A project fails in two teams. In the first, members report what happened, ask what they missed, and revise the process. In the second, people discuss the failure privately, soften the warning, or wait for someone with more authority to speak. The difference is not whether the teams make mistakes. It is whether learning from mistakes becomes a shared process.

Psychological safety helps explain that difference. Amy Edmondson defined team psychological safety as a shared belief that a team is safe for interpersonal risk-taking. When that belief is stronger, people are more willing to ask questions, seek help, test assumptions, disclose errors, and offer dissenting information (Edmondson, 1999). These actions are the visible work of team learning.

Psychological safety is sometimes treated as a promise that work should feel comfortable. That interpretation obscures its purpose. Psychological safety concerns whether candid participation is possible while a team pursues demanding work. It does not remove evaluation, consequences, expertise, or responsibility. It makes relevant information more available for responsible action (Edmondson & Bransby, 2023).

What Psychological Safety Means

Psychological safety is a belief about the likely interpersonal consequences of speaking or acting within a particular group. A member silently asks: If I admit that I do not know, point out a possible failure, or question the plan, how will people respond? The answer is rarely written into a policy. It develops through repeated interactions and becomes part of what members understand about how the team works (Edmondson, 1999).

This makes psychological safety primarily a property of a social system rather than a stable personal trait. Individuals may differ in confidence, history, and sensitivity to rejection, but group dynamics also create shared expectations. The same employee may speak freely in one team and remain guarded in another. Contemporary research continues to find meaningful differences among teams within the same organization (Edmondson & Bransby, 2023).

Trust contributes to this climate, but the concepts are not identical. Trust often concerns what one person expects another person to do in the future. Psychological safety describes the broader interpersonal environment in which members decide whether a question, concern, or unfinished idea can be voiced without avoidable humiliation or rejection (Edmondson, 1999).

Safety Is Not Comfort or Agreement

A psychologically safe team can hold tense conversations. Members may disagree sharply, receive corrective feedback, or face consequences for negligent and harmful conduct. Safety concerns whether people can contribute relevant information and engage in good-faith inquiry. It does not guarantee that every contribution will be accepted or that every decision will feel affirming (Edmondson & Bransby, 2023).

Cohesion is also an unreliable substitute. Irving Janis used the term groupthink for situations in which the pursuit of agreement interferes with careful appraisal. A group can feel close and confident while members censor doubts or assume that silence means consent (Janis, 1972). Psychological safety should make disagreement more discussable, not make agreement mandatory.

How Teams Learn From Mistakes

Why Speaking Up Feels Risky

Questions and error reports carry social meaning. Asking for help may suggest that we lack competence. Challenging an assumption may appear disrespectful. Reporting a mistake can threaten reputation, status, or future opportunities. The potential gain belongs to the team, but the immediate embarrassment may fall on the person who speaks (Edmondson, 1999; Detert & Burris, 2007).

Erving Goffman’s account of face-work helps explain this hesitation. In social encounters, people try to sustain an acceptable image of themselves. Information that disrupts that image can create the experience of losing face, along with shame, confusion, or a loss of composure (Goffman, 1955). The risk is especially strong when an employee must reveal uncertainty before people who evaluate performance or control resources.

These concerns are not necessarily irrational. Teams teach members what candor costs. A sarcastic response, a public rebuke, or a warning that disappears into silence can influence later choices. So can curiosity, appreciation, and visible correction. Psychological safety develops from accumulated evidence about what happens after someone takes an interpersonal risk (Detert & Burris, 2007; Nembhard & Edmondson, 2006).

Psychological Safety and Team Learning

Edmondson treated learning as an activity rather than a possession. Team learning behavior includes seeking feedback, sharing information, asking for help, discussing errors, reflecting on results, testing assumptions, and experimenting. Through these activities, a team discovers gaps in its understanding and modifies what it does (Edmondson, 1999).

When members expect punishment or rejection, protecting themselves can feel more urgent than asking what went wrong. They may hide an error, avoid a difficult question, or move disagreement into private conversations. When that interpersonal cost is lower, more of what team members know can enter the discussion, and problems can become visible while correction is still possible (Edmondson, 1999).

