It Works and We Like Using It

No matter how comfortable and well upholstered a chair is, if a person’s goal is not to sit down but to take a shower they need a shower not a chair. With basic and common sense approaches to system design authors on task analysis and usability outline ways to test and examine real people performing actual or visualized tests so that the resulting system designs both function and do so elegantly; these methods can compliment each other to derive the best from both forms of analysis.

In the early days of the Web, basic text sites, with just simple links sold merchandise; one classic site that sold quite a bit of merchandise was a Christmas gift site, center-justified colored text with many links running straight down a single page decorated with randomly placed candy canes. Certainly such sites were not easy to find, nor easy to use and, in a word, ugly. But they were useful.

One might think by now we were way past questions around online goals, and into customizing beautiful interfaces, but with the plethora of reasons to use Webs the need to evaluate what people’s actual goals are in unique situations and combine ease of use customized with real goals has grown as the world reaches past the 100 million mark of live sites(1). We seek to both have things work and enjoy using them; using these techniques ensures both are likely.

To help designers improve systems Task Analysis focuses on what real people do to achieve their goals (Greenberg, 2003), and Usability Evaluations are based on evaluating how discoverable and possible the interface allows for people to achieve their goals (Krug, 2000). Both have the same goal – usable systems – but by different means. The difference is Task Analysis captures the working environment and demonstrates who the user is and what they are trying to achieve, what their goals are including failures. In the Usability Evaluation method the goal is to help designers increase how well end users can accomplish tasks, and even measure the difference from old systems to new ones.

Following the advice of these authors, from the beginning envision incorporating task centered design to follow and watch how users actually work, and make notes as they demonstrate the Web based software for the design process.

As you learn more about a system you tend to forget that you are making due, using workarounds to make up for system flaws, so it is just as you enter a system that it is best to carefully document and record what your own experience as well as the users immediately.

Paper prototypes are a swift approach to get to the first level of task centered design, and it is with this approach that the beginnings of usability start. Working with teams dispersed worldwide one plan of action is to rely first on local teams, using their feedback begin to construct clickable wire frames, iteratively requesting feedback. Scenario based design (Carroll, 2000) will also help make the designers job easier in terms of what it takes to display lots of data and make it useable.

Several authors make the same point that having just a few people perform some usability tests is better than no data at all, notably Nielson. All in all combining pulling from several methods from beginning to end will make both the process and the results better for everyone.

(1)“Web reaches new milestone: 100 million sites” CNN article located at: http://www.cnn.com/2006/TECH/internet/11/01/100millionwebsites/index.html?eref=rss_topstories

Written for IMT 540

Dr. Dave Hendry's University of Washington's iSchool Website