In my last post I brought up the main factors that determine usability for a product or system. Now it’s time to consider various approaches to testing these components and gathering practicable information on usability.
There are different types of usability tests depending on development stage, goals for testing, and resources.
(1 ) A Heuristic Review involves having a trained usability tester evaluate a program and compare it to standards or guidelines that are generally agreed upon in the industry. http://www.measuringu.com/blog/he.php is a good article on effective Heuristics testing.
(2) In a monitored or formal usability test, a facilitator will set up a physical testing environment. Participants complete tasks while being observed by the facilitator. Data is collected from video/audio of the test session, along with the facilitator notes and any post-test evaluation.
(3) An ad hoc usability test is less prepared than a formal usability test. The participants may be chosen randomly off the street (referred to as hallway testing) and/or the testers may simple be observed using the program without prompts on exactly what tasks to perform. This type of testing will be cheaper and easier to repeat throughout the design process.
(4) A competitive usability test will have participants compare two or more similar programs and answer questions about the performance of each. This type of testing will give the development group a sense of how the program is stacking up to other existing ones.
(5) A usability questionnaire, often administered after a physical usability test of some variety, is a way of getting systematized written feedback from testers. The Software Usability Measurement Inventory (SUMI) is one example of a standard usability questionnaire.
Because the GNOME usability testing internship involves conducting a formal usability test, I want to mention several indispensable components to this type of research. I can’t say that these are hard and fast rules, but they seem to be the precedent for any good monitored usability test.
(1) Usability test are conducted with human participation. The subjects are often going to be representative of the user base for a program, though professional usability testers are needed for some types of tests.
(2) Practical, real world tasks are given for the participants to perform. These tasks are carefully considered and remain consistent for all test subjects.
(3) The participants are monitored while completing a usability test when possible. It’s important that the data come from observing and interacting with the subject performing the tasks rather than a mediated post-task evaluation of the testers experience alone(though that is valuable too).
(4) The combination of facilitator notes, video/audio recording of the test, and tester feedback (verbal or written) is used to create a report providing both qualitative and quantitative data about the usability of the program. This information directly affects the next iteration of the programs development.
As you can see there is a lot of variety when it comes to usability testing. It is crucial that a development team do the research and decide what type(s) of usability testing will work best for the specific circumstances of any given project.