Geography Reference
In-Depth Information
''lines'' experiments to foster an explicit topology. The classic London Under-
ground (LU) map is a good example of this category, where the stations are
regularly spaced and the various lines are separated, put in parallel and made to
adopt a regularized geometry for ease of navigation. In this map, metric distance
and direction are not as important as connection and separate identity. It would be
inconceivable in the LU case to remove the connecting lines, yet the differing
function and the simpler real world transport network of the SRCC display gives a
better result without lines. This is despite anisotropic effects in the South Island
case, i.e. physical barriers to movement emphasising the connecting role of the
roads, suggesting that the simpler geography makes for a more complete mental
model, lessening the necessity for lines.
A final note is that connecting lines, where used, are likely to be more effective
in topological than Euclidean representations due to the flexibility that topological
space affords. Figure
4
from the second case study is a good illustration of what
can happen when the Euclidean metric is followed slavishly.
Section 6.2
offers an
alternative cartographic generalisation solution to the situation—processes such as
displacement and typification represent a departure from the precision require-
ments of Euclidean space, and in doing so, take a step towards topological
representation.
6.1.3 Grouping in Layers and Cartographic Generalization
The use of Multi Layer Displays in case study 2 was intended to make use of the
Proximity Compatibility Principle again, with like-priority data being integrated
on the same physical layer. The aforementioned Trial 24 represented a situation
where this effect did not work. This was attributed to attention tunnelling, where
participants were so focussed on the incident and resource information on the front
layer of the display, they failed to recognise the extent of the harbour on the rear
layer. Therefore, while the MLD appears to be successful in making the important
information salient, it can be at the expense of background data, which in this case
was equally salient.
This finding was reflected in a study by Yeh and Wickens (
2001
) when they
were exploring the reduction of visual clutter through hiding display elements—it
is difficult to decide which are primary information and which are secondary.
Often this is dependent on the situation encountered by operators and as such
information that is considered secondary in some instances may be absolutely
critical in others. Another phenomenon Yeh and Wickens (
2001
) found in their
study was that the time cost of turning information on and off (including deter-
mining if the hidden information needed consideration) outweighed the time
benefits of presenting less information, due to increased cognitive load.
An interesting parallel can be drawn here with the cartographic generalization
process of elimination, the use of which seems to be complicated in a dynamic
display environment. It is not as easy to eliminate (albeit temporarily) display
features in a vital decision-making environment such as that for ambulance
