Lessons learned from 15 years of non-grades-based selection for medical school
Context: Thirty years ago, it was suggested in the Edinburgh Declaration that medical school applicants should be selected not only on academic, but also on non-academic, attributes. The main rationale behind extending medical school selection procedures with the evaluation of (non-academic) personal qualities is that this will lead to the selection of students who will perform better as a doctor than those who are selected on the basis of academic measures only. A second rationale is the expectation that this will lead to a representative health workforce as a result of reduced adverse impact. The aims of this paper are (i) to describe what can be learned about the use of selection criteria other than grades from over 15 years of Dutch experience and (ii) to summarise current knowledge on the issue of adverse impact in relation to non-grades-based selection. Methods: A narrative review was undertaken of the (published) evidence that has resulted from non-grades-based school-specific selection procedures in the Netherlands and from recent explorations of the effect of the use of non-grades-based selection criteria on student diversity. Results: The Dutch evidence is grouped into five key themes: the effect of participation in voluntary selection procedures, the assessment of pre-university extracurricular activities, the use of work samples, Dutch experiences with situational judgement tests and the effects of changing circumstances. This is followed by several lessons learned for medical schools that aim to increase their student diversity. Conclusion: Over the last 30 years, important steps towards reliable and valid methods for measuring non-academic abilities have been taken. The current paper describes several lessons that can be learned from the steps taken in the Dutch context. The importance of sharing evidence gathered around the globe and building on this evidence to reach our goal of predicting who will be a good doctor is acknowledged.