Understanding accidents-from root causes to performance variability

When an accident happens, it is obviously important to understand what caused it in order to take effective preventive measures. Accident analysis always implies an accident model, i.e., a set of assumptions of what the underlying "mechanisms" are. Over the last 50-75 years there have been significant changes in accident models, leading to changes in the methods and goals of accident analysis. In parallel to this development the understanding of the role of humans in accidents, and of the nature of "human error", has also changed. This paper provides an overview of the developments, and outlines the consequences for contemporary accident analysis and prevention.