Benchmarking Psychological Safety: The Unseen Foundation of Procedural Adherence

When a procedural breach leads to a near-miss, the immediate response is often to review the written steps, retrain the operator, or tighten supervision. But in many cases, the root cause is not a lack of knowledge or a poorly written procedure — it is a team culture where people hesitate to speak up when something feels off. Psychological safety, the collective belief that one can take interpersonal risks without fear of negative consequences, is the invisible enabler of consistent procedural adherence. This article explores how teams can benchmark this dimension using qualitative indicators, without relying on anonymous surveys alone.

We focus on high-hazard settings: chemical plants, aviation maintenance, healthcare, and energy operations. In these environments, procedures exist to prevent catastrophic failures. Yet even the best procedures fail if the people executing them are unwilling to question ambiguity, report errors, or challenge authority. Benchmarking psychological safety means looking for patterns in everyday interactions — not just in annual engagement scores.

This content is for general informational purposes only and does not constitute professional safety consulting advice. Organizations should consult qualified human factors specialists for site-specific assessments.

Why This Topic Matters Now

The past decade has seen a shift from compliance-driven safety to culture-driven safety. Regulators and industry bodies increasingly recognize that procedural adherence cannot be enforced through audits alone. A 2023 survey of safety professionals (industry-wide, not a single study) found that over 60% of respondents identified “fear of speaking up” as a contributing factor in recent incidents. Meanwhile, high-reliability organizations like nuclear submarines and air traffic control have long embedded psychological safety into their operating principles.

For most organizations, however, the concept remains abstract. Leaders ask: “How do we know if we have enough psychological safety?” Without a benchmark, teams either assume it exists (because no one complains) or dismiss it as a soft skill irrelevant to hard procedures. The cost of ignoring this foundation is visible in repeated procedural deviations, unreported hazards, and a culture where silence is mistaken for compliance.

Consider a composite scenario: In a chemical processing plant, a technician notices that a pressure gauge reading is slightly outside the acceptable range before starting a batch. The procedure says to proceed if the reading is within 5% of the setpoint, but the technician has seen similar readings precede a valve failure. In a psychologically safe team, they would call the shift supervisor and ask for a double-check. In a less safe team, they might proceed — or quietly log a slightly different number. The outcome depends on the culture, not the procedure.

Benchmarking psychological safety gives teams a way to detect these invisible barriers before they lead to incidents. It is not about measuring a single number, but about observing patterns over time: how often people question, how leaders respond, and what happens after an error is reported.

What the Research Suggests

While we avoid citing specific studies, extensive field observations from human factors practitioners indicate that teams with high psychological safety have lower procedural deviation rates, faster incident reporting, and more effective learning from near-misses. These patterns are consistent across industries, from offshore drilling to hospital ICUs. The challenge is translating these insights into actionable benchmarks.

Who Should Pay Attention

This article is for safety managers, team leads, human factors engineers, and anyone responsible for procedural compliance in high-risk environments. It is also relevant for organizational development professionals who support safety culture initiatives. If you have ever wondered why a team that “knows the procedure” still cuts corners, the answer often lies in the unseen dimension of psychological safety.

Core Idea in Plain Language

Psychological safety is not about being nice or avoiding conflict. It is about creating conditions where people can express concerns, admit mistakes, and challenge assumptions without damaging their standing in the group. In the context of procedural adherence, this means that when a worker sees a discrepancy between the written step and the real-world conditions, they feel empowered to stop and ask — rather than assuming they must follow the procedure blindly or risk being seen as incompetent.

Think of it as a permission structure. Procedures are the map, but psychological safety is the permission to say, “I think the map is wrong,” or “I need help understanding this step.” Without that permission, people follow the map even when they see a cliff ahead. This is not a hypothetical: many major accidents, from chemical releases to aviation mishaps, have involved crews who followed procedures to the letter but never voiced their doubts because the culture did not welcome them.

Benchmarking psychological safety, therefore, is about assessing the strength of this permission structure. It involves looking at real behaviors, not just stated values. For example, in a team meeting, do junior members speak up when a senior person proposes a plan that contradicts a known procedure? In a post-incident review, does the focus shift to blame or to understanding system factors? These observable moments are the raw data for benchmarking.

Key Components of Psychological Safety for Safety Systems

We can break down psychological safety into four observable dimensions relevant to procedural work:

• Speaking up about errors: Will team members report their own mistakes or near-misses without fear of punishment?
• Questioning authority: Can a junior person challenge a senior’s decision regarding a procedure?
• Admitting uncertainty: Do people feel safe to say “I don’t know” or “I need clarification” in front of peers?
• Offering divergent views: Are alternative approaches to a procedure discussed openly, or are they suppressed?

Each dimension can be benchmarked through qualitative observation and targeted questions during debriefs. The goal is not to assign a score but to identify gaps that could undermine procedural adherence.

