Education and debate

How to do itSelect medical students

David Powis, assistant dean, undergraduate education. 

Faculty of Medicine and Health Sciences, University of Newcastle, New South Wales 2308, Australia

The analysis on p 1111 by McManus of some of the factors affecting whether a candidate is offered a place at medical school in the United Kingdom¹ shows clearly that selectors haven't yet got it right.

What follows is a brief guide that might help medical schools and doctors to determine what they want from a selection procedurethat is, what knowledge, skills, and attributes need to be sought in potential medical students, and why. This determination requires data, specifically, on the progress rates of medical students and the reasons for failure or premature withdrawal. With such knowledge the guide may be used to show how admissions committees should devise and operate an appropriately objective student selection procedure.

	Assumptions

Medical school admissions procedures are often not selection procedures but an administrative exercise to limit the number of entrants to the number of places available on the course by means of a movable marks threshold. The assumption is that the greater the number of applicants for places in the medical school the higher the marks threshold needs to be set to limit entrance.

A true selections procedure should try to match applicants with the course of study to maximise the chance of successful completion and to produce a competent and effective graduate professional.

I have assumed in this article that the medical school curriculum is entirely appropriate for its purpose of educating and training future doctors.

	Determining the qualities required in medical students

A list of essential and desirable qualities with the required extent of each defines the profile of people sought for entry to medical school. To maximise the chance of students successfully engaging with the curriculum and completing the course we need to ask eight questions.

	Selection instruments and procedures

The next task is to choose the most appropriate techniques to obtain the information required. Selection instruments and procedures must be valid, reliable, and appropriately used. They should also be objective.

	Possible selection sequence

The following is a proposed sequence for selecting medical students.

Apply the appropriate academic threshold
Determine the prerequisite and desirable subjects and the grades required. Does an applicant really need three A grades at A level to succeed at medical school?

Use written psychometric tests
Use purpose designed tests to identify essential and desirable cognitive and non-cognitive skillsfor example, logical reasoning, problem solving, and critical reasoning. Some tests are commercially available, some may need to be constructed. Tests could be administered in secondary schools under the supervision of teachers, or they could be administered during a visit to the medical school. Psychometric test data could be supplemented by data from the school principal's report on qualities such as common sense and personal reliability.

Invite applicants with the appropriate profile for interview
Applicants with the appropriate profile of academic and other attributes should be interviewed. Interviews should be conducted by trained staff following a predetermined and constant procedure to assess additional personal qualities and attitudesfor example, verbal communication and interaction skills and motivation to be a doctor. In addition, the interview could be used to investigate the potential compatibility of the candidate with the curriculum at a given medical school. For example, how well does the candidate work in small groups? Does he or she prefer didactic teaching to self directed learning? Evidence should be sought from candidates to support their statements.

The interview should be used only to gather information in the predetermined areas: it should not be used as the selection event. If the interview is the selection event the interviewers would be diverted from their objective of gathering information by considerations such as whether they would like to see the candidate at their medical school.

Select on the basis of a combined score
The selection event should be an administrative task. Decisions whether to select are based on a compilation of the collected data in proportions determined by the admissions committee. The committee would rank the qualities in terms of their importance to student progress or ultimate professional excellence and then construct a selection algorithm. For example, it might decide that the personal qualities assessed at interview are the most important followed by those determined by the written psychometric tests. Furthermore, it might have found that above a certain minimum standard the academic achievements of applicants have no positive predictive value for performance as a doctor.

Given that the interview scores are discrete (discontinuous) scoresfor example, an outstanding candidate scores 1 and a poor candidate 5and the written test scores are on a continuum, candidates can be ranked according to their scores and offered a place according to their rank on the list. For example, all candidates whose academic achievements exceed the threshold and who scored 1 at interview are ranked on their test score from top score down to a predetermined minimum acceptable mark. Ranking continues with candidates who scored 2 at interview.

	Conclusion

The fundamental principle is that students are selected according to a profile of qualities considered by the admissions committee to be important in a medical student and in a future doctor. Information relating to achievements and demographic variables that are not explicitly stated in the profile are ignored in the selection decision.

Further reading
Powis DA. Selecting medical students. Med Educ 1994;28:443-69.

	References

1. McManus IC. Factors affecting likelihood of applicants being offered a place in medical schools in the United Kingdom in 1996 and 1997: retrospective study. BMJ 1998; 317: 0000-0000.