Professor Michael Twidale from the University of Illinois came to talk with us this week. Professor Twidale is a very insightful and personable character with a unique perspective on Human-Computer Interaction (HCI). In the case of Illinois Wesleyan, we are interested in improving the usability of our website. We’ve spent considerable time and money converting our campus web pages into a content management system, and now we’d like to evaluate them to be sure that people are able to find and understand the information we present.

I think it is important to note that usability isn’t just about computer systems or user interfaces. It can be an aspect of something as simple as a campus visit. For example, if an alum from the Class of ’71 comes to campus and tries to enter the old Sheean Library, they are puzzled when the lights are off and the doors are locked. Nobody bothered to put a sign or a map in the windows of Sheean – everyone on campus is familiar with the changes of the last few years. These changes are not intuitive to a person from outside our frame of reference.

A couple of the most important aspects of Professor Twidale’s visit are the emphasis he places on making educated decisions about trade-offs in web/information architecture design, and his assertion that testing and fixing usability problems doesn’t have to be slow or expensive.

With regard to trade-offs, Twidale is talking about understanding that making a change to benefit one user will affect other user experiences. If we decide to rename our “Admissions” link to “Enrollment”, those people who have already learned how our website worked will be alienated. On the other hand, that may be the preferred terminology for young people (our primary audience). As long as we consider the impact of a change or a choice, the decision is ours.

I personally think the emphasis on cheap, quick testing is the real gem of this discussion. A room full of people could argue for days about whether to include a bulleted list or a drop-down list, and their decision would ultimately be based on their own personal biases and preferences with regard to interfaces. “Nobody likes drop downs” or “Bullets are confusing” are the types of arguments brought up with people make decisions that haven’t been informed by testing. Get 4 or 5 people who are outside your organization to try it out and see how they interact with your website. Then make the change and see what happens. Don’t steer the test, just observe and ask the subject to think out loud. They will inevitably discover problems that you didn’t even know existed, and perhaps those problems can be fixed in a matter of minutes. It is amazing how much better a web session can “flow” with a few 10-second link name changes.

Read the article Usability at 90mpg, by Michael Twidale and Paul Marty

Another great resource is Jakob Nielson’s web newsletter AlertBox