A Novel 5-Item Survey Tool to Assess Psychological Safety Among Graduate Medical Education Trainees

Background Psychological safety is a critical component of effective learning and team performance in medical education. However, no tool with validity evidence currently exists to measure it specifically among graduate medical education (GME) trainees. Objective To develop a concise, trainee-centered survey tool to measure psychological safety in GME settings and to examine its content and internal structure among graduate medical trainees. Methods We created 9 candidate items based on existing literature and expert feedback. These were incorporated into an annual institution-wide trainee survey at a single academic health care system. Responses were randomly divided into derivation and validation cohorts. Exploratory and confirmatory factor analyses were conducted to produce and obtain validity evidence for the most effective and concise set of items. Results Of the 480 trainees invited to take the survey, 400 (83%) returned completed responses. During the derivation phase (n=200), 4 items were removed for low tolerance, improving model fit. The final 5 items showed strong loading on a single factor. The 5 items assessed trainees’ comfort level with sharing ideas, giving and receiving feedback, expressing uncertainty, and viewing mistakes as learning opportunities. In validation testing (n=200), confirmatory factor analysis supported a unidimensional structure (Comparative Fit Index=0.99, Tucker-Lewis Index=0.97, and standardized root mean squared residual=0.02). The composite score correlated positively with an external measure of psychological safety (r=0.67) and demonstrated high internal reliability (McDonald’s omega=0.91). Conclusions The scores on this 5-item scale demonstrated evidence of validity and reliability for assessing psychological safety among trainees.