Psychological safety

You can usually feel when a team has gone quiet. Meetings end with nods instead of questions. People spot a problem early but wait until it’s too big to ignore. Mistakes get tidied away rather than talked about. Nobody decided to hold back, exactly — it just became the safer choice.

That feeling has a name, and decades of research behind it: psychological safety. It’s the shared sense that your team is a safe place to speak up, ask a “dumb” question, disagree with the boss, or admit you got something wrong. When it’s present, teams learn faster and catch problems sooner. When it’s missing, the smartest people in the room stay silent. This page covers where the idea came from, the validated questions you can ask your own team, and how to actually build it.

Psychological safety is the belief that you won’t be punished or embarrassed for speaking up with ideas, questions, concerns, or mistakes. Harvard professor Amy Edmondson, who has studied it for more than two decades, defines it as “a shared belief held by members of a team that the team is safe for interpersonal risk taking.”

A few things it is not. It isn’t about being nice, lowering standards, or avoiding hard conversations. A psychologically safe team often has more candid disagreement, not less — because people trust that disagreeing won’t cost them. It’s the opposite of a team where everyone agrees in the room and complains in the hallway.

You can often spot low psychological safety before you measure it. People preface ideas with “this is probably a stupid question.” Bad news travels slowly, or only upward through back channels. The same one or two voices dominate every meeting while others go quiet. Retrospectives stay polite and surface-level. None of these mean your team is broken — but they’re signs that the cost of speaking up feels higher than the cost of staying silent, and that’s exactly what this model helps you measure and reverse.

It’s worth measuring because it’s the foundation almost everything else sits on. A team can have clear goals and talented people, but if no one feels safe enough to flag a risk or ask for help, those strengths never get used. Psychological safety is also one of the few team qualities you can change in weeks, not years — small shifts in how a leader responds to bad news move the needle quickly.

This page is for anyone responsible for how a team works together: team leads, engineering managers, Scrum Masters, agile coaches, and people-ops practitioners. If your team feels capable but cautious — full of capable people who don’t quite say what they think — this is the model to start with.

The story starts with a mistake in the data.

In the mid-1990s, Amy Edmondson was a doctoral researcher studying medication errors in hospitals. Her hypothesis was straightforward: better teams should make fewer mistakes. So she measured team effectiveness and counted error rates across nursing units, expecting the best teams to show the cleanest record.

The numbers came back backwards. The teams that rated highest on effectiveness and leadership had more reported errors, not fewer. For a moment it looked like good teamwork caused mistakes — which made no sense.

The explanation, once she dug in, reshaped how we think about teams. The best teams weren’t making more mistakes. They were talking about them. On those units, people felt safe enough to report a near-miss, ask a question, or admit they were unsure — so errors surfaced where they could be caught and learned from. On the weaker teams, mistakes happened just as often but stayed hidden, because owning up felt too risky. What she had actually measured wasn’t error rates. It was who felt safe enough to speak.

Edmondson named the difference psychological safety and tested it properly in a landmark study of 51 work teams, published in 1999 in Administrative Science Quarterly. She showed that psychological safety drives team learning behavior — asking questions, seeking feedback, discussing errors, experimenting — and that this learning, in turn, drives performance. It wasn’t a feel-good idea. It was a measurable link from how safe a team feels to how well it does its work.

The mechanism is worth sitting with, because it explains why this matters for any team, not just hospital wards. Most real work is uncertain: you don’t know in advance which idea will work, which assumption is wrong, or which small problem will become a big one. The only way a team gets smarter about that uncertainty is by surfacing it out loud — and people only do that when it feels safe. Take the safety away and the learning stops, no matter how talented the team is. That’s the chain Edmondson traced: safety enables candor, candor enables learning, and learning is what actually moves performance.

The concept stayed mostly inside academia until 2015, when Google’s Project Aristotle studied 180+ of its own teams to find out what made the best ones tick. Out of everything they measured, psychological safety came out as “by far the most important” of the five dynamics they found. That’s how an academic term from a 1999 hospital study became a word people use in boardrooms — and why it sits at the center of so many team-health models today.

When Edmondson wanted to measure psychological safety, she didn’t ask people “do you feel safe?” — that’s too vague to act on. Her psychological safety survey is a 7-item scale that asks instead about specific, recognizable situations: asking for help, raising a problem, making a mistake, being different. Agree or disagree with each, and a clear picture emerges.

