The best teams reported more mistakes, not fewer
A hospital study meant to count medication errors accidentally uncovered why some teams learn and others just hide.
In the late 1990s, a young researcher named Amy Edmondson was measuring error rates across hospital nursing units, expecting the highest-performing teams to make the fewest mistakes. Her numbers said the opposite: the units rated best by outside observers had higher reported error rates than the ones rated worst.
The explanation took her research in a new direction. The good teams weren't sloppier — they were more willing to say out loud when something had gone wrong, because their culture didn't punish the admission. The weaker teams had plenty of errors too; they just didn't surface, because nobody wanted to be the one who spoke up.
Edmondson named the missing ingredient psychological safety: not a feeling of comfort, but a shared confidence that raising a hand won't get you humiliated. Two decades and over a thousand follow-up studies later, the finding holds up across hospitals, cockpits, and boardrooms: safety is measurable, buildable, and has nothing to do with being nice.
A team that never surfaces its mistakes isn't doing better. It's just further from finding out.