Psychological Safety Without Surveillance

Abstract

Amy Edmondson's foundational research on psychological safety established that teams performing at the highest level share a belief that interpersonal risk-taking is safe. But organizations that attempt to measure psychological safety through surveillance mechanisms — activity tracking, sentiment monitoring, covert behavioral scoring — systematically undermine the very conditions they seek to measure. This paper presents a privacy-first framework for measuring and building psychological safety at scale through opt-in signals, aggregated insights, and development-oriented recommendations.

Key Findings

• Surveillance-style measurement tools reduce the psychological safety they attempt to measure by creating fear and distrust.

• Opt-in signal collection with transparent purpose generates more accurate and actionable data than passive monitoring.

• Aggregated team-level insights preserve employee dignity while providing organizations with actionable health signals.

• Manager behavior — recognition frequency, response time, 1:1 consistency — is a leading indicator of team psychological safety.

• Explainable AI outputs that show reasoning and data sources are more trusted and acted upon by managers than black-box scores.

How this connects to RUDY