This question bank in: English · Easy Read · Español · Deutsch · Français · 한국어

Psychological safety survey questions

The full bank we use, grouped by theme, with what each question measures and why it is worded the way it is. Pick the ones that fit your team and run them as an anonymous survey.

Build a team surveySee what the report looks like

How to choose yours

Every question below is a statement answered on a five-point scale from strongly disagree to strongly agree. That format matters: it asks people to place themselves rather than to summarise a whole team in a word, and it produces a distribution you can read rather than a single score.

Pick seven to ten. That takes about three minutes to answer, which keeps completion rates high, and it is more than enough to show where a team disagrees with itself. Fourteen is a sensible upper limit. You do not need to cover every theme — a survey of five statements about speaking up will tell you more than a survey of twenty covering everything shallowly.

Adapt the wording to your context. A technology team might ask about retrospectives where an aviation team says debriefs and a hospital says handovers. Fitting the language to the team is not a compromise of the instrument; it is the point of adapting one.

Free to use and adapt in your own survey tool: Copy all the questions Print or save as PDF

Speaking up

The behaviour psychological safety enables. Everything else on this page describes conditions; these ask directly whether the thing itself happens.

1. On this team I can raise problems and challenge ideas.Adapted from EdmondsonThe core item. Asks about the respondent's own experience, not the team's reputation.
2. Members of this team feel able to raise concerns they have about our work.Deliberately about others. Where this and the item above diverge, someone is safe while believing colleagues are not.
3. On this team we are usually comfortable talking about disagreements.Disagreement is the visible test of safety — teams that never disagree are usually quiet rather than aligned.
4. On this team we ensure everyone's voice is heard.Distinguishes a team where speaking up is permitted from one where it is actively made possible.

Learning from failure

What happens after something goes wrong is the clearest signal a team gives about whether it is safe to be imperfect in front of each other.

1. If I make a mistake, the team will not hold it against me.Adapted from EdmondsonEdmondson's framing, turned positive. Consequence, not comfort.
2. On this team I can admit mistakes without fearing criticism.The disclosure version: not whether mistakes are forgiven, but whether they can be mentioned at all.
3. In this team we talk about mistakes and how to learn from them.Moves from tolerance to practice. Many teams forgive errors and still never discuss them.
4. When something goes wrong, this team looks at how the work happened, rather than for someone to blame.Systems thinking versus person-blame — the difference between learning and finding a culprit.
5. When this team reviews how work went, we focus on learning rather than blame.About the ritual specifically. Adapt "reviews how work went" to whatever your team calls it.

Innovation

Psychological safety is what allows an unfinished idea to be said out loud. These questions ask whether that happens here.

1. On this team it is safe to try something that might not work.Adapted from EdmondsonThe permission that makes experiments possible.
2. On this team I feel safe offering new ideas, even if they aren't fully formed plans yet.Half-formed ideas are the ones that get withheld first, and the ones worth most.
3. On this team we welcome each other's ideas and give them time and attention.Attention, not just tolerance — being heard and being ignored politely feel very different.
4. In this team's discussions, we ask what we might be assuming and what we might be missing.Whether challenge is built into how the team thinks, rather than left to brave individuals.
5. This team takes time to find new ways to improve our work or processes.Improvement needs slack. A team at full utilisation has no room to try anything.

Team cohesion

The everyday relationships that make risk affordable. Note that cohesion is not the same as safety — a very close team can still be one where nobody says the difficult thing.

1. It is easy to ask other team members for help.Adapted from EdmondsonHelp-seeking is a small, frequent act of vulnerability, which makes it a sensitive early indicator.
2. On this team, it's ok to say "I don't know".Tests whether competence must be performed here.
3. My skills are valued and used in this team.Adapted from EdmondsonBeing useful is part of feeling safe enough to contribute.
4. Other team members value the work I do.Recognition from peers, which usually matters more day to day than recognition from above.
5. Nobody on this team would deliberately act in a way that undermines my efforts.Adapted from EdmondsonEdmondson's item, almost word for word. Deliberately strong wording — agreement should be easy in a healthy team.
6. People on the team prioritise team success over individual success.Where individual incentives compete, admitting weakness becomes expensive.

Inclusion and diversity

Psychological safety is rarely evenly distributed. These questions look for the people for whom it is lower.

