Examiners tend to grade relative to work already seen, research finds

That sobering conclusion by University of British Columbia professor Kevin Eva resulted from a controlled experiment that examined whether physician-educators' grading of younger doctors could be influenced by the quality of those they had previously evaluated.

The study was one of two articles by Eva, to be published Dec. 5 in JAMA (The Journal of the American Medical Association).

Conducted with researchers at the University of Manchester, it asked physician educators in England and Wales to grade videos of young doctors in their first year of post-graduate training. The doctors portrayed in the videos followed scripts to represent three levels of performances – good, poor, and borderline – as they examined patients.

Some physician-educators were primed by first viewing the good performances while others viewed the poor performances. Then both groups were compared on their grading of the borderline performances.

Educators who had been primed by poor performances consistently gave better grades than those who had just watched the good performances – depending on the patient case, the grades were 30 to 100 per cent better.

"This experiment shows that judging someone's performance – whether it's clinical skills, essay-writing, or figure skating – is likely to be relative," says Eva, Director of Educational Research and Scholarship in the Department of Medicine and a Senior Scientist at UBC's Centre for Health Education Scholarship.

"While such assessments are unavoidable in determining whether a student has mastered the required competencies in a field like medicine, we need to take steps to minimize contrast bias," Eva says. "It is important to ensure that there are a sufficient number of evaluations – the more data points, the more reliable the aggregate score will be."

In the second study published in the same edition of JAMA, Eva and collaborators examined the validity of multiple mini-interviews (MMIs) – a dozen, brief encounters in which applicants are asked to solve a problem through discussion with an interviewer, or through interaction with an actor or another applicant. MMIs have come to replace longer, more traditional interviews at most Canadian medical schools (including UBC's Faculty of Medicine).

The researchers found that students accepted to McMaster University, the first Canadian medical school to adopt MMIs, performed better years later on licensing examinations than students whom McMaster rejected but who were accepted at other, non-MMI schools.

Eva is an expert in the education of health professionals but he has also demonstrated how seemingly unrelated factors can affect judgment in a very different domain – men's physical attractiveness.

In 1999, he and a fellow graduate student at McMaster University asked women to view photographs and brief descriptions of men, and to rate their attractiveness on a 7-point scale. Men labeled as married had a mean attractiveness score that was 23 per cent higher than when the same men were labeled as single. As he and his collaborator wrote in CMAJ (Canadian Medical Association Journal): "Our findings do support the notion that being 'taken' influences perceptions of 'goodness."

Journal reference: Journal of the American Medical Association Canadian Medical Association Journal

Provided by University of British Columbia