A note on wording. Edmondson’s original scale deliberately mixed in reverse-keyed items — some statements were phrased so that disagreeing was the healthy answer (for example, the original “If you make a mistake on this team, it is often held against you”). That’s good research design, but it’s confusing on a recurring team pulse, where you want one consistent direction. So the versions below are normalized to positive phrasing: for every statement, agreement always means healthier. Three items were flipped from Edmondson’s reverse-keyed originals; the meaning is preserved.

Here are the seven, the way you’d put them to your team:

Use a simple agree/disagree scale (for example, 1 = strongly disagree to 5 = strongly agree) and keep responses anonymous. Anonymity matters more here than almost anywhere else: asking people to non-anonymously rate how safe they feel to speak up is its own little trap. The score you care about is the team average, plus the spread — a team that’s split between 5s and 2s is telling you something a single average hides.

These questions are adapted from Edmondson’s validated scale; you can read the original 1999 paper for the source wording.

Parabol’s Team Health tool measures five categories: psychological safety, dependability, structure & clarity, meaning, and impact. These aren’t arbitrary — they’re the five dynamics from Google’s Project Aristotle, which is why so many models map onto them.

Psychological safety is the first of the five, and in a real sense it’s the one the others depend on. It’s hard to be dependable if no one feels safe admitting they’re behind. It’s hard to get clarity if people won’t ask the question everyone’s secretly wondering. Edmondson built her seven questions to measure one thing, but a few of them naturally reach into neighboring categories — which is honest, because real teams don’t split into perfect buckets:

If you want the full picture across all five categories, start with Project Aristotle, which is the model Parabol’s categories are built on. Think of psychological safety as the soil: dependability, clarity, meaning, and impact all grow better when it’s healthy, and struggle when it isn’t.

Measuring is the easy part. Here’s a simple way to run it, and then what to do with what you learn.

Run it as a regular pulse, not a one-off. Psychological safety moves. A single reorg, a harsh response to bad news, or one new manager can shift it. Ask the seven questions on a recurring cadence — monthly or each sprint works well — so you’re watching a trend line, not a snapshot. The direction of travel tells you more than any single score.

Keep it anonymous, and say so. People can’t safely tell you they don’t feel safe if their name is attached. Make anonymity explicit, and resist the urge to guess who said what. The moment a team suspects you’re decoding individual answers, the data — and the trust — is gone.

Discuss the results with the team, not about them. The biggest mistake leaders make is reading the scores privately and quietly resolving to “do better.” Share the results openly. Pick one question that scored low and ask, “What’s behind this? What would make a 4 into a 5?” The act of discussing safety openly is itself one of the strongest ways to build it.

Model the behavior you’re measuring. Psychological safety is set more by what leaders do than what they say. A few moves that consistently work:

You don’t need a tool to do any of this. But a tool makes the measuring and the discussing far easier — which is the whole point of running it as a team.

Psychological safety is one of the five categories Parabol’s Team Health check measures. Parabol’s Team Health check lets your team rate how you’re doing across five research-backed categories — psychological safety, dependability, structure & clarity, meaning, and impact — as a quick, recurring, anonymous pulse, then turns the results into a conversation.

Try Parabol Health Check

Psychological safety is one piece of a bigger team-health picture. These related models pair naturally with it.

What is psychological safety in a team?

Psychological safety is the shared belief that a team is a safe place to take interpersonal risks — to speak up, ask questions, disagree, or admit mistakes — without fear of being embarrassed or punished. The term was defined and validated by Harvard’s Amy Edmondson in 1999, and popularized by Google’s Project Aristotle in 2015.

What are the 7 psychological safety questions?

Amy Edmondson’s validated scale measures seven things: whether mistakes are held against you, whether you can raise tough problems, whether people are accepted when they’re different, whether it’s safe to take a risk, whether it’s easy to ask for help, whether teammates would undermine your efforts, and whether your unique skills are valued. You answer each on an agree/disagree scale, ideally anonymously.

Isn’t psychological safety just about being nice?

No — and this is the most common misunderstanding. Psychological safety isn’t about lowering standards or avoiding conflict. Safe teams usually have more candid disagreement, because people trust that speaking honestly won’t be held against them. The goal is candor with respect, not comfort.

How do you measure psychological safety?

Ask the seven survey questions on a recurring, anonymous pulse — monthly or each sprint — and watch the team average and the spread of answers over time. The trend matters more than any single score. Then discuss the results openly with the team, because talking about safety is itself one of the best ways to build it.

Who created the concept of psychological safety?

Harvard Business School professor Amy C. Edmondson introduced and validated team psychological safety in a 1999 study of 51 work teams, published in Administrative Science Quarterly. The idea grew out of her earlier research on medication errors in hospitals, where she found that better teams reported more errors — because they felt safe enough to talk about them.