Unlocking Team Potential: How Psychological Safety Transforms Learning, Performance, and Organizational Resilience
Download the article here:
Listen to a review of this article:
Abstract: This article examines team psychological safety—a shared belief that a team is safe for interpersonal risk taking—and its implications for learning behavior, team performance, and organizational effectiveness. Drawing primarily on Edmondson's (1999) foundational multimethod field study of 51 work teams, along with subsequent meta-analytic, field, and practitioner research, the article traces the evolution of psychological safety from an academic construct to a widely recognized driver of effective teams. It synthesizes evidence that psychologically safe teams are more likely to seek feedback, discuss errors, and experiment, and that these learning behaviors carry the relationship between safety and performance. The article identifies evidence-based organizational responses, including leader coaching and modeling, structural design and context support, error-embracing cultures, team efficacy, and transparent communication, illustrated with examples from technology, creative industries, mining, manufacturing, and healthcare. Finally, it proposes forward-looking pillars for building long-term organizational learning capacity: continuous recalibration of team norms, distributed leadership, and measurement systems that pair safety with accountability.
Here is an uncomfortable truth that most leaders eventually confront: their most talented teams are not always their highest-performing ones. A team of brilliant individuals can underperform a team of capable but connected colleagues if one critical ingredient is missing—the willingness to take interpersonal risks together. To admit a mistake. To ask a question that might seem naive. To challenge a plan that senior leadership has already endorsed.
This ingredient has a name: psychological safety. Introduced as a team-level construct in Amy Edmondson's 1999 study of work teams in a manufacturing company, psychological safety describes "a shared belief held by members of a team that the team is safe for interpersonal risk taking" (Edmondson, 1999, p. 350). The concept grew partly from an earlier puzzle. In a study of hospital nursing units, Edmondson (1996) found that better-led, better-functioning units reported more medication errors, not fewer. The likely explanation was not that they made more mistakes, but that their members were more willing to report and discuss them. That paradox set the stage for more than two decades of research showing that when people feel safe to speak up, teams learn faster and perform better.
Why does this matter now? Organizations operate under accelerating complexity and uncertainty. Hybrid work has weakened many of the informal cues that once signaled interpersonal trust. Artificial intelligence and automation are reshaping the work teams do, demanding rapid adaptation and collective sense-making. In such conditions, the information that matters most—early warnings, dissenting views, and honest reports of what is not working—often sits with people who must decide whether speaking up is worth the risk.
The practical stakes are high. Organizations that cultivate psychological safety enable a virtuous cycle: candor leads to learning, learning fuels adaptation, and adaptation supports sustainable performance. Those that do not risk a vicious cycle of silence, stagnation, and costly failure. This article offers leaders, consultants, and scholars an evidence-based roadmap for understanding, building, and sustaining psychological safety across organizational contexts.
The Psychological Safety Landscape
Defining Psychological Safety in the Organizational Context
At its core, psychological safety is a group-level phenomenon. It is not merely a trait of courageous individuals, nor is it a synonym for niceness or the absence of accountability. Edmondson (1999) distinguished psychological safety from related constructs. It is not the same as group cohesiveness; cohesive groups can suffer from groupthink, in which the desire for harmony suppresses dissent (Janis, 1982). Nor is it identical to interpersonal trust, though trust is related. As Edmondson (2019) later clarified, trust usually describes a relationship between two parties, whereas psychological safety describes a shared sense of whether the group will give you the benefit of the doubt when you take a risk.
Edmondson (1999) measured team psychological safety with a seven-item survey scale. Items included statements such as "If you make a mistake on this team, it is often held against you" (reverse scored) and "It is safe to take a risk on this team." Importantly, members of the same team tended to agree with one another in their responses, supporting the idea that psychological safety is a shared climate—a taken-for-granted sense of how things work in a particular group—rather than simply an individual feeling.
This team-level character helps explain why psychological safety can vary dramatically within a single organization. In the earlier hospital research, nurses on some units described feeling free to report mistakes to their managers, while nurses on other units in the same hospital described being blamed and made to feel on trial for errors (Edmondson, 1996). Same organization, same policies—very different local climates.
Prevalence, Drivers, and Distribution
Interest in psychological safety surged after Google's internal research initiative, Project Aristotle, studied about 180 teams and concluded that psychological safety was the most important of five dynamics distinguishing effective teams, ahead of dependability, structure and clarity, meaning, and impact (Duhigg, 2016; Google, 2015). The researchers found that who was on a team mattered less than how team members interacted and structured their work.
