Communication initiatives frequently are used to inform people about various risks in an attempt to improve their decision making and behavior. People, however, do not readily accept personal risk information, often underestimating their susceptibility to and the severity of the risks. This project is designed to learn more about perceptions of health risks with the ultimate goal of designing more effective messages for health promotion and disease prevention.

The research involves three experiments to learn whether people automatically underestimate their health risks, that is, without awareness, intention, or effort. The first experiment explores whether participants subconsciously primed with self-identity words (e.g., ?I,? ?me?) respond faster than participants primed with third-person (e.g., ?he,? ?herself?) or neutral (e.g., ?it,? ?itself?) words. The second experiment explores whether participants primed with self-identity words will exhibit faster response times when they are mentally taxed (i.e., memorizing an eight-digit number) compared to when they are not (i.e., memorizing a two-digit number). The third experiment will explore whether highly reactant participants (i.e., those who react negatively to restrictions on personal choice) when primed with self-identity words will exhibit faster response times in response to health conditions that are perceived to be highly controllable (i.e., conditions that are preventable if personal action is taken). In all three experiments, the research team will also examine several individual difference variables (e.g., self efficacy, dispositional optimism) and see if characteristics of the health risk (e.g., prevalence) influence risk perceptions.

Project Report

Communication campaigns frequently are used to inform people about various risks in an attempt to improve decision making and behavior. People, however, often underestimate the possibility of experiencing a health risk (i.e., susceptibility) and its seriousness (i.e., severity). This dissertation draws on psychological theory and research to peer into the black box –people’s minds – to uncover the processes involved in how people arrive at personal risk judgments. People attend to personal health risk information in light of their self-schemas (i.e., the knowledge structure that contains all information about oneself). Because the self is almost always regarded in a positive light, health risk messages are deemed inconsistent with the active self-schema leading a person to underestimate her heath risk. These judgments are frequently made, which qualifies them to migrate from consciousness to unconsciousness such that people arrive at these judgments with little to no demand on processing resources. They become automatic. This dissertation, thus, investigates whether people are predisposed to underestimate personal health risks by examining the automatic nature of health risk perceptions. Two pilot studies and four experiments tested two theoretical propositions and five derived hypotheses. The experiments relied on priming to activate the self-schema and measured the effects of this activation on health risk perceptions. Participants were primed with self-identity words (e.g., I, me) and then were asked to make judgments about their personal susceptibility to and the severity of health risks. Based on the premise that any mental process takes time, and that fast reaction times reflect more accessible, automatic, less thoughtful, and efficient processes, reaction time (RT) was assessed. RT was defined as the time in milliseconds that elapses between a participant’s exposure to a health condition (e.g., cancer) on a computer screen and her response (yes/no) using pre-assigned buttons on a buttons box. Controlling for several individual differences (e.g., self-efficacy, dispositional optimism) and health risk characteristics (e.g., prevalence), the primary hypothesis was that participants primed with self-identity words would underestimate their health risks as evidenced by fast reaction times responding no to negative health conditions (e.g., flu) and yes to positive health conditions (e.g., healthy gums) compared to those primed with other (e.g., he) or neutral (e.g., it) words. Results provide evidence that there are defaults in people’s minds when it comes to health risk perceptions. The defaults are: (a) We are likely to experience positive health conditions and unlikely to experience negative ones. (b) We are likely to acknowledge the severity of rare health conditions (e.g., leukemia) but dismiss the likelihood of experiencing them. In contrast, we acknowledge the likelihood of experiencing highly prevalent/common conditions (e.g., flu) but dismiss them as not severe. These judgments about personal risk are considered defaults because participants made these decisions repeatedly with little to no demand on processing resources, a fact attributed to the automaticity of these judgments. This dissertation is one of the first to make the case for the automaticity of risk perceptions. It contributes to the health communication field because it: (a) challenges current behavior change theories that assume rational decision making and that risk perception is a combination of susceptibility and severity rather than two independent concepts; (b) explains the processes involved in judgments of risk and presents a theoretical model that describes how people arrive at biased risk perceptions; (c) uses RT (i.e., an implicit measure) as an index of the mental processes involved in judgments of risk; and (d) has implications for other risk domains (e.g., environmental risks) and effective health messaging.

National Science Foundation (NSF)
Division of Social and Economic Sciences (SES)
Standard Grant (Standard)
Application #
Program Officer
Robert E. O'Connor
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of North Carolina Chapel Hill
Chapel Hill
United States
Zip Code