When it comes to medicine, training bright people simply isn't enough. Their smarts alone don't predict how appropriately they'll prescribe drugs, or how likely their patients who suffer cardiac arrest are to live.
Nine years ago, this theory, coupled with a fear that an overemphasis on marks was driving away male medical students, led professor Harold Reiter to re-imagine the way McMaster University screens the personal traits of medical school applicants. He crafted a new interview style that has become the Canadian standard and appears to have narrowed the program's gender gap.
Now, the same ideas have been used to build an innovative online test that Dr. Reiter hopes will net more students with the character to match their cognitive powers.
In late October, the first 3,548 applicants to McMaster's program took the Computer-based Assessment for Sampling Personal Characteristics, or CASPer. With its advent, the school is putting less emphasis on students' grades than ever before in an effort to groom better, more balanced doctors.
"We do a very good job at making sure the people going to medical school are smart," said Dr. Reiter, McMaster's chair of admissions for medicine. "What we've done very poorly is on the side of personal characteristics."
He fears failure to discern good decision-making, ethics, communication skills and cultural sensitivity will mean universities accept people who are not necessarily going to be the best doctors. A study in the New England Journal of Medicine of complaints to U.S. state medical boards found that "92 per cent of the identifiable complaints were because of personal characteristics - professional qualities, if you will," Dr. Reiter said.
So in 2001, he and his colleagues began designing the Multiple Mini Interview (MMI) to put applicants in challenging, live scenarios at up to a dozen stations, with as many interviewers. The MMI debuted in 2004, and 12 of Canada's 17 medical schools had adopted it by 2008.
Its success inspired Dr. Reiter and fellow professor Kelly Dore to build CASPer as a way to bring identification of those strengths into evaluation of the total number of applicants each year rather than just the 500 selected for interviews. Most sections of the two-hour test present video scenarios, giving students five minutes to answer three questions. Situations might include having to fire an underperforming colleague, or to advise someone who is disputing a doctor's advice, and there are no right answers.
Before CASPer, McMaster applicants answered five questions with personal essays. But nothing stopped them from enlisting outside help, and the essays didn't really predict who would interview well, Dr. Reiter said.
Early results suggest CASPer works better. Second-year medical student Jelena Lukovic, who recently took a pilot CASPer test, said she found it harder and more revealing than the essays she spent a month crafting for her actual application.
"It's more about what's the first thing that jumps into your mind when you see this scenario. I think that gives you a better insight into the individual," she said.
Until now, grade point average counted for about a third of each application, as did the personal essay. But McMaster has such faith in CASPer that its scores are now weighted nearly twice as heavily as grades.
In 2002, nearly 77 per cent of students admitted to McMaster's MD program were women, but by 2009, the school had restored some balance, with 61 per cent women. Dr. Reiter credits the MMI, which has proven its neutrality to an applicant's gender and income level. Given CASPer's many similarities, it is also expected to be unbiased.
"Before CASPer, there has been no test that one could apply in an accessible and equitable fashion to very large numbers of people in a wide geographic dispersion," Dr. Reiter said.
Not everyone is rushing to follow suit, however. The University of Ottawa's application collects only grades and a detailed resumé. Interviews are divided into eight sections, but all in one room, and assistant dean of admissions Gary Hollingworth sees advantages in keeping the same interviewers throughout. He finds CASPer "a very interesting approach," but thinks "there's still a lot of value to the way we do it."
McGill University adopted the MMI in 2008, and "we love it," said Saleem Razack, assistant dean of admissions for medicine. Yet McGill continues to weigh grades heavily and asks for a personal narrative, which it feels is revealing.
Dr. Razack concedes personal narratives can involve "lots of issues," including students hiring professional help to write them. He is keeping a close eye on CASPer, and won't rule out adopting it if it can be made bilingual.
"We wouldn't want the cookie-cutter thing to happen," Dr. Razack said. "I don't think we should be attached to [any method]"
CASPer can be taken from virtually any computer with an Internet connection, which raises authenticity concerns. But McMaster University may have found a solution.
Because each person's neural synapses work uniquely, there are measurable time delays when a given person types combinations of two letters, such as "nt," "in," or "qu." Research from Torino, Italy, suggests recording a small segment of someone's typing is enough to profile their "typing signature," and the chances of forging that signature are one in 10,000.
In theory, a sample of enough of these keystrokes could tell admissions officer whether the person applying is the person writing the test.
When interviews are held in the spring, applicants will be asked to write another CASPer question so McMaster can collect their signatures. But nobody will be busted this time - Dr. Reiter's team is still verifying how exact the science can be.