Student selection
Introduction
Selection seems deceptively easy: if there are more applicants than places, then simply choose the best applicants. In practice, things are rather more complicated. Selection may be:
• a vulnerable process legally and ethically
• open to challenge on grounds such as discrimination
• criticized by society at large
• under-resourced given the implicit expectations of society, the profession and medical schools.
Although student selection is traditionally concerned with entry to medical school, recent years have seen a growing interest in postgraduate selection, where similar problems apply and similar principles and methods can be used.
“For a man to be truly suited to the practice of medicine, he must be possessed of a natural disposition for it, the necessary instruction, favourable circumstances, education, industry and time. The first requisite is a natural disposition, for a reluctant student renders every effort vain.”Why select?
“Selection is of key importance to medical education. What sort of students are recruited at the beginning is a major determination of what kind of doctors come out at the end.”Selection of students by the medical school
The straightforward reason is to choose the best students. Although seemingly simple, that little word ‘best’ hides many subtleties and complexities.
Selection by applicants of medicine as a career
The pool of medical school applicants only contains those who have selected medicine as a career. The very many individuals who did not apply cannot be selected, even if they might have made excellent doctors.
“Although there are reasons for being anxious about medical school selection, not all of the blame can be laid at the door of the selectors. Self-selection and preselection out of the applicant pool is extensive.”Implicit selection of the medical schools by applicants
While schools are selecting students, they are also selecting us, studying a range of schools and deciding which to apply to. An excellent selection system is of little use if the best applicants apply elsewhere. Effective selection begins by encouraging the right students to apply.
Explicit selection of medical schools by applicants
When applicants receive offers from two or more medical schools, it is they who select medical schools, not vice versa.
Selection for a particular course
Increasingly, medical schools are developing courses with different emphases. Courses with large components of problem-based learning in small groups might prefer students who work together cooperatively rather than competitively.
The limits of selection
“Unfortunately the qualities which count for most in medicine are not precisely measurable. The measurable – examination performance at school – neither necessarily relates to these quantities nor guarantees intellectual potential. … In this sea of uncertainty it is not surprising that selection processes are imperfect and open to criticism or that the remedy is not immediately apparent.”
The limits of selection are easily shown mathematically. If selecting on a single criterion (such as intellectual ability) which has a normal distribution of ability, and with a selection ratio of two applicants per place, the optimal selection is shown in Fig. 43.1. The candidates are placed in rank order, and those above the median are selected.
Fig. 43.1 A simple model of selection when there is a single characteristic on which selection is taking place; those above the threshold are accepted, and those below are rejected.
The limits of selection become apparent when two or more criteria are introduced, for example, intellectual ability and communication skills, which are essentially uncorrelated. The distribution is now bivariate normal (see Fig. 43.2) and the aim is to select the best 50% of candidates on the joint criteria. The dashed lines indicate the median for each of the separate distributions.
Selecting candidates to be above a particular threshold on both criteria means they are in the top right-hand corner of the figure. The key point to realize is that the threshold on either criterion will be substantially below the median. In fact, with two independent criteria, selected candidates are only in the top 71% of the ability range, rather than the top 50%, and hence are less able on average than if either criterion on its own had been used. The same conclusion applies also if one allows compensation between the separate abilities (McManus & Vincent 1993). If medical student selection is based predominantly on academic achievement, then for nonacademic factors to be taken substantially into account, academic standards must be lowered.
Medical schools considering nonacademic attributes for selection rapidly develop long lists of desiderata, often containing 5, 10, 20 or even 50 components. The model of Fig. 43.2 can easily be extended to three, four, five or many criteria, when the limits of selection appear with a vengeance. Assuming the criteria are statistically independent, then Table 43.1 shows that as the number of criteria rise, so the proportion of candidates eliminated on any single criterion (shown in the second column) becomes ever smaller. To put it bluntly, ‘if one selects on everything, one selects on nothing.’
Table 43.1
The effects of selection on the basis of multiple criteria (assuming two applicants for every place)
*N: number of criteria; r: selection ratio (e.g. if r = 3, there are 3 applicants for each place and 1/3 applicants are accepted).
Reproduced from McManus IC, Vincent CA: Selecting and educating safer doctors. In Vincent CA, Ennis M, Audley RJ, editors: Medical Accidents, Oxford, 1993, Oxford University Press, pp 80–105.
• Selection should aim at a relatively small number of what can be called ‘canonical traits’ – the three or four stable characteristics that are likely to predict future professional behaviour and can be assessed reliably at medical school application.
• If schools currently select almost entirely on academic ability, then they will have to reduce academic standards in order to select effectively on nonacademic criteria.
• Selection should be recognized as being limited in its power. The really powerful tools for affecting change are education and training (McManus & Vincent 1993).
What are the canonical traits in selection?
“ ‘A’ levels tell us nothing about some of the most desirable attributes of the doctor. The four desiderata are technical competence, human sympathy, wisdom and experience.”Attempts have been made to identify canonical traits for selection (McManus & Vincent 1993).
Intellectual ability
Doctors probably cannot be too intelligent. Meta-analyses of selection in many different occupations show that general mental ability is the best predictor both of job performance and of the ability to be trained (Schmidt & Hunter 1998). Although claims are often made for some minimum threshold ability level which is ‘good enough’, systematic research suggests that ‘more is better’ (Arneson et al 2011).
Learning style and motivation
Students study as university students for many different reasons, and those motivations mean they adopt particular study habits and learning styles. In Biggs’s typology (Table 43.2), both deep and strategic learning (but not surface learning) are compatible with the self-directed, self-motivated approach to learning that is required in the lifelong learning needed in medical practitioners.
Table 43.2
Summary of the differences in motivation and study process of the surface, deep and strategic approaches to study
| Style | Motivation | Process |
| Surface | Completion of the course | Rote learning of facts and ideas |
| Focusing on task components in isolation | ||
| Fear of failure | Little real interest in content | |
| Deep | Interest in the subject | Relation of ideas to evidence |
| Vocational relevance | Integration of material across courses | |
| Personal understanding | Identification of general principles | |
| Strategic/Achieving | Achieving high grades | Use of techniques that achieve |
| Being successful | highest grades |
Communicative ability
Most complaints about doctors involve communication problems, and so it makes sense to include it in selection. Communicative ability has been developing throughout life, so those with poor communication skills even at age 17 will probably respond well to training. Assessment is not straightforward, but interviews, multiple mini interviews, questionnaires and situational judgement tests can all assess communication.
Personality
Many studies have examined the ‘big five’ personality traits of extroversion, neuroticism, openness to experience, agreeableness and conscientiousness. Schmidt and Hunter’s (1998) meta-analysis showed that the best predictor of job performance and trainability, after intellectual ability, was integrity or conscientiousness, not least because highly conscientious people tend to work harder and be more efficient and so gain more and better experience. Conscientiousness may, though, not be a good predictor when creativity or innovation is important. At medical school, conscientiousness better predicts achievement in basic medical sciences, rather than clinical studies or postgraduate activities such as research output (McManus et al 2003).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
