Usability test

The first stage of a usability test consists of familiarizing with the product. The test designer watches demonstrations and follows training, if necessary, to gain a full understanding of the potential users and how they wish to use the product.

The next stage is the creation of a protocol. The protocol is a kind of script with clear instructions for the usability tester(s). It describes in detail all the steps of the usability test: the scope, goals, criteria and settings for the tasks to be performed by a test participant, and the instructions to be followed and supplementary questions to be asked by the tester. Supplementary questions refer to e.g. general comments and opinions, perception of the design, suggestions for improvement…

The practical organization of a usability test starts with the tests scheduling and the recruitment of test participants. These are to be representative of the user population as a whole. Particular attention is given to choosing the right number of participants and to ensuring that a suitable range of skills and experiences is available. A usability test however mostly is not statistically proof. Most usability tests are qualitative evaluations because discovering the biggest design flaws does not require dozens of test sessions.

Usability tests can be set up in a lab, but test results are not less reliable when users are invited to a more familiar environment or visited in their own surroundings. Although several test techniques can be applied (e.g. wizard of Oz, what-if, walk-around...) the most often applied technique is one where participants are asked to think-aloud while carrying out a set of intended tasks. This means that the user explains what he thinks while performing a task. A usability lab facilitates recording the test and the participant's actions, supplementing the test with eye-tracking and partly automating the interpretation of the test results by means of specific software tools.

In the final stage of a usability test the test results are interpreted, reported and optionally transformed into concrete recommendations for improving the design.