Abstract

Introduction: Although there are measurement tools in Spanish to evaluate teacher performance during the initial basic science training years or during medical specialization, there are not as many for the clinical training years, in which the main role of the teacher is to facilitate experiential learning. The MEDUC30 questionnaire is a Spanish tool developed by the Pontificia Universidad Católica de Chile University School of Medicine. It was constructed using the Stanford Faculty Development Program (SFDP) educational framework for evaluation of clinical teachers’ effectiveness by students. MEDUC30 has been used since 2004 at Pontificia Universidad Católica de Chile University and was previously studied using exploratory methods. Objective: To provide evidence of validity and reliability to support MEDUC30’s usefulness in Spanish-speaking contexts, using confirmatory analytical methods. Method: This is an analytical, longitudinal and retrospective study, in which an analysis was performed on 24,681 MEDUC30 questionnaires evaluating 579 clinical teachers. These were completed by medical students from 3rd to 7th year of study, from 2004 to 2015. The questionnaire’s structure was studied by exploratory (EFA) and confirmatory factor analysis (CFA). Measurement invariance was evaluated with multi-group CFA. Results: Four different models were compared, with a bifactor model being the best alternative to explain the data’s structure. It was composed of one general and six domain-specific factors: [i] Patient-Based Teaching, [ii] Communication of Goals, [iii] Evaluation and Feedback, [iv] Promotion of Understanding, Retention and Self-directed Learning, [v] Control of the Session, and [vi] Learning Climate. The overall reliability of MEDUC30 scores was excellent (Cronbach’s ˛ = .98, McDonald’s ω = .98), and that of the six specific factors was very good (Cronbach’s ˛ = .88---.95, McDonald’s ω = .78---.94). Measurement invariance extended over teacher gender, date, semester, year of study, clinical teaching setting, and length of clinical rotation, with all of these variables being sources of population heterogeneity.