Uncertainties in the selection of applicants for medical school.

Decisions about admissions to medical school are based on assessments of the applicants' cognitive achievements and non-cognitive traits. Admission criteria are expected to be fair, transparent, evidence-based and legally defensible. However, unlike cognitive criteria, which are highly reliable and moderately valid, the reliability and validity of the non-cognitive criteria are low or uncertain. Their uncertain predictive value is due not only to their limited validity, but also to the unknown prevalence of the desirable non-cognitive traits in the applicants' pool. Consequently, the use of non-cognitive admission criteria inevitably leads to rejection of an unknown proportion of applicants who have a desirable trait and selection of applicants who lack this trait. We propose that, rather than using non-cognitive admission criteria, admission officers should assist prospective applicants to make informed decisions based on a reflective self-appraisal whether or not to apply to medical school. To this end, medical schools should disseminate information on the strains of medical training and practice, the frequency of medical errors and the most common causes of dissatisfaction and burn-out among practicing physicians. Such information may improve the self-selection process and thereby enrich the applicants' pool for individuals with appropriate motivation. The final selection of medical students may then be based either on past academic achievements, or on a lottery, or on various combinations thereof.