Edmondson's (1999) model identified two antecedents that shape psychological safety: context support, meaning access to resources, information, and rewards; and team leader coaching, meaning supportive, non-defensive leader behaviors. Both were associated with higher team psychological safety, and psychological safety in turn was associated with more learning behavior. This is encouraging for practitioners: it suggests that psychological safety is not a fixed feature of certain kinds of work but something leaders and organizations can deliberately cultivate.
Subsequent research has shown how broadly the construct applies. In their review, Edmondson and Lei (2014) traced psychological safety's links to learning, voice, and performance across many settings. A meta-analysis of 136 independent samples found that psychological safety was associated with information sharing, learning behavior, engagement, and task performance, and that it was shaped by supportive leadership, peer support, and work design characteristics (Frazier et al., 2017).
Organizational and Individual Consequences of Psychological Safety
Organizational Performance Impacts
The performance case for psychological safety rests on a clear mechanism: learning behavior links safety and performance. In Edmondson's (1999) study, team learning behavior—seeking feedback, sharing information, asking for help, talking about errors, and experimenting—was associated with team performance, and learning behavior mediated the relationship between psychological safety and performance.
In practical terms, psychological safety does not directly satisfy customers or produce better products. Rather, it creates the interpersonal conditions under which teams take the actions that produce those outcomes: asking hard questions, testing assumptions, seeking outside feedback, and correcting course. Edmondson (1999) made this point explicitly, describing psychological safety as facilitating appropriate action rather than serving as a direct driver of performance.
Subsequent research has reinforced this logic. A study of Korean work teams found that psychological safety's relationship with team effectiveness operated through team learning behavior and team efficacy rather than directly (Baek et al., 2020). Psychological safety, in other words, is best understood as an enabler: without it, the learning behaviors that drive performance are less likely to occur.
At the organizational level, the costs of psychological safety's absence can be severe. Edmondson (2019) described how fear and silence contributed to organizational failures such as the Volkswagen emissions scandal and the Wells Fargo fraudulent-accounts crisis, in which employees felt unable to challenge unrealistic targets or raise concerns without risking retribution.
Individual Wellbeing and Stakeholder Impacts
The consequences extend beyond organizational performance. Edmondson's (1999) qualitative data contrasted teams in which members interpreted feedback about mistakes as help from colleagues committed to good work with teams in which members experienced tension, felt unsupported, and protected themselves by withholding information. Such self-protection aligns with threat-rigidity theory, which holds that when people and organizations feel threatened, they tend to narrow their information processing and fall back on familiar responses—precisely when flexibility is most needed (Staw et al., 1981).
How feedback is interpreted, as threat or as partnership, shapes employees' experience of work. In psychologically unsafe teams, people spend energy managing impressions rather than solving problems, and the strain of staying silent about concerns can erode engagement and commitment.
For stakeholders outside the team, the stakes can be even higher. In healthcare, psychological safety affects whether staff raise concerns that protect patients. A study of neonatal intensive care unit teams found that leader inclusiveness—leaders' words and deeds that invite and appreciate others' contributions—was associated with greater psychological safety, which in turn was associated with engagement in quality improvement work, and that inclusiveness helped offset the silencing effect of professional status differences (Nembhard & Edmondson, 2006).
Evidence-Based Organizational Responses
Table 1: Evidence-Based Responses for Building Team Psychological Safety
Response | Evidence Base | Key Practices | Illustrative Example |
Leader coaching and modeling | Leader coaching associated with psychological safety (Edmondson, 1999); leader inclusiveness offsets status differences (Nembhard & Edmondson, 2006) | Frame work as uncertain and interdependent; model fallibility; respond productively to candor; ask genuine questions | Google's Project Aristotle and team-level discussion of results |
Structural design and context support | Context support as an antecedent (Edmondson, 1999); work design linked to safety (Frazier et al., 2017) | Ensure resources and information; design for interdependence; separate feedback from hierarchy | Pixar's Braintrust |
Error-embracing cultures | Better units report more errors (Edmondson, 1996) | Distinguish blameworthy from praiseworthy failure; after-action reviews; reward early detection | Anglo American's use of the lekgotla after mine fatalities |
Team efficacy as a complement | Efficacy linked to learning but not beyond safety (Edmondson, 1999); efficacy as a mediator (Baek et al., 2020) | Clear goals; early wins; skill development; connect effort to impact | Toyota's andon system |
Transparent communication and narrative | Shared narratives sustain or erode safety (Edmondson, 2019) | Articulate shared values; share stories of productive failure; co-create norms | Microsoft's "learn-it-all" culture shift |
Leader Coaching and Modeling Behaviors
Team leader behavior is one of the most powerful levers for psychological safety. Edmondson (1999) found that team leader coaching was associated with team psychological safety and, through it, with learning behavior. She argued that team members pay particular attention to leaders' behavior, because leaders' responses signal what is and is not safe.
