How to Measure Psychological Safety: Survey Questions and Metrics

Google's Project Aristotle studied 180 teams over two years to identify what makes teams effective. Technical skills and individual intelligence mattered less than expected. The strongest predictor of team performance was psychological safety.

Harvard Business School professor Amy Edmondson defined the concept in 1999: psychological safety is "a shared belief held by members of a team that the team is safe for interpersonal risk taking." In practical terms, it means people feel comfortable speaking up, admitting mistakes, asking questions, and proposing ideas without fear of punishment or embarrassment.

This guide covers how to measure psychological safety in your organization, what the data means, and how to improve it.

Why Psychological Safety Matters

Teams with high psychological safety outperform on measures that matter to organizations:

Innovation: People share unconventional ideas when they won't be ridiculed for "stupid questions." Research from Edmondson's lab shows psychological safety predicts team learning behavior and innovation output.

Error reporting: In healthcare studies, units with higher psychological safety reported more errors—not because they made more mistakes, but because staff felt safe flagging problems. Hidden errors compound; reported errors get fixed.

Retention: Employees stay where they feel valued and heard. Psychological safety correlates with lower turnover, particularly among high performers who have options elsewhere.

Decision quality: Teams that surface dissent make better decisions. Without psychological safety, groupthink dominates and bad ideas go unchallenged.

Measuring Psychological Safety: The Core Questions

Amy Edmondson's original research used seven items to assess team psychological safety. These remain the gold standard:

The Edmondson 7-Item Scale

Respondents rate agreement on a 1-7 scale (strongly disagree to strongly agree):

1. If you make a mistake on this team, it is often held against you. (reverse scored)

2. Members of this team are able to bring up problems and tough issues.

3. People on this team sometimes reject others for being different. (reverse scored)

4. It is safe to take a risk on this team.

5. It is difficult to ask other members of this team for help. (reverse scored)

6. No one on this team would deliberately act in a way that undermines my efforts.

7. Working with members of this team, my unique skills and talents are valued and utilized.

Scoring: Reverse items 1, 3, and 5 (subtract from 8). Average all seven items. Higher scores indicate higher psychological safety.

Alternative Question Approaches

While the Edmondson scale is validated and widely used, some organizations prefer different framings:

Behavioral Questions

Instead of asking about feelings, ask about behaviors:

• "In the past month, have you held back an idea or question because you were worried about how it would be received?"

• "In the past month, have you seen someone on your team criticized or embarrassed for asking a question or making a suggestion?"

• "When was the last time you disagreed with a decision and spoke up about it?"

Behavioral questions reduce social desirability bias. People might claim they "feel safe" while their actual behavior suggests otherwise.

Manager-Specific Questions

Manager behavior drives psychological safety more than any other factor:

• "My manager responds constructively when I bring up problems."

• "My manager admits when they don't know something."

• "I've seen my manager acknowledge their own mistakes openly."

• "My manager asks for input before making decisions that affect the team."

Interpreting Your Results

Team-Level Benchmarks

Based on aggregated research data and organizational surveys:

Most teams score between 4.0 and 5.5. Scores above 6.0 are rare and indicate exceptional team dynamics.

What to Look For in the Data

Variation across teams: The interesting insight isn't company average—it's the spread. Some teams will score 3.5 while others hit 6.0. Understanding why matters more than knowing the overall number.

Correlation with manager: Compare psychological safety scores with manager tenure and performance. Often you'll find that specific managers consistently lead high-scoring or low-scoring teams.

Demographic patterns: Check if psychological safety varies by gender, race, tenure, or role. Gaps suggest inclusion problems that affect specific groups.

Reverse-scored item patterns: If items about mistakes being "held against you" or "rejection for being different" score poorly while other items score well, you have specific problems rather than general culture issues.

Red Flags in Psychological Safety Data

Watch for these warning signs:

High variability within team: If half the team rates psychological safety at 6 and half at 3, something's wrong. Healthy teams show consensus around shared experience.

Gap between manager assessment and team assessment: Managers often rate their team's psychological safety higher than team members do. A gap over 1.0 points suggests the manager lacks awareness.

Decline over time: Psychological safety can erode quickly with leadership changes, layoffs, or punitive responses to mistakes. Monitor trends, not just snapshots.

Low scores combined with low participation: If only 40% of the team responds and scores are low, the real situation is probably worse. People who don't feel safe often don't respond to surveys about feeling safe.

Improving Psychological Safety

Measurement without action erodes trust. Here's what actually moves the needle:

Leader Behavior Changes

Model vulnerability: Leaders who admit mistakes, say "I don't know," and ask for help normalize these behaviors for everyone. This requires genuine behavior change, not scripted vulnerability.

Respond well to bad news: How leaders react the first time someone shares a problem determines whether they'll ever share problems again. Thank people for flagging issues, even when the issues are frustrating.

Separate learning from blame: When things go wrong, focus post-mortems on system improvement rather than individual fault. "What can we change?" beats "Whose fault was this?"

Ask questions, pause before answers: Leaders who immediately provide answers signal that others' input isn't needed. Asking genuine questions and waiting through uncomfortable silence creates space for others.

Structural Changes

Create explicit feedback channels: Anonymous suggestion systems, regular retrospectives, and skip-level meetings give people multiple avenues to speak up.

Normalize disagreement: Assign "devil's advocate" roles in meetings. Explicitly ask "What are we missing?" and "Who sees this differently?"

Protect early-stage ideas: Create forums where half-baked ideas are welcome. Innovation requires space for ideas that aren't yet fully formed.

Address interpersonal problems: When someone consistently shuts others down, address it directly. Tolerating behavior that undermines safety sends a message.

Psychological Safety and Performance Accountability

A common concern: "If we make everyone feel safe, won't people stop trying?"

This misunderstands the concept. Psychological safety isn't about lowering standards or avoiding difficult conversations. High-performing teams combine psychological safety with high expectations.

Edmondson's research identifies four zones:

The goal is the Learning Zone: people feel safe to take risks AND are held to high standards. The Anxiety Zone (high standards, low safety) burns people out and suppresses the honesty needed to improve.

Holding people accountable for results while making it safe to discuss problems, admit struggles, and ask for help isn't contradictory—it's the combination that drives performance.

Building a Measurement Practice

Recommended Cadence

• Annual: Full psychological safety assessment with all seven Edmondson items plus behavioral questions
• Quarterly: 2-3 key items as pulse check (items 2, 4, and 6 work well for tracking)
• Post-incident: After major events (reorgs, project failures, leadership changes), quick check on key indicators

Who Sees the Data

Define data access clearly before surveying:

• Individual responses: No one except, in some cases, the individual themselves
• Team-level data: Team manager (with minimum response threshold), HR partners, skip-level leaders
• Organization-level data: Executive team, board if relevant

Transparency about data access builds trust. Ambiguity destroys it.

Connecting to Action

Share results with teams within two weeks of survey close. Present data without defensiveness. Ask teams to interpret their own results and propose improvements. Follow up on commitments visibly.

The measurement itself is easy. The cultural change it requires is hard. But organizations that commit to psychological safety—that make it a genuine priority rather than a metric to manage—see the outcomes: better decisions, faster learning, stronger retention, and teams that actually want to do great work together.