Risk Assessment: The Consumer as an Intuitive Statistician - Innovation and Marketing in the Pharmaceutical Industry - page 287

improve the likelihood that the correct inputs are used and those inputs are correctly
integrated to form a judgment. For example, if consumers use a top-down strategy,
then ways in which the base rate is constructed or framed will affect its likelihood
of being appropriately used. On the other hand, if consumers used a bottom-up
strategy, changing the construction or frame of the base rate would have limited
effectiveness, and a more effective strategy would be to strength the link in the con-
sumer's mind between a symptom and a disease. This would increase the symp-
tom's likelihood of being identified and appropriately integrated into the risk
judgment. Thus, the effectiveness of any de-biasing strategy will differ depending
on the biasing input used. Accordingly, understanding whether top-down or bottom-
up methods or a combination of both is used to construct risk is necessary to appro-
priately de-bias consumers and bring their risk estimates in line with reality.
But do errors in risk estimation really matter? Why is it important to bring risk
estimates in line with reality? Underestimating one's risk brings with it the conse-
quences of not seeking or taking treatment, or not engaging in preventative behaviors
(for a review, see Brewer et al. 2007 ). This underestimate of risk has downstream
consequences not only for the individual in terms of longevity and quality of life but
also for their family and social circle, their work place as it can impede their effi-
ciency and effectiveness and lower their overall productivity, lead to higher costs of
treatment at a later stage, and be a cost to the medical infrastructure as well as the
economy and society as a whole. On the other hand, overestimating risk could lead
to consumers behaving like hypochondriacs, seeking and taking more treatment
than they need, purchasing costlier treatments than are warranted, exploiting scarce
medical professional time, being anxious and less productive at work, and concomi-
tantly being a social and economic burden. Thus, the costs of over- and underesti-
mating risk both have consequences that trickle down from the individual consumer
level to society at large.
One of the most robust biases that have been documented in the risk perception
literature is that people have a self-positivity or unrealistic optimism bias such that
they believe negative things, like getting a disease, are less likely to happen to them
than to others (Perloff and Fetzer 1986 ; Raghubir and Menon 1998 ; Weinstein 1980 ,
1987 ). However, there have also been documented cases when people overestimate
risk and are pessimistic (Keller et al. 2002 ; Lin et al. 2003 ). Keller et al. ( 2002 )
showed self-negativity for depressed women's estimates of their relative chances of
getting breast cancer: while depressives' absolute estimates of risk reflected pessi-
mism (vs. the actual base rates), their relative estimates of self vs. another person
reflected self-negativity. Lin et al. ( 2003 ) developed a framework to reconcile
Perloff and Fetzer's ( 1986 ) finding of self-positivity with Keller et al.'s ( 2002 ) find-
ing of self-negativity. They explored the connection between absolute levels of risk
estimates (defining optimism and pessimism) and relative estimates of self-others'
risk (that define self-positivity, self-negativity, or realism). They found that opti-
mists (with low absolute levels of risk) demonstrated self-positivity (low relative
levels of risk). The bias was robust and remained when base rates were provided:
risk estimates were not updated for events that optimists believed were controllable,
and were inadequately updated for events that were believed to be less controllable.