Psychological Safety Diagnosis & Intervention

What it does

Measures a team's psychological safety — whether people fear speaking up, asking questions, or admitting mistakes — with psychometrics, not intuition, then turns it into a concrete intervention plan. The primary diagnostic is Amy Edmondson's 7-item Team Psychological Safety scale (1999, Admin. Sci. Q. — the field's gold standard; 3 reverse-coded items). It computes the team-level mean (never individual), flips reverse-coded items via 8 − raw, and scales to a 0–100 Psychological Safety Score (PSS): PSS = round((mean−1)/6×100). The score maps to four verdict bands — Strong (80–100), Adequate (60–79), Fragile (40–59), At-Risk (0–39) — aligned with Edmondson norms (typical team mean ~5.3–5.5/7; below the 3.5 midpoint signals concern).

It goes beyond a single number. Using the Fearless Organization 4-domain model (Risk & Failure, Open Conversation, Willingness to Help, Inclusion & Diversity) it opens the score into four quadrants and targets the weakest domain first. It labels the behavioral mechanism of low scores via Detert & Edmondson Implicit Voice Theories (authority taboo / proof threshold / retaliation belief / futility belief), and triangulates the survey against error-reporting behavior: high survey + zero error reports = social-desirability flag; low reporting is not a "good team" but, per the reporting paradox, may signal hiding.

Interventions are generated in leader-behavior → ritual → structure order and ranked by RICE = (Reach×Impact×Confidence)/Effort, with a SCARF (Rock 2008) threat lens and Universal Design accommodations for neurodiversity. Output: a score gauge, a 4-domain profile, ranked findings, and a 30/60/90-day plan.

When to use it

• You have survey results but can't interpret them ("is 3.9 good or bad?").
• Suspected silence / error-hiding; "why won't anyone speak up?" after a retro.
• A new-leader transition — building the trust base in the first 90 days.
• Post-incident culture repair — system learning instead of blame.
• Raising neurodivergent participation without exposing any individual.

Method / frameworks

Edmondson 7-item scale (1999), Fearless Organization 4-domain (2019), Detert & Edmondson Implicit Voice Theories (2011), the learning-behavior reporting paradox, Leader Inclusiveness (2006) + SCARF (2008), Universal Design accommodations, and RICE prioritization. Hard-stop ethics: no individual exposure, no reporting at n<5, no therapy, no clinical diagnosis, no turning survey data into performance penalties.