Carolyn Snyder presented about usability testing based on her experience as a usability specialist. The information was basic (meant to be an introduction), but I still learned a couple things.
I was pleased to see the number of hands that went up in the audience when Snyder asked who has ever done a user test. A vast majority of the audience claimed to have done so. Awareness is building.
The main lesson I learned from the presentation was to choose one type of user for a given study. The user should be prioritized as critical to the interface’s success. And the user should be someone who cares about what the interface has to offer. When I did user studies for a school I worked for, I felt like we were trying to choose an eclectic mix of people. But the more I think about it, they still fit the criteria of being people who cared about the interface because we did our recruitments at career nights at the school.
Snyder did talk about the number of people your test should be run on. She presented the Jakob Nielsen thought that five users are enough, but she recommended more. This assumes that picking more test subjects will provide data for a perfect interface. In my opinion, you will never have a perfect interface. There are just too many variables involving types of people and their goals. I agree with Nielsen that testing five users will take care of the majority of your interface’s issues. If you have the resources to test more, go ahead. But if you’re concerned with a major improvement, five users will suffice.