A 2017 meta-analysis combined 136 samples involving more than 22,000 people and nearly 5,000 groups. Across those studies, psychological safety was positively related to information sharing, speaking up, learning behavior, and performance. At the group level, the relationship was stronger with learning behavior than with performance. This pattern suggests that safety helps teams partly by making it easier to exchange and use information. Because many of the studies were correlational, the result does not prove that increasing safety by itself will improve performance (Frazier et al., 2017).

The distinction between psychological safety and team efficacy is useful here. Efficacy concerns whether members believe the team can accomplish its task. Safety concerns whether they can expose uncertainty while trying. In Edmondson’s original study, team efficacy did not predict learning behavior after psychological safety was considered. A capable team can therefore remain a poor learner when confidence makes questions costly or unnecessary in members’ eyes (Edmondson, 1999).

Disclosure Is Only the First Step

An admitted mistake does not automatically teach a team. Disclosure opens the possibility of learning, but the group must still examine what happened, determine what information was missing, and change future action. Research on error-management culture similarly emphasizes communication, rapid detection, analysis, correction, and the sharing of error knowledge, not disclosure alone (van Dyck et al., 2005).

A useful review asks how the task, tools, communication, assumptions, and individual choices interacted. This wider examination does not erase personal responsibility. It prevents the search for a single culprit from ending inquiry before the team understands the conditions that made the error possible. The aim is an explanation detailed enough to guide a different response next time (Reason, 2000).

Debriefing research offers additional support for structured reflection. Across 46 samples, participants who received debriefs performed about 20 to 25 percent better on average than comparison participants. The studies varied considerably, however, and many were quasi-experimental; the result supports reviewing experience but does not establish that any meeting labeled a debrief will work (Tannenbaum & Cerasoli, 2013).

Teams also need a way to carry the lesson forward. A revised checklist, a new handoff, a changed decision rule, or additional training helps close the theory-practice gap. The team should later examine whether the change worked. In this sense, learning depends on feedback loops: action produces information, interpretation shapes adjustment, and the next result tests that adjustment (Tannenbaum & Cerasoli, 2013; van Dyck et al., 2005).

This process requires standards. A team cannot learn unless it can distinguish the intended result from what occurred. Psychological safety is therefore most useful when combined with a clear purpose, reliable information, adequate resources, and expectations that members will respond to what they learn (Edmondson, 1999; Edmondson & Bransby, 2023).

Why Error Reports May Increase

A low error count can look reassuring while concealing an information problem. Reported errors reflect at least two things: what occurs and what people are willing or able to make visible. Research in hospital teams showed that detected error rates depend partly on whether team members notice, discuss, and report errors. A team that makes problems easier to report may therefore appear to have more errors even when greater openness is improving detection and learning (Edmondson, 1996).

More reports do not prove that a team is learning. Reports become useful when they are examined, translated into corrective action, and used to prevent repetition. Error-management research therefore treats communication as one part of a larger process that includes detection, analysis, recovery, and knowledge sharing (van Dyck et al., 2005). Teams should track severity, recurrence, corrective follow-through, and whether risks are being surfaced early, rather than relying on reporting volume alone.

When Low Safety Becomes a Loop

Low-safety teams can become self-sealing. Members keep concerns private because speaking seems risky. The team then receives less contradictory information, so flawed assumptions survive. Poor results increase pressure, and pressure can make leaders and members even more defensive. Silence begins to look like agreement, while the absence of reported problems is mistaken for the absence of problems (Edmondson, 1999).

Constructive responses can create a different pattern. When a well-intentioned question is taken seriously, members gain evidence that inquiry is worthwhile. More information enters the group, correction becomes more likely, and useful responses strengthen the expectation that future candor will be treated constructively. Safety and learning may reinforce one another over time, although the original study measured teams at one point in time and could not test that dynamic directly (Edmondson, 1999).

Research has not established a universal sequence for rebuilding psychological safety after a breach (Edmondson & Bransby, 2023). Naming what happened, inviting correction, and changing the behavior are reasonable responses, but they should not be presented as a validated formula. Repair becomes credible through what the team observes afterward, not through an apology alone.

