Usability Testing

Overview
Usability testing is a process that involves observing users interacting with a document, website, object or other product to improve usability, which is concerned with "how well users can learn and use a product to achieve their goals" [8]. Usability testing, which focuses on "making things more usable and accessible is part of the larger discipline of user-centered design" [7]. User-centered design falls under the even larger umbrella of user-experience (UX design). Because the primary purpose of technical communication is to help a user/reader accomplish a goal, such as understanding a concept or completing a task, knowing how to conduct usability testing can be a useful skill for technical communicators.

Usability is thought of in terms of five basic qualities: [5]
 * 1) Learnability: How easy is it for users to accomplish basic tasks the first time they encounter the design?
 * 2) Efficiency: Once users have learned the design, how quickly can they perform tasks?
 * 3) Memorability: When users return to the design after a period of not using it, how easily can they reestablish proficiency?
 * 4) Errors: How many errors do users make, how severe are these errors, and how easily can they recover from the errors?
 * 5) Satisfaction: How pleasant is it to use the design?

It is important to note however, that it is difficult if not impossible to 'test for' usability because "true usability is invisible. if something is going well, you don't notice it" [7]. For this reason the main aim of usability testing is to identify aspects of the product that cause problems for users in the five aspects of usability. Some Key Principles of usability testing are:
 * Involve users early and often in the product development process [2].
 * Listen and observe carefully to learn about your users needs.
 * Use iterative testing to guide design process [6].

Iteration: While resource and time availability may limit the amount of usability testing that can be included in any given project, an ideal product design process employs several iterations of usability testing. Iteration involves repeating cycles of usability testing, revising the design based on the findings of the tests, usability testing the revised version, and so forth. Even several reiterations of small tests with as few as five participants can provide valuable insights that can vastly improve the usability of the final product. According to Donald Norman, a pioneer in the field of user-centered design "it is far more effective to do one test of five, use the results to improve the system, and then keep iterating the test design until you have tested the desired number of people. This gives multiple iterations of improvement, rather than just one."[6]

Methods
Because usability testing is not exclusive to any specific testing technique or methodology, almost any activity that involves users interacting with a product to glean information about how well the product works for real people can be considered usability testing, The basic components of a usability test include finding representative users, asking the users to perform tasks using your product or design, and observing where they succeed or experience difficulties [5]. As a result, the level of formality, scope, and methodology of a usability test varies greatly depending on the particular situation as well as the researchers needs, abilities or resources. On one end of the spectrum are simple, quick and informal usability tests (such as guerilla testing) that can be carried out by a small team or lone researcher. On the other end are complex, resource intensive usability tests conducted in a laboratory setting that employ advanced software and large research teams. While usability testing methods are highly variable, two examples of informal usability tests that are useful in technical communication are guerilla testing and the 5 second test.

Guerilla Testing
Guerilla testing is an extremely informal method of usability testing that is quick and not resource intensive. To conduct a guerilla test, researcher(s) simply recruit participants by finding people who are easily accessible (co-workers, or people on the street/in a mall), ask them to try and use the product, and make observations about what does or doesn't work well [1]. The observations are then used to improve the usability of the product. Testing with small numbers of users in this fashion may be iterated or repeated to produce increasingly usable designs. Additionally, guerilla testing can be combined with other techniques, such as the five-second test, to provide a more comprehensive view [1].

Five-Second Test
A five-second test is another fast and simple usability test that helps researchers "see how clear and memorable a given moment in the product or service is to users"[1]. In this test, a user is exposed to a document, screen, or product for a period of five seconds. When the time is up, the product is taken away, and the user is asked to describe what they remember, and their impressions of the product. This test offers high level information about users impressions of the product in a very short amount of time. Five-second testing can be used in combination with guerilla testing to rapidly produce rich usability data [1].

1. Test Plan
A test plan "addresses the how, when, where, who, why, and what of your usability test" [7]. It serves as guide for the test process, and as a unifying "stake in the ground" for team members and other stakeholders. The components of a test plan are the test purpose and goals, research questions, participant characteristics, test method/task list, test environment, moderator information, data evaluation measures, and report content. [7]. Some sections can be excluded as relevant for the testing situation, but it is always advisable to create a test plan to clarify goals, questions, and methods that will be used.

