Two nurses, same hospital, opposite worlds
Amy Edmondson went in expecting to find which teams made fewer errors. What she found instead changed how organizations think about mistakes.
In the 1990s, Amy Edmondson was studying nursing teams across two hospitals when the data refused to behave. The teams with the most experienced nurse managers -- the teams considered highest-performing -- were reporting more medication errors, not fewer.
Edmondson's first instinct was that something was wrong with the methodology. But when she interviewed nurses on different wards, the numbers made sense. Two nurses described the same hospital from what felt like different planets.
In one team, a nurse explained plainly: the drugs are toxic, so you're never afraid to tell the Nurse Manager about a mistake. In another ward, her colleague described being called into the office and made to feel accused. You get put on trial, she said.
The better-performing teams weren't making more errors. They were reporting more errors, because their members felt safe enough to say so out loud. Lower-performing teams had just as many mistakes -- possibly more -- but those mistakes stayed buried.
Edmondson named the variable doing the work: team psychological safety. Her 1999 paper in Administrative Science Quarterly defined it as a shared belief that the team is safe for interpersonal risk-taking. Where it was high, people asked questions, flagged problems, and learned from what went wrong. Where it was low, people protected themselves and let problems fester.
Silence in a team isn't safety. It's often the exact opposite.