The name is absent

M. Wilke et al.: Risk Specificity for Risk-Defusing Operators

The number of questions and statements related to the
three types of RDOs (new alternatives, possibilities for con-
trol, and worst-case plans) served as a dependent variable.
We used a category system (see below) to quantify them.

Material

Short descriptions of each scenario were presented to the
participants. We ensured that the selected scenarios did not
overlap in content. The resulting 12 scenarios and their clas-
sification according to type of risk and risk domain are
shown in Table 1.

The construction of the scenarios followed the examples
provided by Huber (2004; Huber et al., 1997; Huber et al.,
2001). Unlike the published procedure, our scenarios were
constructed in such a way that participants were to make
their decisions on a societal level, that is, not for themselves,
but for many others. This procedure was adopted because
it is implausible to imagine individual damage for the glob-
al type of risk. Each scenario contained a statement that a
negative consequence could result from a selected alterna-
tive, but no exact probabilities were given. The two alter-
natives were presented as follows: A riskier alternative was
presented with a safer alternative, the safer one being con-
nected with further potential negative consequences. Also,
no hints at potential RDOs were given. An example of one
of the scenarios (avian flu pandemic, type of risk: cata-
strophic; domain: health) is provided in the Appendix.

Data Collection

The AIS (Huber et al., 1997) method was used as the fun-
damental technique for information presentation and data
collection. Quasi-realistic scenarios were presented, fol-
lowed by two alternatives as described above. Participants
were allowed to ask as many questions as necessary to make
their decisions. Answers were provided according to a fixed
pattern (see below). Participants made their decision after
having collected enough information. Structuring the situ-
ation is an essential part of the decision-making process. In-
formation is not processed automatically; rather, based on
the search for information, a mental representation of the
situation is created. During the AIS, participants decide for
themselves what kind of information they need. We decid-
ed to use the conversation-based version (C-AIS), which
was developed by Ranyard et al. (1999; see also Williamson,
Ranyard, & Cuthbert, 2000) and is based on Ranyard and
Craig’s (1995) interview techniques. The conversation-
based version can be seen as a modification of the AIS. The
difference between the C-AIS and Huber et al.’s (1997) AIS
lies in the role of the experimenter, who acts as an inter-
viewer and gives verbal (instead of written) answers that
are read from standardized templates. This version has the
advantage of creating a more natural social interaction, but
also disadvantages due to problems in verbalizing and ex-
perimenter effects. To avoid any of the known potential dis-
advantages, the experimenters underwent interviewer train-
ing. The interview was non-directive, that is, conducted
without any evaluative comments or statements. Questions
for which there were no answers in the answer templates
were answered spontaneously depending on the context and
immediately added to the template. In some cases, the ex-
perimenter answered, “There is no information available.”
To further reduce experimenter effects, we followed the rec-
ommendation given by Ranyard et al. (1999), Williamson
et al. (2000), and Huber et al. (2001) to use standardized
statements like “What do you mean exactly?” or “Could
you please pose the question in more concrete terms?” (in
order to elicit a more specific question) and “What are you
thinking about right now?” (to elicit thinking aloud when a
participant had been silent for over 1 min).

Williamson et al. (2000) showed that C-AIS is a well-
suited process-tracing method that, unlike thinking aloud
techniques, does not provoke reactivity. We thus selected it
as our basic method. According to Huber et al. (1997) and
Williamson et al., the C-AIS can be combined with think-
ing aloud, following Ericsson and Simon (1980, 1993). The
thinking aloud instruction asks participants to verbalize all
ideas they have while working on the task, so that their cog-
nitive processes can be assessed afterwards. However, Rus-
so, Johnson, and Stephens (1989) found that simultaneous
verbalization affected cognitive processes and decision ac-
curacy. Further criticism of the thinking aloud technique
concerned the fact that not all cognitive processes can be
verbalized (Harte, Westenberg, & van Someren, 1994;
Westenberg & Koele, 1994). However, if the participant is
given clear instructions to communicate thoughts without
analyzing them, the independence between verbalization
and primary task - that is, the validity of the data - can be
assumed (Ericsson & Simon, 1993). According to Harte et
al., the experimental setting should include warm-up tasks
as well as ensure that enough time is allowed for partici-

Table 1

List of 12 Scenarios Grouped According to Risk Domain and Type of Risk

Swiss J Psychol 67 (1), © 2008 by Verlag Hans Huber, Hogrefe AG, Bern

<<   <   2   3   4   5   6   7   8   >   >>