Information Technology Reference
In-Depth Information
missed [49]. Using collaborative software that allows evaluators to view
each other's results while working, and later discuss them, has also been
shown to provide benefits. This process generates less problem
duplication, saving time for the evaluator and evaluation manager. It
allows evaluators to agree on how to classify results. This saves time in
aggregating the final problem list and allocating severity ratings [50].
Tasks and Scenarios
In heuristic evaluation, evaluators decide how to proceed. They are not to
use the system to perform real tasks, but rather “just inspect it freely.” The
effectiveness of this has been criticized [51], because results are dependent
on the skills, experiences, motivation, and curiosity of each evaluator.
Another approach is to provide evaluators with typical tasks or usage
scenarios to direct the inspection. The procedure chosen depends on the
evaluation's goal. If its goal is to acquire feedback on the usability of the
visualization - and the evaluators are double experts, domain experts, or
real users with their own goals - it may be possible to allow them to use
the system without any further instruction. They may be provided with the
goal of the visualization, and permitted to discover tasks of their own to
drive and test the inspection. If the evaluators are fairly naive with respect
to the domain of the visualization, the technique itself, and the aim of
using it, then more control is needed. It is necessary to assist them to
guarantee a careful inspection [51] by providing the steps a real user
would follow when performing typical tasks, and making sure all
important features are examined. When the goal is to answer a specific
question, or investigate whether a specific requirement or usability goal
has been achieved, scenarios and tasks are needed to ensure certain issues
are covered by the inspection.
When and How to Use
Heuristic evaluation can be performed at any stage of development, and is
fairly easy to learn and conduct. It requires neither extensive resources in
terms of participants, nor fully functioning and reliable software.
Heuristic evaluation can be used when a controlled study is not ideal,
the aim is feedback on usability issues, performance is hard to measure
objectively, or objectives and variables are not defined. It is also ideal
when it is either impossible to engage real users for an empirical study, or
when time and resources are short. Heuristic evaluation can be used as a
complement to more controlled studies. In an iterative design process, one
Search WWH ::
Custom Search