Geography Reference
In-Depth Information
independent variable. When differences in users are considered, the main focus has been on
differences between novice and expert users, while other factors such as the users' comfort
with uncertain information and their experience in making decisions are often downplayed.
Blenkinsop
et al.
(2000) assessed user perception of a variety of uncertainty representa-
tions, including both interaction and animation, using results of a fuzzy classification of
satellite imagery. The study examined the performance of two user groups, one expert and
one novice, in determining classification uncertainty. The results were framed primarily
in terms of effectiveness of the representation form (grey-scale images, histograms, ani-
mations, linked graphics and other methods) and not explicitly according to the differing
profiles of the user groups, but the differences in users were discussed. Cliburn
et al
. (2002)
developed a visualization environment to allow decision makers to visualize the results of
a water-balance model. That study found that the complexity and density of some of the
representation methods seemed to overwhelm novice users, while experts were able to use
the detail more readily in decision making. For example, they suggest that intrinsic methods
provide a more general representation of uncertainty data, which non-expert users may pre-
fer over more-detailed extrinsic representations. Aerts, Clarke and Keuper (2003) evaluated
two visualization techniques showing model results and the associated uncertainty, includ-
ing a toggling animation and a side-by-side static comparison. Their focus was specifically
on the end user, in this case urban planners, and on what uncertainty representations they
found most useful.
A review of these studies and those from other disciplines suggests that the effectiveness
of uncertainty communication may be as dependent on the user and context as it is on
the representation method or the problem domain. Thus, we advocate the reorientation of
research in uncertainty visualization to an approach that places the user at the centre of
inquiry, and maintains a goal of making uncertainty information as usable as possible to
the diverse community of users of geographic information.
14.2.2 The pervasiveness of uncertainty
The study of uncertainty is one of only a handful of themes that can be called truly
interdisciplinary; almost all academic pursuits, from history to mathematics to philos-
ophy to economics to linguistics, must consider uncertainty, whether it be variability,
fuzzy categories, doubt, inaccuracy, vagueness or any of the other forms uncertainty takes
on. Uncertainty is so much a fundamental piece of human experience that social an-
thropologists are discussing the possibility of an anthropology of uncertainty (Boholm,
2003; Anderson, 2006). Researchers in a broad array of disciplines - psychology, eco-
nomics, political science, medicine, information science, as well as geography - have
considered the processes by which decision making occurs under uncertain conditions
(Molden and Higgins, 2004; Dequech, 2006). Although uncertainty takes on many dif-
ferent meanings throughout these disciplines, there are several concepts and theories that
exist that may provide a foundation for uncertainty visualization and communication in
GIScience. In order to inform a user-centred approach to the representation of uncer-
tainty for decision making (a thankfully smaller but still daunting research theme than
'uncertainty' in general), we now turn to a review of uncertainty research in the psychol-
ogy of decision making, information science (specifically information seeking) and risk
communication.
Search WWH ::
Custom Search