In designing this site I tried to use the techniques of "discount usability engineering". (Nielsen and Sano, 1994) These techniques are good for when a usability must be conducted in a short time frame. The methods are informal and rely more on the designer's ability to observe users than on statistics. But the methods are fast and can provide good insight into usability problems that exist.

Since this site is fairly simple design, with only two levels of pages, the point of the study was less serious concerns about usability but more about the experience of conducting a study.

I skipped the usual card-sorting test, where test subject sort randomized cards with the contents of the site into conceptual categories which make sense to the subject. The designers can then use these categories to determine the organization of the site. My site had very few objects, which seemed to fall into a natural organization, so a card-sorting test was unnecessary.

I did try to keep to a unifying metaphor of a term paper to organize the information. Thus I have categories for essay, bibliography, etc... I felt that this metaphor made sense, and would be familiar and so easy to navigate.

The first test I conducted was an icon intuitiveness test. I made icons for each of the areas on the page and placed them for viewing on the web. I then asked volunteers to describe what they thought the icons represent. Where the responses matched the intended meaning the icons were judged adequate. But in other cases icons were described as suggesting something other than the intended meaning and so they were changed.

After the organization of the site was complete, but before all of the content was finished, I observed more volunteers as they did a walk-through test. They were asked to find a number of pieces of information on different pages. All were able to find the proper page with minimum difficulties. Interestingly, one subject used each of the available navigation methods (side-bar, up arrow, and back button) instead of a consistent method of moving about. Therefore, I judge that the choice of navigation methods is a benefit to users and does not confuse them.

This project gave me a chance to learn by doing. Since the site was small, no serious difficulties were uncovered. But one lesson was learned about the conduct of tests. The walk-through tests must have well defined test moves in order not to confuse the user with vague instructions.

These pages were constructed using SuSE Linux, NEdit and VI text editors. They have been tested under Linux with the Mozilla and Konqueror browsers, under Windows 98 using Internet Explorer, and under Macintosh using Netscape. I have attempted to make them accessible to any system and browser. If you encounter any problems, please let me know.