This attentiveness means that a single dismissive reaction to a question, or one punitive response to an admitted error, can erode trust built over months. Conversely, when leaders model fallibility—asking genuine questions, acknowledging uncertainty, and responding with interest rather than judgment—they signal that interpersonal risk taking is welcome.
Effective leader practices include:
Frame the work accurately: Acknowledge the uncertainty and interdependence in the team's task, making clear that every voice is needed.
Model fallibility: Saying "I don't know" or "I may be missing something" lowers the perceived cost of speaking up.
Respond productively to candor: Express appreciation when someone raises a concern, focus on learning rather than blame, and reserve sanctions for genuinely blameworthy conduct (Edmondson, 2019).
Ask genuine questions: Move from rhetorical questions to authentic inquiry, such as "What are we missing?" or "What would you do differently?"
Practice inclusiveness: Explicitly invite input from those with less status or experience (Nembhard & Edmondson, 2006).
Google offers a well-known example of translating research into practice. After Project Aristotle identified psychological safety as the most important dynamic of effective teams, Google encouraged teams to discuss their results on team-dynamics surveys and agree on concrete changes, rather than treating psychological safety as an abstract value (Duhigg, 2016; Google, 2015). The lesson is that psychological safety is built through specific, repeated team practices and leader behaviors, not through proclamations.
Structural Design and Context Support
Edmondson's (1999) model positioned structural features—context support and team design—as antecedents of psychological safety. Teams with adequate access to resources, information, and rewards were more likely to develop psychologically safe climates. This makes intuitive sense: when teams must fight for basic resources or operate without critical information, insecurity and defensiveness tend to increase. Meta-analytic evidence likewise links work design characteristics, such as autonomy and interdependence, to psychological safety (Frazier et al., 2017).
Structure is not destiny, however. Edmondson's (1999) qualitative data suggested that some teams developed strong learning behavior despite structural obstacles, while others struggled even when resources were adequate—an indication that interpersonal climate interacts with, rather than simply follows from, structure.
Effective structural approaches include:
Ensure resource adequacy: Provide teams with the information, materials, training, and decision authority they need.
Design for interdependence: Tasks that require genuine collaboration create natural opportunities for the interactions through which safety develops.
Separate feedback from hierarchy: Create forums where ideas are evaluated on their merits, regardless of who offers them.
Create feedback loops: Build regular, non-punitive feedback from customers, peers, and leaders into team routines.
Pixar Animation Studios illustrates deliberate structural design for candor. Its Braintrust meetings bring directors together with experienced colleagues to review films in progress. Feedback focuses on the work, not the person, and the director retains authority over what to do with it. By separating candid evaluation from formal power, the structure makes challenging feedback a contribution to shared work rather than a threat (Edmondson, 2019).
Error-Embracing Cultures and Learning from Failure
Perhaps no aspect of psychological safety is more counterintuitive—or more important—than its relationship to errors. Edmondson's (1996) hospital research showed that units with better leadership and relationships reported more errors, likely because staff there were willing to discuss them, creating opportunities for correction and system improvement. Errors that go unreported cannot be learned from.
Effective error-embracing approaches include:
Distinguish types of failure: Separate blameworthy failures, such as deliberate rule-breaking, from praiseworthy ones, such as well-designed experiments that did not work or honest mistakes in novel situations (Edmondson, 2019).
Conduct after-action reviews: Debrief projects with a focus on "What did we learn?" rather than "Whose fault was it?"
Celebrate early detection: Recognize people who identify problems early, especially when the problems are uncomfortable.
Seek outside perspectives: Invite colleagues from outside the team to review work, providing fresh input that insiders may hesitate to offer.
