Choosing tomorrow’s rural doctors: An analysis of medical student selection in an Australian rural context

This thesis explores the evolving landscape of medical school admissions, focusing on the integration of non-academic selection tools to support equity and rural workforce sustainability. Against a backdrop of shifting societal expectations and educational reform, the study examines how non-academic measures contribute to identifying “fit for purpose” candidates, likely to thrive in training and serve underserved communities upon graduation. This case study of the University of Wollongong’s MD program, in Australia, investigates how admissions processes can better align with social accountability, rural intent, and clinical competencies required for good medical practice.The literature review traces the evolution of medical admissions from academic focused models to more holistic frameworks that incorporate non-academic attributes. It critically examines the limitations of cognitive metrics such as GPA and GAMSAT, highlighting equity concerns and relevance to clinical performance. The review explores structured interviews, MMIs, SJTs, and emerging tools like psychological testing and virtual asynchronous interview models, reviewing the evidence of their psychometric properties and alignment with desired graduate competencies. Particular attention is given to rurality, socioeconomic disadvantage, and the imperative to select candidates who meet the needs of the community they serve. These insights inform the thesis’s focus on equitable, evidence-based selection strategies in medical education. The methodology and research approach is outlined in detail in Chapter 3.Building on a description of the selection methods used at UOW, Chapter 4 presents an analysis of academic and non-academic admissions measures, examining their relationships with demographic variables, rural origin, and educational background. Drawing on a dataset of over 19,000 applicants to the UOW MD program, the analysis reveals that GPA and GAMSAT demonstrate modest predictive validity for written assessments during medical school, but weaker associations with clinical performance. MMI scores varied significantly by rurality and degree background, with biomedical science applicants and rural candidates consistently underperforming. Gender and age influenced academic selection measures and interview outcomes in more nuanced ways. These findings highlight systemic disparities in selection outcomes and underscore the need for admissions frameworks that balance predictive validity with equity, particularly for applicants from structurally disadvantaged and geographically remote backgrounds.Chapter 5 analyses the predictive validity of MMI scores for clinical performance, using OSCE outcomes across 14 cohorts at UOW. MMI scores showed strong correlations with Year 3 and Year 4 OSCE performance, particularly in domains related to clinical reasoning and structured communication. Attribute-level analysis revealed targeted associations between MMI constructs, such as leadership, insight into medicine, and problem solving, and specific OSCE competency domains. These findings affirm the utility of MMI in identifying candidates with early-stage clinical competence and support its role in multimodal selection frameworks. The chapter also highlights construct overlap and blueprinting challenges in assessing non-academic attributes.Chapter 6 analyses the Casper situational judgment test as a non-academic admissions tool in medical selection. Using data from over 12,000 applicants, it explores Casper’s use in Australian medical school selection and evaluates associations with demographic, academic, and performance outcomes. Casper showed modest correlations with MMI and OSCE scores, and minimal overlap with GPA or GAMSAT. Regression analyses suggest Casper adds unique value but may be influenced by contextual factors such as age, gender and prior degree. These findings support Casper’s role in holistic admissions while underscoring the need for ongoing evaluation of fairness and predictive utility in diverse applicant cohorts.Chapter 7 analyses predictors of impeded academic progress in the UOW MD program, analysing demographic, academic, and experiential variables across 915 students. Lower GPA, GAMSAT, and MMI scores were all associated with academic difficulty, alongside older age, rural background, and prior degrees in science or humanities. Surprisingly, students with graduate degrees also showed increased risk of impeded academic progress. While statistically significant, these predictors offer limited utility for exclusionary selection. Instead, highlighting the need for targeted support mechanisms to uphold equity goals. The study underscores the tension between predictive selection and inclusive admissions, advocating for proactive interventions over restrictive thresholds.Together, these findings contribute to a growing body of evidence supporting the use of multifactorial, context-sensitive admissions frameworks in medical education. The thesis argues that selection tools must be evaluated not only for predictive validity but also for their alignment with institutional purpose, social accountability, and workforce needs. By integrating equity considerations into selection design medical schools can better fulfil their obligation to produce competent, compassionate, and community-responsive graduates that are “fit for purpose” for each school’s medical program and eventual community.