Geoscience Reference
In-Depth Information
the specific character of the visual variable to respond to notarial tasks and in the
context of 3D visualisation instead of 2D plans. To limit the comparison of the
large number of contributors to perceptual design theory, three of them, well-
suited to our experiments, were selected. First, the work of Jacques Bertin, who
published a book on semiology of graphic in 1967 (Bertin
1967
), which is cer-
tainly one of the main references found in the literature. As a cartographer and
geographer, he proposed a theoretical foundation for map visualisation and his
work has influenced many scientists. Although some criticisms are expressed
about his work (MacEachren et al.
2005
), it was judged to be an excellent frame-
work to support a comparison between 2D and 3D about the suitability of visual
variables. Secondly, the work of Carpendale (
2003
) is interesting since she pro-
posed material additional to Bertin's works and revised it while taking into
account the migration of paper map to screen visualisation. The last reference is
the work of Ware (
2004
) who described the theory of information visualisation by
taking into account 2D and 3D object representation and visualisation. His work,
less experimental than Bertin's but in support of his principles, is based on physi-
ological, perceptual and cognitive psychology research without explicitly con-
sidering Bertin's visual variables. Both Bertin and Ware refer to Gestalt theory
(Koffka
1935
); an important foundation in pattern perception. Obviously these
three frameworks are much more complete compared to our work and we do not
claim to propose a similar agenda. However, these references appear compatible
and complementary to our experiments and could provide us solid foundations to
plan future works.
The comparison will only take into consideration the selective characteristic of
the visual variable since this is the only easily estimated concept that is compara-
ble between our work and the three others references. The ability to make choices,
the selective characteristics of visual variable is nevertheless the main influence to
performance of a particular task. A visual variable could be said to be selective if
users could easily distinguish the isolation between two groups of symbols solely
by their difference in this visual variable. The main question to be asked when
estimating the ability of visual variables is: Does a change in the visual variable
allow users to select one object from a group of objects?
Note that, in his framework, Ware did not include the notion of selection. For
comparison purposes, we extrapolate his term “categorization (nominal scale)”
used in his book as the most pertinent for isolation or selection between two
groups of symbols. Thus, the performance of visual variables for selective pur-
poses could be deduced from the rules of Ware has followed:
• Color: Ware (
Chap. 4
) concluded that “
color is extremely useful in data visu-
alization. It is excellent for labeling and categorization, but poor for display-
ing shape, detail, or spatial layout. When small symbols, text, or other detailed
graphical representations of information are displayed using color on a dif-
ferently colored background, always ensure luminance contrast with the back-
ground
”. Color is judged to be useful for selective purpose since in our case
study, categorization was involved.