How Leadership and Status Shape Psychological Safety

How Leaders Shape the Cost of Candor

Leaders have unusual influence because their reactions are highly visible. A formal invitation to speak matters less than what happens when someone delivers unwelcome information. If the leader becomes defensive, attacks the messenger, or quietly penalizes dissent, the team receives a lesson. If the leader asks what can be learned and follows through, that response also becomes evidence (Detert & Burris, 2007; Nembhard & Edmondson, 2006).

Edmondson’s manufacturing study found that leader coaching and organizational support were associated with psychological safety. Detert and Burris later studied employee voice in a restaurant chain. Employees were more likely to speak up when managers seemed open to ideas and willing to act on them, and psychological safety helped explain that connection. Broad transformational leadership showed less consistent results (Edmondson, 1999; Detert & Burris, 2007). These findings distinguish visible managerial openness from a general leadership style. Team members need observable evidence that input is welcome. Leadership matters, but work design, resources, colleagues, status relations, and organizational practices also shape the cost of speaking.

Leader-expressed humility offers one relevant pattern. Bradley Owens and colleagues describe expressed humility through accurate self-appraisal, appreciation of others’ strengths, and teachability. Their research associated leader-expressed humility with a team learning orientation. A leader who can acknowledge limits gives members evidence that not knowing is compatible with competent leadership (Owens et al., 2013).

Managerial Assumptions and the Climate for Voice

Psychological safety also belongs to the wider study of organizational psychology, which examines how leadership, group processes, status, and organizational practices shape behavior at work. A question that appears individual, such as why a particular employee remained silent, may have an organizational answer. Team members may have learned that raising concerns brings little response, threatens their standing, or violates an unwritten expectation about who is permitted to challenge a decision (Edmondson & Bransby, 2023; Detert & Burris, 2007).

Douglas McGregor’s Theory X and Theory Y offers a useful lens for examining these expectations. McGregor argued that management practices reflect assumptions about whether employees must be closely controlled or can exercise self-direction and responsibility. Psychological safety is not equivalent to Theory Y, and a Theory Y orientation does not guarantee candor. The connection lies in how managerial assumptions become visible through work design, participation, and responses to employee initiative. Those experiences give team members evidence about whether questions, dissent, and admitted uncertainty are genuinely welcome (McGregor, 1960; Gannon & Boguszak, 2013; Detert & Burris, 2007).

How Status and Hierarchy Shape Participation

The interpersonal cost of candor is not distributed evenly. People with less formal authority, lower professional status, or greater dependence on a supervisor may face more uncertainty about whether their knowledge will be respected. In 23 neonatal intensive care units, physicians reported higher psychological safety than nurses and respiratory therapists, although the size of that status gap differed across teams (Nembhard & Edmondson, 2006).

Leader inclusiveness helped explain that variation. When physician leaders explicitly invited and appreciated others’ contributions, differences in psychological safety between higher- and lower-status professionals were smaller; psychological safety was also associated with greater engagement in quality-improvement work. Because the study was conducted in specialized health-care teams, it does not justify a universal causal claim. It does show why a formally open door may be insufficient: invitations carry different meaning when status determines who usually gets heard (Nembhard & Edmondson, 2006).

Accountability and Psychological Safety

Psychological safety and accountability describe different features of a team’s climate. Safety concerns whether members can tell the truth about the work. Accountability concerns whether people meet appropriate standards and respond responsibly when they do not. A team can be high or low on either dimension. An agreeable climate without clear standards may produce little learning, while strict standards without candor can hide the information needed to apply them intelligently (Edmondson & Bransby, 2023).

Separating conduct from humiliation is central. An inadvertent error, a normalized shortcut, and a knowing violation may require different responses. James Reason contrasted a person approach, which locates failure in individual shortcomings, with a systems approach, which asks how conditions and defenses allowed an error to reach its consequences. A systems view does not make individuals irrelevant; it keeps blame from ending the investigation too early (Reason, 2000).

Psychological safety also needs a responsible purpose. Frazier and colleagues noted that low interpersonal risk can sometimes enable undesirable behavior when team norms are unethical. Edmondson and Bransby similarly caution against treating psychological safety as the goal itself. Its value lies in helping people communicate about the work as part of a disciplined process directed toward responsible performance (Frazier et al., 2017; Edmondson & Bransby, 2023).