1. I feel able to be myself at work.The cost of masking falls unevenly, and it is paid before any work gets done.
2. On this team we understand and respect each other's differences.Respect as practice rather than policy.
3. On this team we ensure no one is left out.Exclusion is usually structural — meeting times, socials, informal channels — rather than intended.
4. Team members welcome ideas that are different from their own.Cognitive diversity only pays off if difference survives contact with the group.
5. Members of this team never reject others for being different.Adapted from EdmondsonStrongly worded on purpose. Any disagreement here is worth a conversation.

Transparency

Uncertainty is expensive. When people cannot see what is expected or why decisions were made, they fill the gap with guesses, and cautious guesses produce silence.

1. On this team I understand what is expected of me.Clarity is a precondition for confident action.
2. It is clear what this team's priorities are, and why.The "and why" matters — priorities without reasons cannot be applied to new situations.
3. When this team makes decisions that affect me, I understand why they have been taken.Explained decisions are arguable; unexplained ones are only survivable.
4. On this team, I have a good understanding of whether I am meeting expectations.Not knowing where you stand makes every other risk feel larger.
5. On this team we are judged on what we achieve, not on how busy we appear.Where visible busyness is the currency, nobody admits to having capacity — or to being overloaded.

Voice and power

These four are ours rather than adapted, and they are the ones we would most encourage you to include. They ask not whether people speak up but what governs it: whether reactions are predictable, whether speaking has been worth it, and how much depends on who is in the room.

1. I can usually predict how this team will respond when someone raises a difficult issue.predictability Unpredictability silences people even when the actual response would have been fine. A team can be safe and illegible at the same time, and the fix for that is clarity rather than courage.
2. When I speak up in this team, it tends to be worth it.worth The whole ledger: was it heard, did it cost, did anything change. Deliberately broader than "was I listened to".
3. In this team, people can disagree with the most senior person as easily as with anyone else.gradient Measures the power gradient without asking anyone to state their own position.
4. What I feel able to say in this team depends on who else is present.gradient · reverse-scored The only reverse-scored item here: agreement indicates a steeper gradient. If you use it, score it accordingly, and expect it to be one of the most revealing answers you get.

The seven adapted from Edmondson (1999)

Amy Edmondson's 1999 study of work teams introduced a seven-item scale for team psychological safety. In it, 427 members of 51 teams in a manufacturing company rated each statement on a seven-point scale from "very inaccurate" to "very accurate". Three of the seven were worded negatively, so that people could not score well by agreeing with everything. A team's score was the average of its members' answers.

Seven questions in our bank are adapted from those items. They differ in three ways: the wording, all seven are phrased so that agreeing means safety, and they are answered on this survey's five-point agree–disagree scale. That is why they are not her validated scale, and why a score from them can't be set against published results. For research, or for comparison with other studies, use her original items unchanged.

When you build a survey, a button ticks all seven at once, alongside anything else you choose.

Source: Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. The items are listed in the paper's appendix.

The follow-up questions that matter more than the scores

Knowing that people hold back tells you little on its own. Knowing why tells you what to do, because the reasons need opposite responses. When someone reports finding it hard to speak up, ask a short follow-up:

• It's hard to predict how people will react — fog rather than fear. The fix is making reactions legible and consistent.
• I've seen it go badly before — the price is known and high. The fix is changing what happens next time, publicly.
• It gets acknowledged, but nothing changes — futility. The fix is closing loops visibly, including when the answer is no.
• It depends who is involved — the power gradient. The fix works on who can say what, in whose presence.
• Not sure — it's just how it feels — a real answer, not a missing one. Silence often runs below awareness, and a team that cannot name its reason needs language before it needs a plan.

Ask the same question of the high scores, too. When someone says speaking up here works, ask what makes it possible — that tells a team what to protect, which almost no survey ever asks.

Running these as a survey

Our tool runs them for you: pick the statements, adapt the wording, add your own, and send a link. Answers are anonymous by architecture — no names, no accounts, no timestamps — and results stay sealed until the survey closes.

What comes back is the part worth having. The report shows distributions rather than a score, so you can see where the team disagrees with itself. It tells you why people held back, and what is already working and worth doing more of. And it ends with a short set of experiments chosen against what your team actually reported, ready to take into a conversation.

Build a team surveyRead a sample report

There is also a free three-minute self-check if you would rather look at your own experience first. It stores nothing at all.