30
M. Wilke et al.: Risk Specificity for Risk-Defusing Operators
1. New alternatives: search for alternatives to the options
presented, whereby the alternatives share the positive as-
pects - but not the negative consequences - of the op-
tions presented (e.g., using biological instead of chemi-
cal pest control).
2. Possibilities for control: check whether the risk of neg-
ative consequences can be reduced by either preventing
the negative event from happening or breaking the causal
link between the event and its negative consequences
(e.g., wearing protective clothing at dangerous work-
places).
3. Worst-case plans: search for action plans that would neu-
tralize negative consequences should they occur (e.g.,
backing up important files or buying insurance).
These RDOs are indicators of the cognitive process of de-
cision making. For worst-case plans, besides the mental ef-
fort involved, additional costs arise, for example, time
(backing up the data) or money (paying for insurance). The
RDO new alternatives differs from the other two operators
because it characterizes risk management in terms of an en-
largement of the problem space. The other two RDO vari-
ants can be seen as active risk management. Huber et al.
(2001, p. 411) comment: “Thus, risk defusing can be at-
tempted in two ways: (a) by reducing the probability of the
negative outcome, and (b) by transforming the negative out-
come into an outcome state with a higher utility. Both ways
increase the expected utility of the alternative.” Subsequent
articles by these authors only refer to the two general strate-
gies of pre- and post-RDOs, which contribute to the pre-
vention of or compensation for the damage (e.g., Huber,
2004; Huber & Huber, 2003). Our study, however, will use
the differentiation explained above.
To analyze decision-making behavior in everyday risk
situations, Huber, Wider, and Huber (1997) used the method
of active information search (AIS). In this procedure, par-
ticipants can pose questions to the experimenter to receive
more information about two alternative actions in a written
quasi-realistic risky scenario. A description of the method
and empirical results from its application can be found in
Huber (1997, 2004), Huber et al. (1997), Huber et al. (2001),
as well as in Huber and Macho (2001). Because we used
this technique in our own study, a short description can al-
so be found below.
In classical decision-making research, the subjective
value of consequences and the subjective probability of their
occurrence are essential for the decision maker (Edwards,
1954, 1961). This position differs from the assumptions
about the decision-making process in the quasi-realistic
risky situations used by Huber (2004). In Huber’s situations,
a comparison of participants’ decision-making behavior
when they are provided with all the information they need
(e.g., in the lottery experiments in classical decision-mak-
ing research) with that when they are to request additional
information from the experimenter (as in the AIS method)
shows that most decision makers do not request informa-
tion concerning probabilities. Huber and colleagues con-
cluded that classical decision-making tasks induce the use
of probability information by the way the information is
presented. In quasi-realistic tasks, the information about po-
tential negative consequences suffices. Furthermore, the
majority of decision makers even search for additional ac-
tions which could reduce the probability of the incidence
of risky events (Huber, 1997; Huber et al., 1997; Huber et
al., 2001; Huber & Macho, 2001). Active risk-defusing be-
havior seems to play the more important role in decision
making.
The question concerning the conditions under which
which type of RDO will be used is still open (Huber, 2004).
Studies have demonstrated substantial differences in the
number of questions posed, that is, in the frequency of
searching for different RDO types in different scenarios
(e.g., Huber et al., 1997; Huber et al., 2001). There is a clear
need for the development of a theory that explains decision-
making behavior in different situations, that is, why deci-
sion makers show different decision-making behavior in
different types of situations. A typology of risky situations
would be helpful: It would provide us with a formal char-
acterization of task types as well as a comparison of deci-
sion-making behavior in different types of situations so that
we could determine whether risk-defusing behavior is a
universal feature of decision making under uncertainty. To
this end, the first step is the classification of risks (see pi-
lot study below).
The literature suggests the existence of domain-specific
differences in decision-making behavior. Huber and Macho
(2001), for example, showed that more questions concern-
ing probability were asked in the domain of medicine than
in the domains of social issues and business. This is in ac-
cordance with results concerning naturalistic decision mak-
ing about domain-specific strategies used in the decision-
making process (Payne et al., 1993; Pennington & Hastie,
1993; Rettinger & Hastie, 2001; Schmitt & Klein, 1998).
But research on such differences in the RDO search is still
lacking. Therefore, we want to study this question more
closely and look for differences in the search for RDOs de-
pending on risk domain and type of risk. Our intention is to
find out how strongly those factors affect the search for dif-
ferent RDO types.
Pilot Study
In order to investigate the search for RDOs in different types
of risky situations, we needed a formal classification of
risks. We started with the risk typology of the German Ad-
visory Council on Global Change (Wissenschaftlicher
Beirat der Bundesregierung Globale Umweltrisiken
[WBGU], 1999). This typology, formulated by experts,
characterizes the threat of risks using several criteria: prob-
ability of occurrence, extent of damage, certainty of as-
sessment, persistency, irreversibility, and delay effect. We
were interested in how non-experts would rate different
risks using these criteria and which criteria they would con-
Swiss J Psychol 67 (1), © 2008 by Verlag Hans Huber, Hogrefe AG, Bern