Usability Testing

From Tango Desktop Project
Revision as of 19:49, 18 March 2005 by StevenGarrity (talk) (Make a numbered list a real numbered list)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

The development of the Tango Project will include rigorous usability testing. We believe that doing such testing, in conjunction with careful application of the relevant design guidlines and heuristic evaluation, is the best way to ensure that we create a usable desktop.

There are a lot of myths surrounding Usability Testing, namely that it is difficult to do, and requires complicated equpment. Neither of these things is necessarily true; the links below provide an introduction to a simple, cost-effective method of usability testing. Please review.

Why is Usability Testing important for Tango?

There are three main reasons:

  1. It provides us with a mechanism for determining scientifically how to make a more usable desktop, and,
  2. It enables more people to contribute to Free Software, regardless of development skill or artistsic talent, and,
  3. It allows us to harness the diversity of our home communities to mutual benefit; it allows members of our home communities to teach us about their needs, and it helps us to craft our software to better serve them.

How does Usability Testing work?

Conducting a usability test is a lot of running a role playing game. You make up a story, and you ask the people who you are testing to go along with the story, and to try to accomplish something specific with the bounds of the story. The following links explain what the various parts of a usability test are, and how you, yes you, can run one.

  • Writing Tasks: The core of a usability test is the set of tasks that the test subject will be asked to complete during the test. These tasks should be based on operations that someone in the real world might want to use your software to accomplish.
  • Making Software Prototypes: After you've written the tasks for testing your software, you are ready to make software prototypes for your test subject to interact with. These prototypes can take many different forms, from low-fidelity sketches, to software running on a computer. The mockup form which is appropriate for your test depends on where in the design process your software is.
  • Picking Test Subjects: In theory, anyone who is able to understand your tasks and interact with your software prototype could be a test subject for your usability test. However, the biases and experiences of your test subjects will color your results. In practice, you are likely to get the most useful input from people who are actually members of your target audience.
  • Running the Test: A big and hearty section which discusses how you and your test subject should interact during your test.
  • Analyzing Test Results: - After you have finished your test, it may be tempting to start hacking immediately. However, some of your work will be wasted if you don't first take some time to summarize your results.