Information Technology Reference
In-Depth Information
Discussion and Conclusion
Issues considered in this chapter include: what heuristics to apply, who
and how many evaluators to engage in evaluation, and what procedures to
deploy during execution that are crucial for the outcome. The fact that it is
a discount method (quick, cost-effective, and fairly easy to conduct etc.)
speaks in its favour for several reasons. However, there are drawbacks: the
method is subjective and depends on the evaluators, and one cannot expect
to find all problems. Nevertheless, the method can be useful and
information visualizers are encouraged to consider using it more often.
The method can generate feedback at all stages and is suitable for
assessing qualitative aspects of use and high-level cognitive tasks. It can
address the questions: what is important to support exploration for gaining
ideas and insight; and what is affecting cognitive load while working?
Such questions cannot be fully answered by objective quantitative
methods measuring performance. They need to be assessed and discussed
in terms of qualitative dimensions expressed by heuristics and guidelines.
Heuristics can be useful in other ways to obtain feedback, e.g., used as
statements in questionnaires, or as categories when analysing and
reporting outcomes from using other methods such as interviews and talk-
aloud evaluations.
To improve the method for information visualization, there are clear
directions for future work. The heuristic sets in the section
Heuristic Sets
have, with a few exceptions, not been validated. Lack of repeated use of
the same heuristic set makes it difficult to judge the usefulness of the
heuristics, and to compare results between different information
visualization techniques or different evaluations of the same technique.
This could identify patterns of usability issues and improve the ability to
judge the validity of results.
There is no consensus about what kind of heuristics could be useful for
assessing information visualization. Initial work to find a small set that is
general enough to be used for evaluating techniques could be continued
and validated. Similarly, there are no studies to investigate how many
evaluators are needed for heuristic evaluation in information visualization.
Search WWH ::
Custom Search