Making Uncertainty Usable: Approaches for Visualizing Uncertainty Information - Geographic Visualization: Concepts, Tools and Applications - page 284

information control. Instead of making uncertainty information available at detailed levels
for all uses and users, it might be useful to evaluate whether experience should determine the
extent and methods of uncertainty communication available for a given user, problem or use.
Any decision problem - the alternatives, consequences and probabilities involved with a
particular decision, governed by the available data and the relative uncertainty of the data
-is framed by the decision maker's concepts associated with a particular course of action
(Tversky and Kahneman, 1981). Thus, the same decision problem can be framed in multiple
ways - either by the same person who may have multiple or changing goals, or by many
different decision makers, each of whom has a different perspective, expertise and conceptual
model about the data and the phenomenon. When faced with decisions under uncertainty,
individuals often revert to heuristics , or abstract mental rules, rather than statistics, to
determine a course of action. In terms of time and information requirements, heuristics serve
to efficiently generate satisfactory outcomes in situations that a decision maker frequently
encounters. The individual learns to apply the heuristics that result in the most favourable
outcomes; these repeatedly used rules reduce the complexity of assessing the alternatives
and potential outcomes. Of course, there is no guarantee that, in any specific instance,
heuristics will always generate the most favourable outcome, or that they are applicable for
new situations or problems (Patt and Zeckhauser, 2000). Because they are used and reused
in different decision problems, if they are incorrect they can result in systematic errors and
bias in decision making (Tversky, 1974; Tversky and Kahneman, 1974).
Individual cognition, therefore, significantly informs the decision framework used to
solve problems. Understanding the extent to which (and the problems for which) heuristics
are used by decision makers when solving problems with uncertain information should
be a consideration in the design of methods for communicating uncertainty. For example,
individuals who, because of their particular decision frameworks, employ decision rules may
be more likely to expect discrete alternatives, or 'scenarios'. Alternatively, others who prefer
statistical explanations of uncertainty may be more likely to expect continuous alternatives,
or 'ranges' of outcomes.
Risk management, uncertainty, and communication
Risk communication is defined as an interactive process of information and opinion ex-
change between individuals, groups, and institutions. These exchanges can include multiple
'risk messages', which are written, verbal or visual statements about a risk, and the expres-
sion of concerns, reactions and opinions about the risk (Patt and Dessai, 2005). The risk
messages are specifically tailored to the parties involved; for example, a risk message does not
necessarily involve statistical probabilities, which may be appropriate for a scientist, but not
for less expert users of the information. Figure 14.2 presents several examples of potential
risk messages.
Effective risk communication involves incorporating the potential cognition conflicts
and knowledge of the users of the information into the risk message. Before scientists or
organizations can effectively communicate risk, there needs to be an awareness of the decision
frames and heuristics that individuals and groups use when evaluating alternatives and
making decisions, as well as the potential actions that could be taken, and the consequences
that may result (Grabill and Simmons, 1998). For example, Patt and Dessai (2005) evaluated
the effectiveness of risk communication techniques for climate change data, based on a