Anglo American, the global mining company, offers an example from a very different industry. As Edmondson (2019) describes, after a series of fatalities in its South African operations, the company's leadership halted operations at its most dangerous mine and drew on the lekgotla, a traditional South African forum for community deliberation, to engage miners, managers, and union representatives in candid discussion about safety. By building on a culturally resonant practice rather than importing a generic training program, the company created conditions in which people at every level could voice safety concerns. The example shows that psychological safety must be adapted to its cultural and occupational context.
Investing in Team Efficacy as a Complement
Edmondson's (1999) model also included team efficacy—the shared belief that the team is capable of performing its work. Efficacy was associated with learning behavior on its own, but it did not add explanatory power once psychological safety was taken into account. Later research found that efficacy can act alongside learning behavior as a pathway through which psychological safety relates to team effectiveness (Baek et al., 2020). The practical logic is that people are most likely to speak up when they feel safe from interpersonal punishment and believe the team can do something useful with what they say.
Effective efficacy-building approaches include:
Set clear, compelling goals: Teams need to believe their work matters and that success is achievable.
Build early wins: Initial successes, even small ones, create confidence that compounds over time.
Provide skill development: Invest in both technical and teamwork capabilities.
Connect effort to impact: Help teams see how their learning behaviors translate into tangible outcomes.
Toyota's production system illustrates the pairing of high standards with safety to speak up. Its andon system allows workers on the assembly line to signal a problem and, if necessary, stop the line when they detect a quality issue (Liker, 2004). The practice works because signaling a problem is treated as a contribution to quality rather than a disruption or a personal failure. High expectations for performance and continuous improvement are paired with a system that makes raising problems normal and expected.
Transparent Communication and Organizational Narrative
How teams and organizations tell their story matters. Edmondson's (1999) qualitative data showed that high-learning teams described their experiences in terms of shared purpose and mutual support, while low-learning teams described theirs in terms of blame and self-protection. Such narratives are not mere byproducts of psychological safety; they actively sustain or erode it.
Effective communication approaches include:
Articulate shared values explicitly: Teams benefit from explicit conversations about who they want to be as a group.
Share stories of productive failure: When leaders discuss their own mistakes and what they learned, they normalize the link between error and growth.
Co-create team norms: Invite team members to agree on how they will interact, disagree, and support one another.
Use regular check-ins: Brief, structured moments to surface concerns, risks, or requests for help build a habit of candor.
Microsoft offers an example at organizational scale. After becoming chief executive in 2014, Satya Nadella sought to shift the company's culture from a "know-it-all" orientation to a "learn-it-all" one, drawing on research on growth mindset and emphasizing curiosity and learning from mistakes (Nadella et al., 2017). The consistent narrative from the top signaled that admitting what one did not know was a strength rather than a weakness, giving teams permission to experiment and question existing approaches.
Building Long-Term Organizational Learning Capacity
Psychological Contract Recalibration
Psychological safety does not develop once and persist forever. Edmondson (1999) argued that it is shaped by a team's history—by events that show members whether interpersonal risk is worthwhile. Changes in leadership, membership, strategy, or external conditions can all disrupt an established climate. Psychological safety is therefore maintained through the same behaviors that create it: consistent modeling, productive responses to candor, and visible consequences when norms of respect are violated.
For organizations, this means psychological safety cannot be treated as a one-time initiative. It must be embedded in leadership development, onboarding, performance management, and the routines of daily team interaction. Leaders can periodically assess team climate, using instruments such as Edmondson's (1999) scale, and treat declining scores as early warning signals.
Distributed Leadership for Sustainable Safety
Psychological safety need not depend on a single formal leader. Edmondson's (1999) qualitative data included teams in which members, rather than the formal leader, generated much of the team's learning culture. This suggests that the behaviors that build safety can be distributed across the team.
This has important implications for organizational design:
Develop coaching capability at every level: Invest in facilitation and coaching skills for individual contributors who serve as informal leaders, not only for managers.
Rotate facilitation roles: Allowing different members to lead retrospectives, feedback sessions, or problem-solving discussions builds collective ownership of the team's learning culture.
Value relational competence in promotion: Promote and reward leaders who create safety as well as those who deliver results; the evidence suggests the two are intertwined (Frazier et al., 2017).
Continuous Learning Systems and Measurement
Edmondson's (1999) cross-sectional design could not capture the self-reinforcing cycles that likely characterize psychological safety's development. She proposed that psychological safety enables learning behavior, and that learning behavior which goes unpunished in turn reinforces psychological safety. Understanding and managing these cycles requires ongoing feedback mechanisms, measurement practices, and reflective routines, not just training events.
