The medical program at McMaster University is one of the best in Canada. While everyone likes to toot their own horn, they’re simply telling the truth when they say:

…a group of innovative educators developed an undergraduate medical program that stirred controversy and defied convention by emphasizing self-directed learning. McMaster created a revolution in health care training with the establishment of a medical school that pioneered a problem-based learning (PBL) curriculum, which has since influenced health care education worldwide.

So what makes it special?:

Extensive peer-reviewed research has proven the effectiveness of two of those assessment methods:

Development of clinical skills is assessed by Objective Structured Clinical Examinations (OSCEs). Students participate in one OSCE in each year of the program. Each 12-minute OSCE station presents the student with a unique clinical case which will test particular skills such as history-taking, physical examination, communication skills, test interpretation, medical decision-making and responses to ethical dilemmas. Each student receives feedback from the assessor as well as overall scores for the stations.

and:

Formative assessment is provided to students through the Personal Progress Index (PPI) which all students in the program complete 3 times each year. This is 180-question multiple choice online test which allows students and the program to assess progressive knowledge acquisition. The questions test a broad range of basic and clinical science and fundamental principles of medicine.

The payoff for all of this is that while most university MD programs are four years long, McMaster’s is only three, but produces doctors who are just as capable by every available measure as those who’ve spent a third longer in school. Another (which is just as important to me) is that Mac takes in students from a much wider range of backgrounds than most medical schools, again because research showed that doing so worked.

What started me on this exploration was a story. The story is that when the program at Mac was designed, its creators started by surveying doctors to find out what they still remembered and used five years after finishing medical school, and only included those things in the curriculum (the theory being that if nobody uses it, there’s no point teaching it). I haven’t been able to confirm that story, but I do know that there was a lot of protest from other medical faculties when Mac began innovating this way.

Trying to track down that “five years out” story led me to a question: what would an undergraduate software engineering program look like if it was designed and run this way? What would we teach if we looked at what practitioners do, use, and remember five or more years after leaving school and focused on teaching that, using problem-based learning with frequent formative assessment, one-to-one pairing with experienced practitioners, and everything else that McMaster and its imitators do?

I suspect such a program would have more emphasis on project management, design (both of software and of interfaces), tools, and systems than the programs we have now. I suspect there’d be less about esoteric data structures and analysis of algorithms. I also suspect that there would be an outcry among existing computer science faculty, much as other medical schools criticized Mac’s program for dumbing down the curriculum despite repeated demonstration that they were doing nothing of the kind. Finally, I think that outcry would lead to software engineering education finally being separated from computer science education. The former would focus on training practitioners who need to understand and apply research that’s relevant to their work, while the latter would focus on training researchers, just as medical schools are distinct from research departments of cell biology and human physiology.