Measuring Balint Group learning effects in medical students: A validation of the Balint Group Questionnaire for Medical Students in English

Abstract Background Balint groups are an established method for strengthening doctor-patient relationships and building professional resilience, particularly in undergraduate medical education. Despite their increasing use, a validated English-language tool to measure the specific learning outcomes for medical students has been conspicuous by its absence. This study sought to validate an English translation of the German Balint Group Questionnaire (BGQ-E) for use within this population. Methods A cross-sectional validation study was undertaken involving 349 fourth-year medical students at the University of Birmingham. After their Balint group sessions, students completed the 15-item BGQ-E. To examine the instrument's psychometric properties and factor structure, a suite of analyses was conducted, including descriptive statistics, reliability analysis, as well as exploratory and confirmatory factor analysis (CFA). The CFA was specifically used to test various competing models, including two-factor, three-factor, and second-order factor structures. Results The BGQ-E demonstrated excellent internal consistency, with a Cronbach's alpha of 0.93. While the initial analysis pointed towards a three-factor structure, the CFA revealed that the factors were too highly correlated, indicating a lack of discriminant validity. The model that provided the best fit was a second-order factor model with a dominant general factor, which has been termed "Balint Process and Learning," and two distinct, albeit related, subscales: "Individual Learning" and "Group Learning". All items loaded significantly onto their respective factors, and both subscales showed strong internal consistency (alpha values of 0.92 and 0.79, respectively). Conclusion The BGQ-E is the first validated English-language tool for measuring the effects of Balint groups in a medical student population. It is a reliable instrument with a robust underlying structure. For research and evaluation, it is recommended to use the total score as the primary measure, with the two subscales providing valuable supplementary insight into the dynamics of individual versus group learning. Future research should focus on longitudinal validation and testing this tool in diverse healthcare populations.