Edmondson's (2019) later work offers a useful framework: a two-by-two matrix crossing psychological safety with performance standards or accountability. High safety with high standards produces a learning zone, in which teams aim high and speak candidly about progress. High safety with low standards produces a comfort zone; low safety with high standards produces an anxiety zone; and low levels of both produce apathy.
Organizations seeking to sustain learning should:
Measure psychological safety regularly: Use validated team-level instruments and track trends over time.
Pair safety with accountability: Accompany safety initiatives with clear performance expectations, rather than treating them as alternatives to rigor.
Create forums for cross-team learning: Let high-learning teams share practices so that safety-enabling norms spread.
Study change over time: Track how team climate evolves through leadership transitions, reorganizations, and new technologies.
Conclusion
More than two decades after Edmondson's (1999) foundational study, research on psychological safety has only grown more relevant. The construct she defined—a shared belief that the team is safe for interpersonal risk taking—has proven to be one of the most robust and practically consequential findings in organizational behavior, with meta-analytic evidence linking it to learning, information sharing, engagement, and performance (Frazier et al., 2017).
Yet knowing about the construct is not the same as practicing it. Many organizations profess to value psychological safety while their daily practices undermine it, and the gap between stated values and lived experience remains wide.
The evidence suggests several practical takeaways for leaders:
Psychological safety is a team-level climate that leaders shape daily: It is built through how leaders respond to questions, mistakes, and dissent in real time, not through workshops or posters.
Learning behavior is the mechanism: Safety without expectations for feedback seeking, experimentation, and error discussion risks producing comfort without improvement.
Structure matters but is not destiny: Resources and thoughtful design help, but team climate also depends on interpersonal practice.
Safety is not the absence of accountability: The best teams combine high safety with high standards, creating a learning zone in which candor and rigor coexist.
Sustaining safety requires ongoing investment: Embed it in leadership development, measurement, and daily team routines.
The organizations that thrive in the coming decades will be those in which people feel confident enough to say "I made a mistake," "I don't understand," or "What if we tried something different?"—and are met with curiosity rather than punishment. That is the promise of psychological safety, and the evidence suggests it is a promise worth keeping.
Research Infographic
References
Baek, P., Kim, N., & Heo, G. (2020). How psychological safety affects team performance: Mediating role of efficacy and learning behavior. Frontiers in Psychology, 11, Article 1581.
Duhigg, C. (2016, February 25). What Google learned from its quest to build the perfect team. The New York Times Magazine.
Edmondson, A. 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.
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383.
Edmondson, A. C. (2019). The fearless organization: Creating psychological safety in the workplace for learning, innovation, and growth. Wiley.
Edmondson, A. C., & Lei, Z. (2014). Psychological safety: The history, renaissance, and future of an interpersonal construct. Annual Review of Organizational Psychology and Organizational Behavior, 1, 23–43.
Frazier, M. L., Fainshmidt, S., Klinger, R. L., Pezeshkan, A., & Vracheva, V. (2017). Psychological safety: A meta-analytic review and extension. Personnel Psychology, 70(1), 113–165.
Google. (2015). Guide: Understand team effectiveness. re:Work.
Janis, I. L. (1982). Groupthink: Psychological studies of policy decisions and fiascoes (2nd ed.). Houghton Mifflin.
Liker, J. K. (2004). The Toyota way: 14 management principles from the world's greatest manufacturer. McGraw-Hill.
Nadella, S., Shaw, G., & Nichols, J. T. (2017). Hit refresh: The quest to rediscover Microsoft's soul and imagine a better future for everyone. Harper Business.
Nembhard, I. M., & Edmondson, A. 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.
Staw, B. M., Sandelands, L. E., & Dutton, J. E. (1981). Threat-rigidity effects in organizational behavior: A multilevel analysis. Administrative Science Quarterly, 26(4), 501–524.
Jonathan H. Westover, PhD, Chief Research Officer (Nexus Institute for Work and AI); Co-Founder & Chief Workforce and Learning Officer (Future State University); Founder & CEO (Human Capital Innovations); Professor of Organizational Leadership & Change (UVU). Read Jonathan Westover's executive profile here.
Suggested Citation: Westover, J. H. (2026). Unlocking Team Potential: How Psychological Safety Transforms Learning, Performance, and Organizational Resilience. Human Capital Leadership Review, 39(4). doi.org/10.70175/hclreview.2020.39.4.4
