M. Wilke et al.: Risk Specificity for Risk-Defusing Operators
37
Table 3
Number of People Asking at Least One Question for Each Category and Risk Type
Type of risk
|
Category |
Norm |
med |
cat |
_______glob |
χ2(3) value |
|
Background knowledge / experience |
18 |
24 |
16 |
21 |
1.86 |
|
Attitude / rules / principles |
5 |
12 |
13 |
25 |
15.04* |
|
Situation |
82 |
85 |
82 |
72 |
1.21 |
|
Probability |
53 |
45 |
48 |
20 |
15.64* |
|
Negative consequences |
82 |
81 |
76 |
84 |
0.43 |
|
Positive consequences |
62 |
42 |
38 |
59 |
8.61* |
Note. *p < 0.05. Type of risk: norm = normal, med = medium, cat = catastrophic, glob = global.
an equally distributed exploration of the different RDO op-
tions, but decision makers, as for the medium type of risk,
looked more intensively for new alternatives, in other
words, they went beyond the given frame of reference. It
might be that the normal type of risk with its combination
of classification criteria does not differ enough from the
medium type of risk. For the medium type of risk, most
questions and statements were about new alternatives. We
are not sure if this result can be interpreted as a phenome-
non specific for the medium type of risk because according
to the logit analysis, the catastrophic and global types of
risk also had a similar influence on new alternatives. Ac-
cording to Huber et al. (2001), new alternatives is the most
prominent risk-handling behavior. Therefore, our reported
result may be less due to the influence of the medium type
of risk but more a consequence of this RDO working as a
“general-purpose strategy.”
For the catastrophic type of risk, logit analysis shows a
significant positive influence on worst-case plans, but at the
same time the influence on new alternatives is stronger. In
comparison to other types of risk, worst-case plans are the
most frequently mentioned here. With a high extent of dam-
age and short delay of effects, decision makers seem to want
to reduce the consequences of damage beforehand. Instead
of possibilities for control, which is used least frequently,
new alternatives (the general-purpose strategy) is searched
for. It might be that worst-case plans are searched for in cas-
es in which no adequately evaluated new alternatives are
available. Participants may have tried to generate at least a
plan for the catastrophe. This explanation supports our as-
sumption that worst-case plans might be one way of risk-
defusing in catastrophic risks.
For the global type of risk, the results of the logit analy-
sis follow our expectations that possibilities for control is
the most frequently used RDO. No other type of risk had
such a strong influence on it. Why is the global type of risk
so strongly related to it? First, one can assume that the con-
trol of global risks is the only “correct” risk-specific pref-
erence (a normative solution). Second, it could be caused
by the construction of the scenarios. In our pilot study, we
found that global risks were evaluated most coherently ac-
cording to the evaluation criteria. It could be that global risk
is the most homogenous type of risk in the present study.
Altogether, there are obvious aspects of risk type specifici-
ty, but other factors (risk domain, interaction) were also in-
fluential.
Consequences for Active Risk-Defusing
Behavior
The results for the influence of risk domains and the inter-
action between risk domains and types of risk demonstrate
strong effects on the number of questions and statements
concerning different RDOs. We interpret these results as
support for domain-specific effects. Those effects are in line
with the results reported by Huber et al. (2001), Huber and
Macho (2003), and Payne et al. (1993) on risk domain-spe-
cific decision behavior.
The main difference between the present study and oth-
ers can be seen in a comparison of the effects of different
factors on RDOs. The interaction between type of risk and
risk domain is especially important for the frequent use of
new alternatives and possibilities for control because it de-
pends on the specific scenario. Most worst-case plans are
found in the risk domain of ecology, however. The connec-
tion between RDO search processes and clearly defined
types of risk allows a specification and extension of the ap-
proach to RDOs: The scope of this approach could be ex-
tended from everyday risk up to more global risks used in
our scenarios.
Exploratory Research Questions
The exploratory questions were analyzed with respect to
risk-specific frequency distributions. The results allow the
following conclusions: Attitudes, values, and principles are
mentioned significantly more often for global risks. This is
a plausible procedure: The evaluation of global risks, which
cannot be assessed for long-term effects, relies more strong-
ly on general individual attitudes, principles, and rules than
any other type of risk. For questions from the situation cat-
egory, there are no significant differences between the types
of risk. This implies that, for all four types of risk, situation
influences will be checked in a similar way. Nearly all par-
Swiss J Psychol 67 (1), © 2008 by Verlag Hans Huber, Hogrefe AG, Bern