Assessment of observed general practice trainee performance: Confirmatory factor analysis of the General Practice Registrar Competency Assessment Grid

BACKGROUND AND OBJECTIVES: Validity of workplace-based assessments (WBA) is important, yet rare in general practice. The General Practice Registrar Competency Assessment Grid (GPR‑CAG) assesses registrar in‑consultation performance, as observed during external clinical teaching visits. GPR‑CAG content validity and initial internal consistency is established. This study aimed to confirm the GPR‑CAG factor structure. METHOD: Confirmatory factor analysis was conducted using factors and items from an exploratory factor analysis. Separate analyses were conducted for Term 1 (GPT1; 16 items) and Term 2 (GPT2; 27 items) registrar data (n=827 and n=796, respectively). RESULTS: Analyses supported the existing GPT1 GPR-CAG factor structure, with fit indices showing good fit of the data to the model and most factor loadings being moderate‑to-high. For GPT2, a single model change was strongly supported, after which model fit was good, the factor structure coherent and most item loadings moderate-to-high. DISCUSSION: This study supports the GPR-CAG as a psychometrically robust, evidence-based WBA.