Why It Is Often Invisible

Psychological safety is hard to see because the absence of speaking up looks like agreement. A team where no one questions a procedure may appear compliant, but the silence could mask unspoken concerns. The only way to detect it is to create opportunities for people to speak and observe whether they do. This is why benchmarking relies on behavioral markers, not just self-report surveys.

How It Works Under the Hood

Benchmarking psychological safety requires a systematic approach to observing and interpreting team interactions. It is not a one-time audit but an ongoing practice. The core mechanism involves three steps: identifying benchmark indicators, collecting data through naturalistic observation, and analyzing patterns against a reference model.

First, teams need to define what “good” looks like in their context. For a safety-critical team, good psychological safety might mean that during a pre-job briefing, at least one person raises a concern about the procedure. In a less safe team, the briefing might be silent except for the leader. These indicators are context-specific but share common themes: the frequency of questions, the tone of responses, and the follow-up actions after a concern is raised.

Second, data collection should be embedded in existing routines. After a shift handover, supervisors can note how many times a procedure was questioned. During incident investigations, facilitators can track whether the discussion focuses on systemic causes or individual blame. These observations can be recorded in a simple log, without needing complex software.

Third, analysis involves comparing current patterns to a desired benchmark. For instance, if the benchmark is “every team member has spoken up at least once in the past month,” and the data shows only senior members have spoken, that indicates a gap. The analysis should look for trends over time, not just snapshots.

Common Benchmark Indicators

Practitioners often use the following qualitative indicators to gauge psychological safety:

• Voice-to-silence ratio: In meetings, what proportion of attendees contribute questions or concerns?
• Leader response type: When a concern is raised, does the leader thank the person, ask clarifying questions, or dismiss it?
• Error reporting rates: Are near-misses reported voluntarily, and do they increase after a safety intervention?
• Post-incident language: Do reports use blame-oriented language (e.g., “failed to follow procedure”) or system-oriented language (e.g., “procedure did not account for condition X”)?

These indicators are not perfect, but they provide a starting point for conversation. The key is to use them consistently and compare across teams or time periods.

The Role of Leadership

Leaders set the tone for psychological safety. A leader who penalizes questions, even subtly, will quickly suppress speaking up. Conversely, a leader who explicitly invites concerns and responds with curiosity reinforces the permission structure. Benchmarking should therefore include leader behaviors: how often do they ask open-ended questions, acknowledge their own uncertainty, or thank people for raising issues?

One practical method is to have leaders self-assess after each team interaction, using a simple checklist: Did I create space for questions? Did I respond defensively? Did I follow up on a concern? Over time, these self-assessments can be aggregated to identify patterns.

Worked Example or Walkthrough

Let us walk through a composite scenario of a maintenance team in a power generation facility. The team is responsible for a critical turbine overhaul, following a detailed procedure with over 200 steps. Historically, the team has had a low rate of procedural deviations, but recent near-misses have raised concerns.

The safety manager decides to benchmark psychological safety over a four-week period. They use the following approach:

1. Observation during pre-job briefings: The manager attends three briefings and notes how many questions are asked by technicians versus the lead. In the first briefing, only the lead speaks; technicians nod. In the second, one technician asks about a torque specification that seems low. The lead checks the procedure and confirms it is correct, thanking the technician. In the third, two technicians raise concerns about access to a bolt.
2. Review of incident reports: The manager looks at near-miss reports from the past three months. Out of 12 reports, 10 were filed by supervisors, only 2 by technicians. This suggests that technicians are not reporting their own near-misses.
3. Post-shift debriefs: The manager implements a 5-minute debrief after each shift, asking: “Did anything about the procedure feel unclear or risky today?” Initially, responses are minimal. By the third week, technicians start mentioning small issues, like a missing tool or a step that was hard to read.

After four weeks, the manager compiles observations. The pattern shows that while speaking up is increasing, it is still mostly about minor issues. No one has questioned a major step or challenged the lead’s decision. The benchmark suggests moderate psychological safety — enough to report small problems, but not yet enough to challenge authority or admit significant uncertainty.

The manager shares this with the team and together they identify actions: the lead will explicitly ask for challenges on critical steps, and the team will practice “stop and discuss” scenarios during training. Three months later, a second benchmark shows increased questioning and a near-miss report from a technician that prevents a potential bolt failure.

Lessons from the Walkthrough

This example illustrates that benchmarking is iterative and requires patience. The initial silence was not a sign of compliance but of low psychological safety. By creating structured opportunities and observing responses, the team was able to surface hidden concerns. The key was not to judge but to gather data and act on it.

Adapting to Different Team Sizes

In larger teams, benchmarking may focus on sub-teams or shifts. It is important to compare like with like: a night shift may have different dynamics than a day shift. The indicators should be consistent across comparisons to avoid misinterpretation.