Practices That Support Psychological Safety

Invite Information That Could Change the Plan

A vague request for feedback is easy to answer with silence. More useful questions identify the uncertainty: What are we assuming? What could make this fail? What information would change our decision? Who sees the problem differently? These questions are examples rather than a tested script. They apply the broader evidence that visible managerial openness and explicit invitations make speaking up more credible (Detert & Burris, 2007; Nembhard & Edmondson, 2006).

Respond Before Defending

The first response should keep the conversation open long enough to understand what happened. A leader or teammate can ask for the sequence of events, clarify what the speaker observed, and thank the person for raising the concern. These active-listening responses do not require agreement with the interpretation. They help prevent immediate defensiveness from closing the inquiry before the evidence is heard. In an experiment involving initial conversations, active-listening responses increased the speaker’s feeling of being understood relative to advice or simple acknowledgment. Because this was not a team-learning study, its implication for teams remains limited (Weger et al., 2014).

Review Errors Without Erasing Responsibility

A review should examine both the system and the choices within it. The team can ask what made the action seem reasonable at the time, what safeguards failed, and what obligations were neglected. This produces a more accurate account than automatic blame or automatic absolution. Error prevention remains important, but resilient systems also assume that some errors will occur and build defenses that detect, contain, and learn from them (Reason, 2000).

Close the Learning Loop

Speaking up becomes credible when people can see what followed. Explain what changed, what remains unresolved, and why a suggestion was not adopted when that is the decision. Structured debriefs can help a team examine results and set the next goal, but the review must remain aligned with the work and lead to action. Visible follow-through turns candor from a symbolic value into part of the team’s operating process (Tannenbaum & Cerasoli, 2013; van Dyck et al., 2005).

Limits of Psychological Safety Research

Edmondson’s foundational study examined 51 teams in one manufacturing company. It combined several methods and linked psychological safety with learning behavior and independently rated performance. Because the survey measured teams at one point in time, it could not establish which factor came first or show how safety develops and erodes over time (Edmondson, 1999).

The later meta-analysis found consistent positive associations across a much larger literature, but many of its studies collected several measures from the same respondents at one point in time. Relationships with performance were smaller when performance was measured objectively. The authors also found signs that studies with positive results may be overrepresented in the published literature. These limitations support confidence in the general relationship while arguing against simple causal promises (Frazier et al., 2017).

The 2023 review identified 185 recent empirical papers and concluded that psychological safety had become a mature construct. It also found important gaps: relatively few intervention studies, limited longitudinal evidence, and too little research on fluid, remote, and hybrid teams. Psychological safety appears especially consequential in complex, knowledge-intensive work. Its effects may be less pronounced when tasks are highly routine and feedback is already built into the work (Edmondson & Bransby, 2023).

The additional studies point in a similar direction, but each has limits. The leadership and health-care studies depended heavily on employees’ perceptions within particular organizations. The error-management studies found relationships with performance but could not fully show how those relationships developed. Many debrief studies compared groups without randomly assigning participants. Together, the findings support a plausible learning process, but they do not prove that one intervention will work for every team (Detert & Burris, 2007; Nembhard & Edmondson, 2006; Tannenbaum & Cerasoli, 2013; van Dyck et al., 2005).

A Few Words by Psychology Fanatic

Mistakes expose more than a faulty action. They expose what we do not know and how much we depend on other people to see what we missed. That exposure can feel like a threat to competence. A team may respond by protecting appearances, or it may protect the process through which incomplete knowledge becomes more accurate.

Psychological safety does not make failure harmless. It makes failure more discussable. The difference matters because an error cannot guide a team while it remains hidden. When people can speak honestly and remain responsible for what follows, a mistake has a chance to become something more useful than blame. It can become part of the team’s capacity to learn.

Associated Concepts

• Groupthink: The pursuit of agreement can suppress dissent and realistic appraisal, making it an important contrast with psychological safety.
• Feedback loops: Team responses to candor become information that shapes whether members speak the next time.
• Theory X and Theory Y: McGregor’s framework examines how managerial assumptions about employees shape control, participation, responsibility, and the conditions under which people contribute.
• Humility: Accurate self-appraisal and teachability can help leaders acknowledge limits and receive correction.
• Active listening: Attentive, clarifying responses can help a speaker feel understood long enough for the information to be examined.
• Losing face: Face concerns help explain why admitting uncertainty or error can carry an interpersonal cost.

