Study Participants and their Treatment

Hello again, everyone! Now that you've made a wireframe and set up your own usability lab, it's about time to actually sit down and test your design. In order to test your design, however, you'll need victims study participants! This all might sound like a big deal, but surprisingly enough...

You only need five participants in order to discover 85% of your usability problems (Neilsen 2000) (it's a bit of a contended point (Spool 2001), but five participants is a generally accepted standard in the industry).

Of course, the more users you have, the better (Neilsen 1993), but at the very least you should have at least one person review your design before you do anything with it.

All right, so you know you have to have a certain number of participants now. But who should they be? If you can, try and ask people from your website's potential target demographic to help you out. If that doesn't work, friends and family will do, too. But when you're recruiting participants, you need to make sure they're being honest with their feedback.

Your research participants will probably be eager to please, and will look for ways to praise your design rather than critique it. The critique, however, is what you really want out of this- praise is nice, but it doesn't help you improve your design at all. In order to avoid this problem, make sure to let your participant know that you want to know what they do and do not like about your design.

There are certain other things that you, as a moderator, need to do when you are conducting usability tests as well, mostly so that you can get your participants to give you high quality feedback.
A few of the more important ones are:

•  Don't ask leading questions. If you're asking questions that are biased in one way or another, your participants will be more likely to answer the way you want them to, even if it's not how they truly feel. 
• Don't try and direct your participants. Beyond giving the participants a task to accomplish, try not to force them to traverse the design in a certain way. If it's usable, they will easily be able to find out how the design is intended to work; if not, then you've just discovered something you need to work on when you revise the design.
• Don't be biased with your praise. Praising your participant is good- but make sure to praise "failures", as well. Repeated failure is frustrating, but if your participant understands that their mistakes are actually a good thing, they will hopefully feel better about the times when they aren't sure what they're doing.

Sources:
Spool, J. and Schroeder, W. Testing Websites: Five Users is
Nowhere Near Enough. In CHI 2001 Extended Abstracts,
ACM, 2001.

Nielsen, Jakob and Landauer, Thomas K. A
Mathematical Model of the Finding of Usability
Problems. INTER CHI "93 206-213.

Nielsen, Jakob Why You Only Need To Test With Five
Users, Jakob Nielsen's Alertbox, March 19, 2000,
http://www.useit.com/alertbox/20000319.html