Information Technology Reference
In-Depth Information
Freitas et al. [40] stressed the importance of assessing visual
representation and the interaction mechanisms provided by an information
visualization technique. Based on previous work by Scapin and Bastien,
who introduced criteria for evaluation of the ergonomic quality of
interactive systems [41], they proposed a set of ergonomic criteria that
encompass both aspects. Four classes of criteria (or heuristics) were
proposed for evaluating the usability of a visual representation:
completeness
,
spatial organization
,
information coding
,
state transition
;
and three classes for the interaction mechanisms:
orientation and help
,
navigation and querying
,
and data set reduction
. In subsequent work [42],
the criteria were used in an evaluation that also compared them with
Nielsen's 10 heuristics, and Scapin and Bastien's ergonomic criteria. This
set has been applied and validated. When proposed, the set was tailored to
evaluate hierarchical data representations; the authors suggest their
applicability for other techniques as well.
It is evident that there is an ongoing effort in developing heuristics for
information visualization. All the sets proposed so far cover important
aspects of use. Each researcher must decide which set is appropriate to
use. Still, there exists no general set applicable to any visualization
technique that is equivalent to Nielsen's set of
Ten Usability Heuristics
for
the HCI community. In [7] the authors presented a first effort in deriving
such a set. Based on 63 heuristics from six published sets (including most
of the sets above), they synthesized 10 heuristics that, as a set, had the
greatest explanatory power (using the same approach as Nielsen). The
resulting set consisted of the following heuristics (listed by name and
reference to its original set):
information coding
(Freitas et al.),
minimal
action
(Scapin and Bastien),
flexibility
(Scapin and Bastien),
orientation
and help
(Freitas et al.),
spatial organization
(Freitas et al.),
consistency
(Scapin and Bastien),
recognition rather than recall
(Nielsen),
prompting
(Scapin and Bastien),
remove the extraneous
(Zuk and Carpendale),
and
data set reduction
(Freitas et al.). As with Nielsen's work, this set was
shown to provide the best heuristics (out of 63) for explaining known
problems (here a pool of 74 problems). The set has subsequently been
applied for evaluation, for example, to explain unknown problems in [43]
and [44]. Further evaluation, refinement, and validation are needed until
there is a standardized set like
The Ten Usability Heuristics
[28]
.
Search WWH ::
Custom Search