Who Will Succeed in Dental School? Predictors of dental school performance

Abstract

Background: Selection of students with the highest potential of success is a very challenging process because selection is carried out among a highly academically qualified pool of applicants exceeding the number of places available. Additionally, evidence about the incremental and predictive validity of admission assessments and personal attributes assessed at admission is limited.

Objectives: To address this, a systematic review for evidence of the predictive validity of selection methods was completed, the incremental and predictive validity of admission assessments and whether any of the assessments are biased towards or against certain individuals were then investigated.

Methods: This was a retrospective cohort study using data of four cohorts at the University of Leeds, School of Dentistry. Data analysis included univariate and multivariate analysis. Outcome measures included academic and clinical performance. Predictor measures included personal statement, BMAT and MMI scores in addition to the socio-demographic characteristics of participants.

Results: Hierarchical regression models revealed that BMAT was the only admission assessment that contributed significantly in increasing the variance. Sections 3 and 2 were the most predictive. Additionally, MMI and BMAT significantly predicted on course 3rd to 5th year clinical practice and 2nd to 3rd year academic scores. Empathy, communication, insight and presentation stations were the most predictive of students’ performance. None of the admission assessments showed evidence of bias against gender, widening participation or ethnic groups.

Conclusion: The findings demonstrated evidence of incremental and predictive validity of BMAT as an admission test. They also revealed the necessity to re-evaluate the MMI structure, particularly the skills assessed and the tasks used to assess them, to improve its validity. The research has also highlighted the need to identify and provide appropriate support to individuals at greater risk of low performance and the necessity for adequate admissions data management at the University to facilitate future studies.