In 1999, Harvard Business School professor Amy Edmondson published research that would become one of the most influential findings in organizational psychology. Her study of medical teams found that higher-performing teams actually reported more errors than lower-performing ones — not because they made more mistakes, but because they felt safe enough to discuss them. She named the condition that allowed this: psychological safety. Google’s Project Aristotle, a large-scale study of what made its teams effective, found that psychological safety was the single most important factor differentiating high-performing teams from others.
Table of Contents
- What Is Psychological Safety?
- The Research Evidence
- What Psychological Safety Is Not
- Leader Behaviors That Build or Destroy It
- Building a Psychologically Safe Culture
- Frequently Asked Questions
What Is Psychological Safety?
Psychological safety, as defined by Edmondson, is a shared belief held by members of a team that the team is safe for interpersonal risk-taking. Specifically: people feel they can speak up with ideas, questions, concerns, and mistakes without fear of punishment, humiliation, or rejection. It is not the absence of conflict but the presence of enough safety to engage in authentic communication — including disagreement, error disclosure, and challenging questions.
The concept addresses a fundamental human challenge: every time you speak up at work, you risk being judged as ignorant (if you ask a basic question), incompetent (if you reveal a mistake), disruptive (if you challenge a norm), or negative (if you raise a concern). These interpersonal risks are manageable when the cost of silence exceeds the perceived cost of speaking. In most workplaces, chronic silence is the rational response to the perceived risks of authenticity — which means organizations systematically lose information, ideas, and early warning signals that could prevent problems.
The Research Evidence
Edmondson’s research across hospital units, manufacturing plants, and product development teams consistently finds that psychological safety predicts learning behavior and performance. Teams with higher psychological safety report errors more readily, learn from them faster, innovate more successfully, and show higher customer satisfaction. Google’s Project Aristotle — analyzing over 180 Google teams — found psychological safety was the most important of five key team effectiveness factors, without which the other four (dependability, structure and clarity, meaning, and impact) could not operate effectively.
What Psychological Safety Is Not
Psychological safety is frequently confused with trust, niceness, or low standards. It is none of these. Trust is about expectations of another’s behavior; psychological safety is about a group climate that makes risk-taking safe. Niceness is surface-level pleasantness that can coexist with deep psychological unsafety — teams where conflict is avoided by suppressing authentic views are not psychologically safe; they are silently toxic. Low standards is perhaps the most important misconception: high psychological safety combined with high performance standards produces what Edmondson calls a “learning zone” — the conditions for both engagement and excellence. High standards without psychological safety produces the “anxiety zone” — high performance pressure with fear of speaking up, the most dangerous combination in complex work.
Leader Behaviors That Build or Destroy Psychological Safety
Building behaviors: Acknowledging your own fallibility and uncertainty; modeling intellectual humility by asking questions rather than always having answers; explicitly inviting input (“What am I missing?”); responding non-defensively to bad news and mistakes; following up on suggestions; explaining reasoning behind decisions; expressing genuine curiosity about others’ perspectives. Destroying behaviors: Reacting to bad news with blame or punishment; humiliating people who make mistakes publicly; dismissing concerns or questions; proceeding without explanation or consultation; rewarding only successful outcomes and punishing exploratory failures; interrupting, dismissing, or ignoring contributions.
Building a Psychologically Safe Culture
Edmondson identifies three things leaders must do consistently: frame work as a learning problem rather than an execution problem (acknowledging uncertainty and the need for all perspectives), acknowledge their own fallibility (modeling that not-knowing is safe), and proactively invite participation (asking directly for input rather than waiting for it). The invitation matters because speaking up feels risky; being explicitly asked to contributes removes the first barrier. According to Edmondson’s research published in her book The Fearless Organization, psychological safety is not about removing accountability but about creating the conditions where people can learn from mistakes faster than the cost of those mistakes accumulates.
Frequently Asked Questions
Is psychological safety the same in all cultures?
The underlying dynamics operate cross-culturally — people everywhere manage the risk of speaking up versus staying silent. But baseline levels of interpersonal risk-taking and the specific behaviors that create safety vary by culture, organizational context, and industry. Cultures with high power distance (strong deference to hierarchy) typically have lower baseline psychological safety and require more deliberate intervention by leaders to create conditions for authentic input. The specific behaviors leaders use to invite participation may also need cultural adaptation.
How can you measure psychological safety in a team?
Edmondson’s seven-item Psychological Safety Scale is the most widely used instrument, assessing agreement with statements like “If you make a mistake on this team, it is often held against you” (reverse scored) and “It is easy to speak up about problems and tough issues on this team.” Qualitative signals also indicate low safety: long silences after difficult questions, ideas always being attributed upward rather than to their originators, errors discovered late, and a pattern of “parking lot conversations” (important things said in private after meetings that were not said in the meeting).
