One suggestion for approaching this issue appears in:
Newhall, C.G. and R.P. Hoblitt, 2002, "Constructing
event trees for volcanic crises", Bulletin of
Volcanology, vol. 64, pp. 3-20.
These authors deserve the credit for the ideas mentioned in this message,
but any errors or misunderstandings are mine.
Table 5 (p. 16) provides the risk of death per year at various ages, from
various occupations, and due to various causes such as smoking, cancer, and
drowning. These risks, expressed as a probability of dying per year, could
then be compared to the risk of death per year from natural hazards, such as
staying in a volcanic risk zone or from a specific weather phenomenon. The
public or decision maker would have a straightforward comparison to
something more familiar than an abstract number.
(As well, Chris Newhall mentioned to me the book "Risk-Benefit Analysis",
2001, by R. Wilson and E.A.C. Crouch, Harvard Press, 2nd ed. which has more
complete tables of the risk of dying per year.)
The paper also discusses aspects of communicating the uncertainty of the
risk estimates and different tolerances for such uncertainties.
Issues of concern with this approach are noted, including:
-The numbers for comparison should represent the
audience in terms of being appropriate for the
location, culture, and types of familiar activities.
-Compared risks must be similar. Comparing involuntary
risks with voluntary risks may be inappropriate.
-Different people in the same situation have different
tolerances for the same (perceived or actual) risk.
-"Some individuals will conclude that if their personal
risk from the volcano is no higher than another
acceptable risk, then the volcanic risk is also
acceptable. Others might prefer to add the volcanic
risk to their other risks, and then decide whether the
new cumulative risk is acceptable" (p. 17).
To give a practical example, we may claim that the risk of death is immense,
being the same as that from smoking two packs of cigarettes a day. The
person whom we are trying to convince to evacuate (or to issue an evacuation
order or to enact a mitigation strategy) could respond: "Oh, no problem,
then. I smoke three packs a day". Similarly, we may indicate that the risk
is so dangerous that it is equivalent to being a helicopter pilot, only to
met with the response "My mother, my son, my neighbour, and I are all
helicopter pilots and we have never had a problem".
By making the risk too familiar, we may lose the power and fear of extreme
risk. On the other hand, is that worse than creating confusion with an
obscure number without much meaning? Is our goal to recommend an action and
to convince people to take that action or to ensure that the decision-makers
understand as much information as possible for them to make the decision?
And if we do not wish to recommend, but that is requested or demanded?
I look forward to anyone's further thoughts or sources,
Ilan
_________________________________________________________________
STOP MORE SPAM with the new MSN 8 and get 2 months FREE*
http://join.msn.com/?page=features/junkmail
|