Edge Cases and Exceptions

Benchmarking psychological safety is not straightforward in all contexts. Some teams face structural barriers that suppress speaking up regardless of culture. For example, in highly hierarchical organizations, such as military units or traditional manufacturing, the chain of command may discourage questioning. In these settings, even a psychologically safe sub-team may still hesitate because of external pressures.

Another edge case is when psychological safety is high but procedural adherence is still low. This can happen if the team feels safe to deviate from procedures without consulting others. In such cases, the culture may have shifted from “safe to speak up” to “safe to ignore rules.” Benchmarking must therefore distinguish between constructive speaking up and unconstrained rule-breaking. The indicators should focus on whether concerns are raised before action, not after.

Cultural differences also play a role. In some national or organizational cultures, direct questioning of authority is seen as disrespectful. Teams may express concerns indirectly, through hints or body language. Benchmarking methods must be sensitive to these communication styles. For instance, observing whether people use hedging language (“I might be wrong, but…”) can still indicate psychological safety, even if the tone is deferential.

Finally, there is the risk of over-surveying. If teams are asked to fill out lengthy psychological safety questionnaires every month, they may become fatigued and provide rote answers. Qualitative benchmarking, when done lightly, avoids this pitfall. The goal is to integrate observation into normal work, not add extra burden.

When Not to Benchmark

If an organization is in the midst of a major crisis, such as a layoff or a regulatory investigation, psychological safety may be temporarily low due to external stress. Benchmarking during such periods may yield misleading results. It is better to wait until the situation stabilizes, or to interpret results with caution.

Similarly, if leadership is not committed to acting on the findings, benchmarking can backfire. Team members may perceive it as surveillance or a check-the-box exercise. Before starting, ensure there is a clear plan for how results will be used to improve conditions.

Limits of the Approach

Qualitative benchmarking has inherent limitations. It relies on observer interpretation, which can introduce bias. Two observers might rate the same interaction differently. To mitigate this, teams should use multiple observers and calibrate their assessments through discussion. A simple rating scale (e.g., low, medium, high) with behavioral anchors can improve consistency.

Another limit is that benchmarking captures only observable behavior, not internal beliefs. A person may feel psychologically safe but choose not to speak up for other reasons, such as fatigue or lack of time. The benchmark may underestimate the true level of safety. Conversely, someone may speak up frequently but still feel unsafe — they may be high in assertiveness but low in trust. The benchmark may overestimate safety.

Moreover, benchmarking is not a substitute for deeper culture change. It can identify gaps, but closing them requires sustained effort in leadership development, process redesign, and trust-building. A team that scores well on one benchmark may regress if leaders change or if external pressures increase.

Finally, there is no universal benchmark that fits all teams. What works for a surgical team may not work for a refinery crew. Each team must define its own indicators and reference points. This customization makes it harder to compare across organizations, but it improves relevance for the team itself.

Balancing Qualitative with Quantitative

While we emphasize qualitative methods, some teams find value in combining them with simple quantitative measures, such as the percentage of meeting time devoted to questions, or the number of near-miss reports per quarter. These numbers add rigor but should not replace the nuanced understanding gained from observation. The best approach is to use both, with qualitative insights driving interpretation of the numbers.

Reader FAQ

Q: How often should we benchmark psychological safety?
We recommend a baseline assessment, then periodic check-ins every 3-6 months. More frequent checks may be useful during major changes, such as new leadership or new procedures.

Q: Can we benchmark without a dedicated human factors specialist?
Yes. Many teams start with a simple log of speaking-up events during meetings. The key is consistency and a willingness to reflect on the data. However, if the team is large or the stakes are high, involving a specialist can improve objectivity.

Q: What if our benchmark shows low psychological safety? What then?
Low scores are not a failure; they are a starting point. Focus on one or two specific behaviors to improve, such as leaders inviting questions. Small changes, like a leader pausing for 10 seconds after asking for concerns, can have a big impact.

Q: How do we handle a team that resists benchmarking?
Explain the purpose clearly: it is not about judging individuals but about making the team safer. Involve team members in choosing indicators. If resistance persists, start with anonymous methods, such as a suggestion box, and gradually move to more open discussions.

Q: Can psychological safety be too high?
In theory, yes, if it leads to complacency or excessive informality. But in practice, most teams are far from that point. The risk of low safety is much greater. If you see signs of overconfidence, refocus on the purpose: speaking up to improve adherence, not to bypass it.

Q: Is benchmarking enough to improve psychological safety?
No. Benchmarking is a diagnostic tool. Improvement requires deliberate action: training, feedback, and changes in leadership behavior. Use the benchmark to identify priorities, then invest in those areas.

Q: How do we know if our benchmarks are accurate?
Cross-check with multiple sources: observation, incident reports, and informal conversations. If all point in the same direction, you can be more confident. If they conflict, investigate further. Accuracy improves with practice and calibration.