioned at the end of
the previous chapter ,there is a whole set of problems that you're unlikely to find by bringing in users;
you need to go to their environment and watch them.
Here's one of my favorite examples. A few years ago, one of my colleagues consulted with a company