Risk Assessment: The Consumer as an Intuitive Statistician - Innovation and Marketing in the Pharmaceutical Industry - page 304

symptoms are not unlike the ambiguous symptoms of depression. Type II diabetes
is characterized by “feeling tired or ill, frequent urination (especially at night),
unusual thirst, weight loss, blurred vision, frequent infections, and slow healing
of sores. The symptoms of type 2 diabetes develop gradually and are not as notice-
able as in type 1 diabetes” ( http://my.webmd.com/content/article ) . Notably, the
list omits “tingling hands and feet,” a symptom that is less ambiguous and shares
many of the characteristics of the “thoughts of suicide/death” symptom in the
depression inventory (i.e., it has high causal clarity, low frequency, is a present/
absent event rather than a state, etc.). This example also highlights the intriguing
possibility that the interpretation of the symptoms may itself be contingent on
actual demographics (or feelings or base rates) that are the α of the bottom-up
process.
Further, it is possible that the detection potential of a signal is contextually deter-
mined. It is a function of the other signals that surround it as well as its own innate
ability to predict. The concept that signals vary as a function of their degree of exis-
tence is conceptually similar to STD's “threshold of detectability.” The degree of
existence of a symptom is an innate aspect of a signal, but the manner in which it is
perceived to be informative is again contextually determined. The consistency
between an event occurring and how frequently it is expected to occur affects the
perceived strength of a signal. But, as the expectation of occurrence is itself a func-
tion of whether the event is a state of being or a specific event, specific events are
perceived to be stronger signals than are states of being.

10.14.2 Interactive Effects of Proximity

The effect of personal relevance on risk perceptions is nuanced. Prior research on
persuasion suggests that personal relevance of the risk might activate a variety of
defense mechanisms to protect the self when the individual's safety and health are
threatened (Brown and Smith 2007 ; Kiviniemi and Rothman 2006 ; Leffingwell
et al. 2007 ; Liberman and Chaiken 1992 ). For example, Freeman et al. ( 2001 )
found that smokers exposed to antismoking videos responded with a significant
amount of defensive processing, including information derogation. Therefore,
greater proximity may not always lead to perceptions of higher risk and may in
fact lead to the well documented inverse U-shaped function where risk percep-
tions increase as the level of proximity increases and then reduce as it becomes
too close for comfort.
The effect of geographic proximity may be contingent on social proximity, and
other dimensions of proximity may also interact. Specifically, when geographical
proximity to the origin of the risk is high, consumers perceive a greater risk to them-
selves (Fischhoff et al. 2003 ). When the risk is close by, the closer its social origin,
the higher should be perceptions of risk. However, when the risk is geographically
distant, then unless the risk is high at an absolute level and socially proximate, it
would be difficult for people to imagine themselves at risk.