2. Participant Recruitment
Participants selected for usability testing should be representative users of the product being tested. In other words, participants should be, or be essentially the same as the end users, or the people who will ultimately use the product. Questionnaires or surveys are often used to ensure that appropriate participants are chosen. The matter of how many participants are needed "is one of intense interest and discussion in the usability community" [3]. The general consensus among researchers in the field is that tests including as few as five participants can detect 80% of major usability problems, while tests with ten participants will detect 90%. [3].

3. Testing
During the testing phase, the usability test is conducted as outlined in the test plan. Participants complete the tasks or other activity, and data/observations are collected. Testing can occur in public, in a participants home, in a laboratory, office, or other setting.

4. Analyze Data
Depending on what stage of development a document/product is in, what the research seeks to learn, and what testing methods are used, testing may produce quantitative or qualitative data, or both [9]. Quantitative data includes hard numbers, such as the percentage of users that were able to successfully complete a task, the average amount of time users took to complete a task, or the number of mistakes that were made [9]. Qualitative data includes open-ended data, like users thoughts and opinions about the product. Numerical data can be tabulated and categorized in spreadsheets to facilitate analysis. Similarly, qualitative data can be organized based on theme, feeling, problem, or other organization strategy.

5. Report Findings and Recommendations
Findings are the implications that can be drawn from the data, and can be general or specific. For instance, a general finding may be that a certain part of the design is confusing to users. A specific finding might be that users could not read a piece of text because of the font used. Recommendations are the design changes suggested to remedy the problems revealed in the findings. Depending on the situation, technical communicators may communicate findings and recommendations to other team members in a meeting or conversation, or they may be included in a formal report produced after the test has been completed.

Think Out Loud
While most usability tests revolve around observations about the users actions made by researchers, 'think out loud' protocol also relies on the participants own thoughts during the test. By asking users to 'think out loud' or 'think aloud' as they attempt to complete tasks or explore a document or product, technical communicators can glean valuable information about the users thought processes. Because it is not always obvious why users experience difficulties with certain tasks, insights from think aloud tests are particularly useful in illuminating aspects of a design that contribute to user confusion or misunderstanding [4].

Moderated vs. Un-moderated
In a moderated usability test a member of the design team (a moderator) is present as users engage with the product or use it to complete a task. Depending on the specific usability test, moderators may read instructions, guide the user through a task, or ask questions about the users thinking process. Moderated tests can conducted in-person or remotely. Common moderated usability tests include: Card sorting, phone/video interviews, lab testing, or guerilla testing. In an un-moderated usability test a user interacts with a product or uses it to complete a task on their own, without guidance from a member of the testing team. Un-moderated usability research might utilize basic observation, recorded sessions, or software that automatically tracks a users keystrokes, eye-movements, or other activity.

Remote vs. In-Person
Testing can be conducted either in person or remotely. In an in-person test session, researchers observe the users directly, and in real time as they interact with the document/product. In a remote test session, the researcher might record the session (video and/or audio), observe remotely via video chat type software, or interview the user over the phone.

Design Phase
Usability tests can be grouped according to research goals and the product development phase in which they are typically utilized. Exploratory testing aims to gain broad level information about users goals and needs which can be used in early phases of product development. At this stage, users may interact with a paper-prototype or another early design prototype. Assessment or summative testing explores how well a user can perform specific tasks with a prototype or complete product and is typically employed early or midway into the design process. Validation or verification testing is used to compare users performance with established benchmarks. For example, a validation test might seek to determine what percentage of users can successfully complete a task using their product, with successful completion defined in terms of a time benchmark (e.g. in less that five minutes). In comparative testing, users interact with two versions of a product to determine which is more usable. Comparative testing can be used at any point in the design process, and in combination with the other forms of testing; a validation test comparing two different product versions for example. The following table illustrates the different types of usability tests, their respective goals and the design process stage in which they are commonly employed [7].