By Amy C. Edmondson
Learning to fail productively
Subject
Themes
A PhD student at Harvard in 1993 runs a statistical analysis expecting to find that better hospital teams make fewer medication errors. The opposite appears: teams rated highly for teamwork report more errors, not fewer. Shame and self-doubt follow, but pushing through the anxiety produces a pivotal question. What if better teams do not make more mistakes but simply feel safe enough to report them? A double-blind research assistant confirms dramatic differences in openness across hospital units. In some, nurses speak freely about errors; in others, mistakes mean punishment and humiliation. The correlation between openness and reported error rates is nearly perfect. This insight becomes the foundation of psychological safety, a concept now supported by over a thousand research papers showing that teams where people feel safe to speak up perform better, experience less burnout, and in healthcare settings, see lower patient mortality.
A PhD student at Harvard in 1993 runs a statistical analysis expecting to find that better hospital teams make fewer medication errors. The opposite appears: teams rated highly for teamwork report more errors, not fewer. Shame and self-doubt follow, but pushing through the anxiety produces a pivotal…
Published in 2023, Right Kind of Wrong arrives at a moment when the rhetoric of embracing failure has saturated business culture, entrepreneurial communities, and popular self-help literature. Silicon Valley's mantra of "fail fast, fail often" has become a cliche, repeated at conferences and in…
Not all failures are created equal, yet most people treat them as though they are. The instinct to lump every setback into a single category of "things that went wrong" obscures crucial differences in cause, context, and value. A taxonomy that distinguishes between intelligent, basic, and complex…
Better hospital teams appeared to make more medication errors. That counterintuitive finding, which nearly derailed a young researcher's career, turned out to contain one of the most important insights in modern organisational science: better teams do not make more mistakes; they report more. The…
The human mind is remarkably poorly equipped to learn from failure. Several deeply embedded cognitive biases distort perception, and an automatic emotional response, shame, actively shuts down the reflective thinking that learning requires. Understanding these internal obstacles is not merely…
When something goes wrong, the first instinct in most organisations and cultures is to find someone to blame. A nurse administers the wrong medication; fire her. A bank employee transfers $900 million instead of $8 million; discipline the team. Thousands of Wells Fargo staff create fake accounts;…
Pilots train in flight simulators that look and feel like real cockpits. Athletes prepare through scrimmages where the stakes are low but the learning is high. Scientists run small-scale experiments before committing to large studies. In each case, the principle is the same: create environments…
On 6 April 1951, cardiac surgeon Dr. Clarence Dennis wheeled five-year-old Patty Anderson into a state-of-the-art operating room at University Hospital in Minnesota. Dennis had spent years developing a heart-lung bypass machine, a contraption so complex it required sixteen people to operate. Its…
Dr. Jennifer Heemstra runs a DNA research laboratory at Emory University where experiments fail roughly 95 percent of the time. On a summer Zoom call in 2021, her hands fly up, palms open, as she declares, "Failure is a part of science rather than a judgment." She tells her lab students on their…
On 11 August 2020, three Citibank employees accidentally transferred $900 million instead of the $8 million that was owed to companies managing a loan for Revlon. The senior manager who approved the wire transfer failed to check all the boxes necessary to override an automatic default mode in the…
Captain Pastrengo Rugiati was a congenial man who loved the sea. When one of his officers received news of a child's birth, Rugiati would order the ship's whistle blown, fly a coloured ribbon from the mast, and throw a party for the off-duty crew. That Friday night in March 1967, he lingered on deck…
In 1982, Ray Dalio was thirty-three years old, armed with a Harvard MBA, and riding seven years of phenomenal returns at Bridgewater Associates, the investment firm he had founded at twenty-six. He was a frequent guest on national business news programmes, prized for his ability to predict long-term…
A seven-by-ten-foot rug lies on the floor of a large room, divided into a grid of nine rows and six columns. Each square will either beep loudly or remain silent when stepped on. The task is straightforward: find a path of contiguous silent squares from one end to the other. There is no way to…
Spencer Silver was trying to develop a powerful adhesive for aircraft construction at 3M's central research laboratory near Minneapolis in 1968. Experimenting one day, he used more than the recommended amount of a chemical reactant and was astonished to find he had created a thin, weak substance…
Barbe-Nicole Ponsardin Clicquot was widowed at twenty-seven in early nineteenth-century France. Born in Reims in 1777, she lived in an era when women could not own property, sign contracts, or make household financial decisions without a husband's consent. Widowhood, paradoxically, granted her the…
By Amy C. Edmondson
Subject
Themes
A PhD student at Harvard in 1993 runs a statistical analysis expecting to find that better hospital teams make fewer medication errors. The opposite appears: teams rated highly for teamwork report more errors, not fewer. Shame and self-doubt follow, but pushing through the anxiety produces a pivotal question. What if better teams do not make more mistakes but simply feel safe enough to report them? A double-blind research assistant confirms dramatic differences in openness across hospital units. In some, nurses speak freely about errors; in others, mistakes mean punishment and humiliation. The correlation between openness and reported error rates is nearly perfect. This insight becomes the foundation of psychological safety, a concept now supported by over a thousand research papers showing that teams where people feel safe to speak up perform better, experience less burnout, and in healthcare settings, see lower patient mortality.