medical education, medical students, well-being measure, bifactor model, confirmatory factor analysis


Accreditation standards for U.S. medical education programming require that training programs promote trainee wellness and well-being, although constructs such as psychological distress and depression commonly serve as proxies for well-being. A direct measure of subjective well-being would be invaluable to inform programming efforts to promote medical trainees’ well-being and advance the study of the well-being construct itself. This study investigated the structural validity of subjective well-being as measured by the Well-Being Scale in a sample of 548 osteopathic medical students. Results of the confirmatory factor analysis indicated that subjective well-being is best represented by a bifactor model with a general subjective well-being factor and five well-being domain factors and that allowing items from two well-being domains to cross load brought significantly stronger model-data agreement. Some scale items were respecified to become pure indicators of the general well-being factor based on their patterns of factor loadings, and the final bifactor model brought an acceptable fit. To our knowledge, this study is among the initial research efforts to investigate the use of a holistic subjective well-being instrument in the medical student population. The use of a direct, multidimensional subjective well-being measure is foundational to medical education programs’ efforts to promote well-being, future investigations of predictors of medical students’ well-being, and longitudinal analyses of medical students’ well-being over their matriculation.