Should the Emphasis of Public Health Messages Shift Based On Topic? A Relative Importance Analysis

Theoretical Background and research questions/hypothesis:  Through social interaction and communication, perceptions of risks can be amplified and attenuated (Kasperson & Kasperson, 1996; Renn, 1991; Renn, Burns, Kasperson, & Kasperson, 1992).  One framework often used to explain the subjective nature of risk perceptions is the psychometric paradigm.  This framework predicts that characteristics like dread, newness, and potential for catastrophic effects vary across risks in measurable and predictable ways (Slovic, 1987).  Furthermore, these characteristics comprise two factors, along which risks can be plotted.  It would be particularly useful to know if risks existing in one quadrant require different risk communication strategies than risks in another quadrant.  The psychometric paradigm does not provide guidance as to how the qualitative factors of risks might be influenced through communication, or whether such communication could move a risk into a quadrant that may be more easily targeted in a public health campaign. RQ₁:  Can emotional appeals affect evaluations of the risk characteristics identified in previous research as fundamental to risk perceptions? RQ_2a:  Which risk characteristics have the greatest impact on risk perceptions? RQ_2b:  Does the impact of particular risk characteristics vary by type of risk being evaluated?

Methods:  Non-probability sampling was utilized to recruit 560 adult participants to complete an online survey.  Those who agreed to participate were randomly assigned to a message condition--either a distracted driving message targeting anger, a distracted driving message targeting fear, a tritiated water message targeting anger, or a tritiated water message targeting fear.  After viewing the message, participants completed a brief questionnaire assessing emotional responses, risk perceptions, and evaluations of various risk characteristics.

Results:  Univariate ANOVAs indicated perceptions of the risk of traffic accidents were more effectively influenced by the message induction.  Interestingly, traffic accident risk perceptions were affected by both distracted driving messages and tritiated water messages.  Next, relative importance analyses were conducted to determine the unique contributions of each of the risk characteristics and expressed emotion on risk perceptions.  The results were quite different across risks.  Nuclear energy risk perceptions were largely driven by expressed fear, with evaluations of the possibility of fatal consequences and estimations of catastrophic potential also strong predictors.  Traffic accident risk perceptions were also most strongly predicted by expressed fear, but effects were more evenly distributed across predictors. 

Conclusions:  Risk characteristics identified in previous research as fundamental to risk perceptions can be influenced by risk communication efforts.  The extent of the influence varied by risk as did the importance of the various risk characteristics.  Together, these results suggest there are specific characteristics for different types of risks that are likely to be most fruitful in targeting change.

Implications for research and/or practice:  Theoretically, the data suggest the psychometric paradigm framework could benefit from the incorporation of messaging as a predictor of risk perceptions and evaluations of risk characteristics.  Practically, the data suggest health communication efforts could benefit from the strategic selection of risk characteristics to target, based on the type of risk at hand.