Student selection to a rural veterinary school. 2: predictors of student performance and attrition.

ABSTRACT

OBJECTIVE: To evaluate the extent to which current selection processes at Charles Sturt University, New South Wales, Australia, predict performance within and attrition from the veterinary degree program. DESIGN: Retrospective evaluation of application details and student performance data. METHODS: Database records of 424 students entering the veterinary program were retrieved from university records, including all students graduating (n = 356) or leaving the program without completing (n = 68) between 2005 and 2016. Demographic data were related to results of selection processes and achievement within the degree using univariable and multivariable general linear and logistic regression analyses. RESULTS: The grade point average achieved over the 6 years of the degree was influenced by academic achievement prior to entry, gender (females performed ≈ 2% better than males) and interview scores. Preceptor evaluation of final-year clinical performance was associated with interview scores, gender (males performed ≈ 4% better than females) and residential address at the time of application. Attrition for personal reasons was more common for female students and students who had completed a prior degree, and students who experienced academic or personal attrition had lower written application scores. Factors that have been previously linked to poor academic outcomes, including rural or low socioeconomic background, were not associated with adverse student outcomes. CONCLUSION: Selection processes did not systematically disadvantage students admitted to the degree program. Gender differences in personal attrition and academic and clinical performance warrant further evaluation. Factors in addition to academic aptitude predicted student success.