> Crucially, the researchers showed all respondents information about the number of deaths that had historically occurred in the worst accident associated with

> Crucially, the researchers showed all respondents information about the number of deaths that had historically occurred in the worst accident associated with the technology. This is important for nuclear power, since accidents are rare but can have dire consequences if they do occur. This seems like a flawed study. It implies that previous nuclear accidents are sufficient data to form a precise “statistical risk” level. And if people don’t decide based on those statistics alone they must be irrational and just “anxious”. That is not a scientific conclusion. Nuclear accidents are rare, but potentially extreme. Just one more major accident, or the first case of terrorism, could significantly change the statistics. Thus It is possible that people are being rational when they recognize that we have too few data points to form a precise risk model for nuclear power.