References

Detert, James R.; Burris, Ethan R. (2007). Leadership Behavior and Employee Voice: Is the Door Really Open? Academy of Management Journal, 50(4), 869-884. DOI: 10.5465/AMJ.2007.26279183.
(Return to Main Text)

Edmondson, Amy C. (1996). Learning From Mistakes Is Easier Said Than Done: Group and Organizational Influences on the Detection and Correction of Human Error. The Journal of Applied Behavioral Science, 32(1), 5-28. DOI: 10.1177/0021886396321001.
(Return to Main Text)

Edmondson, Amy C. (1999). Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2), 350-383. DOI: 10.2307/2666999.
(Return to Main Text)

Edmondson, Amy C.; Bransby, Derrick P. (2023). Psychological Safety Comes of Age: Observed Themes in an Established Literature. Annual Review of Organizational Psychology and Organizational Behavior, 10, 55-78. DOI: 10.1146/annurev-orgpsych-120920-055217.
(Return to Main Text)

Frazier, M. Lance; Fainshmidt, Stav; Klinger, Ryan L.; Pezeshkan, Amir; Vracheva, Veselina (2017). Psychological Safety: A Meta-Analytic Review and Extension. Personnel Psychology, 70(1), 113-165. DOI: 10.1111/peps.12183.
(Return to Main Text)

Gannon, Dave; Boguszak, Anna (2013). Douglas McGregor’s Theory X and Theory Y. CRIS—Bulletin of the Centre for Research and Interdisciplinary Study, 2013(2), 85–93. DOI: 10.2478/cris-2013-0012.
(Return to Main Text)

Goffman, Erving (1955). On Face-Work: An Analysis of Ritual Elements in Social Interaction. Psychiatry, 18(3), 213-231. DOI: 10.1080/00332747.1955.11023008.
(Return to Main Text)

Janis, Irving L. (1972). Victims of Groupthink: A Psychological Study of Foreign-Policy Decisions and Fiascoes. Houghton Mifflin. ISBN: 978-0-395-14002-4.
(Return to Main Text)

McGregor, Douglas (1960). The Human Side of Enterprise. McGraw-Hill. ISBN: 9780070450929
(Return to Main Text)

Nembhard, Ingrid M.; Edmondson, Amy C. (2006). Making It Safe: The Effects of Leader Inclusiveness and Professional Status on Psychological Safety and Improvement Efforts in Health Care Teams. Journal of Organizational Behavior, 27(7), 941-966. DOI: 10.1002/job.413.
(Return to Main Text)

Owens, Bradley P.; Johnson, Michael D.; Mitchell, Terence R. (2013). Expressed Humility in Organizations: Implications for Performance, Teams, and Leadership. Organization Science, 24(5), 1517-1538. DOI: 10.1287/orsc.1120.0795.
(Return to Main Text)

Reason, James (2000). Human Error: Models and Management. BMJ, 320(7237), 768-770. DOI: 10.1136/bmj.320.7237.768.
(Return to Main Text)

Tannenbaum, Scott I.; Cerasoli, Christopher P. (2013). Do Team and Individual Debriefs Enhance Performance? A Meta-Analysis. Human Factors, 55(1), 231-245. DOI: 10.1177/0018720812448394.
(Return to Main Text)

van Dyck, Cathy; Frese, Michael; Baer, Markus; Sonnentag, Sabine (2005). Organizational Error Management Culture and Its Impact on Performance: A Two-Study Replication. Journal of Applied Psychology, 90(6), 1228-1240. DOI: 10.1037/0021-9010.90.6.1228.
(Return to Main Text)

Weger, Harry, Jr.; Bell, Gina Castle; Minei, Elizabeth M.; Robinson, Melissa C. (2014). The Relative Effectiveness of Active Listening in Initial Interactions. International Journal of Listening, 28(1), 13-31. DOI: 10.1080/10904018.2013.813234.
(Return to Main Text)