#1 The presence of trained specialists.
Presumably the key component that determines the length of time to
conduct a CW will be the amount of experience the team has with the method.
Having a trained specialist who is experienced with interface inspection
and someone that has the knowledge regarding the problem solving model
would greatly increase the turnaround time. Although a usability
specialist is not required, having individuals with exhaustive knowledge
of the method is important for its success and efficiency (Lewis, Polson,
Wharton, and Rieman; 1990).
#2 The complexity of the system.
Having a system that is extremely complex will increase the time for
the CW because it will require more time to develop the task list, the
user profile, and the task sequence. Increased complications in any one
of these areas will result in increased evaluation time and undoubtedly
require more explanation and interpretation.
#3 Stage of development.
Depending on what stage in development the system is being tested,
the length of time to complete the CW will vary. For early prototypes,
the length will likely be shorter due to its decreased complexity and limited
features. More complex systems will require more preparation
and evaluation time (see #2 above).
#4 Budget
Obvious the budget is going to determine to a large extent the turnaround
time for completing the CW. If the organization has the resources
to hire a trained specialist, then presumably the CW will be faster.
However, the strengths of the CW lie in its usefulness for non usability
specialists and therefore can reduce the cost of UI evaluation.
#5 Time frame for completion
Having a varying time frame for completion will determine the depth
or detail that the CW can be done in. Having a long time will allow
thorough inspection of each task, while a shorter time frame may require
the elimination of some tasks and a focus on the most important ones (Rowley
and Rhoades, 1992).
#6 Detail required
This factor is closely linked to #4 and #5 as having a higher budget
and a longer time frame will allow the CW to be performed in more detail.
However, depending on the evaluation goals that are set out, the CW can
be performed in more or less detail.
#7 Who the audience is
Having a clearly intended audience will influence how the CW is conducted
and how the results will be presented. It is possible that the CW
may not be useful in its entirety if the audience does not comprehend the
basic cognitive problem solving model. In this situation, a different
walkthrough may be more appropriate or the interpretation may be more informal
and less technological.
For example, Rowley and Rhoades (1992) demonstrated
that by modifying the CW process slightly to include video equipment and
an interactive evaluation session, a faster process could be achieved.
They compared to information derived from both methods and found that although
the faster form (labeled the cognitive jog through) highlighted the same
main problems, it lacked the detail that the CW provided. Similarly,
Rieman et al. (1991) developed an automated version of the CW. Results
indicate that the methodology can be introduced in about an hour and the
automation allows a user action to be analyzed in as little as 5 to 10
minutes. Therefore, the degree that the structure and guidelines
of the CW are followed depends a great deal on the goals of the evaluator(s)
and the time frame in which they have to work.