Early identification of first-year students at risk of dropping out of high-school entry medical school: the usefulness of teachers’ ratings of class participation

• Published in: Advances in health sciences education : theory and practice Vol. 24; no. 2; pp. 251 - 268
• Main Authors: Araújo, Alexandra M., Leite, Carlos, Costa, Patrício, Costa, Manuel João
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.05.2019; Springer; Springer Nature B.V
• Subjects: Academic Failure; Academic Success; Adolescent; Age Factors; At Risk Students; Behavior; College students; Colleges & universities; Dropout Characteristics; Dropout Prevention; Dropout Rate; Education; Education, Medical, Undergraduate - statistics & numerical data; Female; Health sciences; Higher education; Humans; Hypotheses; Interpersonal Relations; Learner Engagement; Learning; Longitudinal Studies; Male; Medical Education; Medical Schools; Medical students; Portugal; Potential Dropouts; Ratings & rankings; Residence Characteristics; School Admission Criteria - statistics & numerical data; School dropouts; Schools, Medical - statistics & numerical data; Sex Factors; Socioeconomic Factors; Stress, Psychological - epidemiology; Student Attitudes; Student Dropouts - statistics & numerical data; Student Experience; Student Participation; Young Adult; Portugal; Medical education; Engagement; First year; Class participation; Dropout
• DOI: 10.1007/s10459-018-9863-9
• ISSN: 1382-4996

Dropping out from undergraduate medical education is costly for students, medical schools, and society in general. Therefore, the early identification of potential dropout students is important. The contribution of personal features to dropout rates has merited exploration. However, there is a paucity of research on aspects of student experience that may lead to dropping out. In this study, underpinned by theoretical models of student commitment, involvement, and engagement, we explored the hypothesis of using inferior participation as an indicator of a higher probability of dropping out in year 1. Class participation was calculated as an aggregate score based on teachers’ daily observations in class. The study used a longitudinal dataset of six cohorts of high-school entry students (N = 709, 67% females) in one medical school with an annual intake of 120 students. The findings confirmed the initial hypothesis and showed that lower scores of class participation in year 1 added predictive ability to pre-entry characteristics (Pseudo- R 2 raised from 0.22 to 0.28). Even though the inclusion of course failure in year 1 resulted in higher explanatory power than participation in class (Pseudo- R 2 raised from 0.28 to 0.63), ratings of class participation may be advantageous to anticipate dropout identification, as those can be collected prior to course failure. The implications for practice are that teachers’ ratings of class participation can play a role in indicating medical students who may eventually drop out. We conclude that the scores of class participation can contribute to flagging systems for the early detection of student dropouts.