Assessing Usability 3: Inspection
Whereas the assessment methods identified in Part 2 of this series, Discovery, provide the most benefit at the very beginning of the concept development for a product, the inspection methods detailed in this section are really only useful when you have actually started design.
And while these methods (Heuristic Evaluations, Cognitive Walkthroughs, and Pluralistic Walkthroughs) can be used at any stage of the design process, they typically provide greater value at the beginning of design. The earlier you can correct usability-related issues in the design phase, the more solid your foundation will be for the rest of the product.
The methods described below are typically internal assessments. That is, they don't require outside users for evaluation. Rather, they rely on usability experts (either internal to the team or from an outside firm) and other project team members to identify and evaluate the usability issues that need to be corrected as design progresses.
Heuristic Evaluations
"Heuristic Evaluation" is a simply an important-sounding term for having a group of experts evaluate the usability of a product design using a list of commonly-accepted principles or guidelines for usable design.
Unfortunately, this technique is often used in place of a more formal usability test (with real users), as it is cheaper and easier to do. Although we advise against the elimination of a more formal user-based usability test, budgets are rarely unlimited. Therefore, Heuristic Evaluations have become, in many instances, the primary usability assessment being conducted on products. If this is the case with your product, it is important to make sure that a number of experts, preferably with multidisciplinary backgrounds, conduct the evaluation. Studies have shown that 5-6 experts should be adequate to capture the bulk of the usability issues associated with a particular design. Further, by having multidisciplinary experts do the evaluations, different types of of usability issues (i.e., interaction, navigation, readability, information design, accessibility, etc.) can be addressed within a single evaluation.
The experts should each evaluate the product on their own, to avoid biases, and should each employ the same usage scenarios (explicity defined user roles and example activities help the evaluator adopt the proper "user perspective") for evaluation. Once the individual evaluations are done, the experts should jointly classify and rate the severity of the issues uncovered.
Note that Heuristic Evaluations are not a "do it yourself" approach to usability assessment. Although established heuristics do exist, they are somewhat general in their description and require the skills and knowledge of an experienced evaluator to be correctly applied to an individual product design.
Cognitive Walkthroughs
As with Heuristic Evaluations, a Cognitive Walkthrough requires experts to adopt specific usage scenarios when evaluating a design. The walkthrough is typically conducted on a prototype system with a representative set of user goals. Again, it is much cheaper to correct usability problems earlier in the product-development lifecycle - once the design has reached a certain point, modifications can easily double the entire project cost, depending on the magnitude of the changes needed.
In a Cognitive Walkthrough, experts proceed through the product design, trying to achieve their goals and complete the identified tasks as if they were users of the final product. Each step in the task is deliberately and intently examined for any "frictions" - or barriers - to users being able to understand the system, complete required tasks, and accomplish their goals. Examples can include an input screen where there is a lack of the right type of data needed to make a knowledgeable choice between options, counter-intuitive control operation where the user's perception/action expectations are violated, or situations in which prior knowledge required to complete a crucial task may not be provided by the current system, but instead relies on a transfer of knowledge from an earlier - or completely different - system.
Cognitive Walkthroughs can be conducted with either a very broad or a very granular focus, depending on the importance of the task. The more crucial a task is to the overall success of the product, the more benefit gained in decomposing that single task into smaller sub-tasks, providing the opportunity for a more detailed evaluation. It is recommended to also examine the users' goals - independent of the tasks needed to achieve them - both for decomposition and evaluation purposes.
Pluralistic Walkthroughs
Pluralistic Walkthroughs are similar to Cognitive Walkthroughs in that during the evaluation the participants step through usage scenarios, identifying barriers to a user's ability to achieve their goals. A main difference, however, is that Pluralistic Walkthroughs are typically conducted by a group of people that often includes members from the different parts of the design/development team and actual end-users.
In these group walkthrough sessions, the Usability Expert will primarily facilitate the discussion that takes place between the various members of the design team and between the team and the end-users. Besides the fact that more evaluators will typically find more problems, the group setting provides a unique opportunity to have people with varying skill sets and perspectives to discuss the cause and potential impact of each of the problems.
Be warned that these discussions can, at times, be difficult to facilitate, especially if members of the design team are present. Often, the facilitator must work to help the designer overcome the feeling of being persecuted (i.e., everyone is "ganging up on them" to point out the deficiencies in their design) and take a more positive approach to the discussions.
An interesting artifact of Pluralistic Walkthroughs is that they can double as an educational session for many of the participants. By listening to other perspectives on what is considered to be a usability problem and how serious or trivial that problems might be, group members often come away from the session with a better understanding of how difficult is can be to accurately fix usability problems.
Finally, you may find that the interaction afforded by the group setting allows the team to develop solutions for the problems more quickly than with the previous two methods.
Next Steps: User Testing
The final article in this series, Assessing Usability 4: User Testing discusses some more formal, user-based approaches to usability testing.