Editorial sample: This reading has not yet undergone Mind Assembly's editorial review. Verify all sources, interpretation and permissions before public release.
The foundational study
Edmondson (1999) studied 51 teams in one manufacturing company. The study examined psychological safety and team efficacy, the group’s belief that it could perform its work. Teams with higher psychological safety showed more learning behaviour, such as seeking feedback or discussing mistakes. The association held after accounting for team efficacy in the statistical model.
The study also reported a pattern consistent with learning behaviour linking psychological safety to team performance. That pattern supports the proposed explanation. Establishing the direction of the effect would require a design that can distinguish it from other explanations, such as a supportive team leader influencing both safety and learning.
What does a larger evidence base add?
Frazier et al. (2017) pooled evidence from 136 independent samples representing more than 22,000 people and nearly 5,000 groups. Their meta-analysis compared links between psychological safety, conditions such as leadership, and work outcomes. It gives readers a way to judge whether associations extend beyond one workplace. The pooled results still depend on how the original studies sampled teams, measured the variables and addressed alternative explanations.
Ask what level was measured
A survey response reflects an individual’s perception of a team. An average can represent a shared team perception if there is adequate agreement among members and the study explains how it assessed that agreement. Two teams can have the same average while one shows broad agreement and the other shows sharply different experiences. Those differences matter when people with less authority may face greater costs for speaking up.
Questions for a journal club
- How did the researchers assess psychological safety and learning behaviour? Could a common data source inflate their association?
- What evidence would help establish whether psychological safety precedes changes in learning behaviour?
- What might change if the same question were studied in a student group where peers also assess one another’s work?
Sources & further reading
- Edmondson (1999) — Psychological Safety and Learning Behavior in Work Teams ↗
- Frazier et al. (2017) — Psychological Safety, a Meta-Analytic Review and Extension ↗
- Edmondson & Bransby (2023) — Psychological Safety Comes of Age ↗