Risk Assessment: The Consumer as an Intuitive Statistician - Innovation and Marketing in the Pharmaceutical Industry - page 289

In the current context, we use the term to signify the individual risk factors,
behaviors, or symptoms that probabilistically contribute to overall risk for an indi-
vidual. The theory of reasoned action (cf. Ajzen and Fishbein 1975 ; for applications
to risk perception, see Fishbein and Middlestadt 1989 ; Fishbein et al. 1994 ; for a
recent meta-analysis, see Albarracin et al. 2001 ) was the first theory applied to
understand how consumers assess risk using what we term the “bottom-up” process.
The risk factors that a consumer could typically use as inputs are their prior behav-
iors (e.g., having unsafe sex in the context of AIDS) and whether or not they have
symptoms that characterize a disease. Of course, individual differences in demo-
graphics (e.g., age, gender, and race), family history (given that certain diseases like
thyroid disorders may have a genetic aspect), and geographical location (e.g., inci-
dence of skin cancer is greater in Australia and New Zealand than in India) would
also factor into their risk estimates. The manner in which individuals use informa-
tion about their behaviors or symptoms to construct their level of risk is covered in
this section. (How individual differences in demographics, family history, and loca-
tion among others affect risk perception is covered in the subsection of conditional
base rates where top-down processes are combined with bottom-up processes to
form an integrated perception of risk).

10.5 Inputs to the Model: Behaviors, Existing Symptoms,

and Feelings

The theory of reasoned action suggests that people make judgments as if they are
performing cognitive algebra, such that they identify a set of attributes, assign weights
to them, and then aggregate across these weighted attributes. In the domain of risk
perception, this suggests a process whereby people identify their risk factors and then
integrate them by assigning weights to each identified behavior or symptom.
The bottom-up approach involves identifying and interpreting causal factors
(e.g., symptoms, behaviors, and situational factors), and then aggregating them to
assess the extent to which a person is at risk. The model follows a signal detection
theory paradigm with the addition that the signal needs to be interpreted prior to
being identified as present.
The model formulation is as follows:

Risk =+ + a Sd ii e

where

i = the specific symptom/behavior for the disease
S = presence of a symptom, risk behavior, or preventive behavior for a disease
δ = diagnosticity of the symptom/behavior for the disease
a = other factors such as demographic and genetic predispositions
ε = random error