Learning before incidents: What we can learn from understanding variation in performance

We tend to define safety as ‘a lack of incidents’. When incidents occur, we assume people and processes performed poorly or differently compared to the times when incidents didn’t occur. What if that assumption is wrong?

This video explores concepts like work as done vs. work as imagined, and variation in performance. It explains how large variations in performance naturally lead to undesirable outcomes.

Rather than measuring the outcomes, if we could instead measure the variation in performance directly, we might be able to predict where incidents are likely to occur. We could also address the causes of this variation before incidents occur.