Risk perception is the extent to which an individual feels that they are at risk of developing a health problem. Risk perception is often related to family experience.  If no one in your family has ever had breast cancer, you may feel like you are at low risk to get it. If your grandfather was overweight and never had any heart problems, you may not be worried about the fact that you have put on a few kilos (or pounds) too many.

Psychologists suggest that one of the reasons that we have this irrational perception of risk is due to optimism bias. We tend to overestimate the number of our health-protective behaviours and underestimate the significance of our health-impairing habits.

Optimism bias means that people have a tendency to judge their own susceptibility to a health problem as lower than the risk of the same problem for others. One explanation for optimism bias is that it reduces the anxiety that would be generated by admitting to personal vulnerability. With this theory, the optimism bias increases when a disease is considered serious. In addition, people tend to make predictions about the future based on earlier experiences, e.g. ‘I haven't had skin cancer; therefore, I won't get it in the future either’. Health psychologists postulate that higher levels of optimism bias make it less likely for an individual to change their health behaviours.

Weinstein carried out a study of 88 undergraduate psychology students (32 males and 56 females) to test their level of optimism bias. Most of the participants did not smoke at all, had no family members with diabetes and did not drink alone.

He asked participants to fill out a questionnaire about eleven health and safety risks: diabetes, heart attack, alcoholism, suicide, lung cancer, other forms of cancer, mugging, injury in an auto accident, high blood pressure, tooth decay, and ulcers.

The participants were asked to answer the following question: Compared to other students of my sex, my chances of developing … are: much below average, below average, slightly below average, average, slightly above average, above average, much above average ( - 3 to +3)  In addition, they were asked their level of worry about the chance of developing the health problem and their likelihood of adopting precautions.

The study was experimental.  There were three conditions in which the participants answered the questionnaire.

• In the control group (n = 31), participants were simply asked to simply fill out the questionnaire.
• In the "own-risk" group [OR] (n = 29), participants were told that they would be asked about “various factors that influence the risk of health.  In some cases these risk factors are well-established; in other cases, it’s not clear and they are only possible risk factors.”
• In the "informational" group (n = 28), participants were shown the list of risk factors and told, "Each question has a red arrow or number which shows the response of a typical Rutgers male or female based on the data we gathered earlier this semester."

The results were as follows:

• The control group had a high level of optimism bias, under-estimating their level of risk. They also had the highest level of worry; however, interest in prevention was low. 
• The OR group was the least worried and the least interested in precautions. This was an increase in optimism bias.
• The IR group had the most realistic estimates of their level of risk. There was no significant pattern of worry or attitude toward prevention.

The results showed that the "own risk" group felt that they had least chance of developing the health problem and were the least worried of the three groups. An unexpected finding was that merely rating oneself on risk factors substantially increased optimistic biases. The results also showed that when participants were given information about the risk status of their peers, the optimistic bias was significantly reduced.

Merely rating oneself on risk factors substantially increased optimism biases. Ironically, this study shows that discussions of health and safety risk factors may actually make people more unrealistic. The results also showed that when participants were given information about the risk status of their peers, the optimistic bias was significantly reduced.

The study is a well-controlled experiment, leading to high internal validity. We can assume that the IV had a direct effect on the DV.

As the study was highly standardized, it may be easily replicated in order to establish the reliability of the data.

The data was self-reported, so demand characteristics may have played a role in the study. 

The sample was mostly made up of students who did not smoke and did not drink alone.  This may not be representative of the larger student body.  Therefore, it may not be possible to generalize the findings to university students.  In addition, they were students of psychology, so there may have been a tendency toward expectancy effects.

Since the study was done under controlled conditions, the students may have seen this as only a hypothetical or academic exercise. Being primed to think about their own health - and the family history of health problems - may have changed the findings.  Cognitive biases have the greatest effect when we do not have enough time or information in order to make a decision.

The study does not account for the disposition or previous experience/education of the sample. These factors, rather than one's level of optimism bias, may have accounted for the results of the study.