Thinking about risk is hard and complicated. Just thinking about some risks can make you feel afraid, and that feeling probably influences your assessment of the risk. I've also read somewhere lately (no cite) that we overestimate the likelihood of events that are widely publicized, that we have paid a lot of attention to.
Mind Hacks reports on a study about influences on risk perception in daily life. Participants were given pagers and randomly asked for in-the-moment assessments of risk and predictions of consequences; later some were asked to review the same situations and re-assess the risk, after having experienced the event and its consequences.
Generally, risks were perceived to be short term in nature and involved "loss of time or materials" related to work and "physical damage".
Interestingly, everyone rated the severity of risk as about the same, but women were more likely to think that the worst consequence was likely to occur.
Furthermore, the better the mood of the participants (both male and female), the less risky they thought their activity was.
Most of that doesn't surprise me: I think it would be hard to fight the tendency for mood to influence perspective, and focusing on the now after an abrupt interruption (the pager going off) would also focus perspective on short-term issues. Women thinking the worst was likely certainly conforms to my experience (I do it); I wonder about larger implications on subjects like confidence, courage, and exercising choice. For example, since a woman expects a worse outcome, will she generally choose options with the least severe worst outcome? Should we count women's courage higher than men's for the same choices, since the women are choosing to risk what they believe to be a worse consequence?
The risk assessments from after-the-fact tended to be much lower, showing that people tend to judge things to be more risky 'in the heat of the moment'. Both of these findings demonstrate the importance of emotion in risk judgements, suggesting that it forms another source of information, along with more calculated rational estimates.
There's a lot of thinking meat here for people trying to understand and change their own risk assessment process, or to train others to respond during episodes of risk.