Marking system at McGill doesn't make the grade

Montreal Gazette

Tuesday, February 22, 2005

You're about to board an airplane when the check-in clerk offers you a choice. The second flight bound for your destination has a two-hour layover. The first flight leaves immediately and is non-stop, but there is a catch - the instrument panel was designed by someone who scraped through a postgraduate course in artificial intelligence on a technicality.

I don't know about you, but I'll be waiting in the lounge.

Jeremy Cooperstock, a McGill University engineering professor, says safety issues and academic integrity have been playing on his mind a lot lately - ever since the dean decided to inflate the grades of three students in his AI course without his consent.

"I'm very disturbed," said Cooperstock, reached by phone in Europe, where he's on sabbatical from McGill's Centre for Intelligent Machines and department of electrical and computer engineering.

"Usually, I believe it's best to keep our dirty laundry to ourselves, but I really feel this brings disrespect on our institution and its academic integrity."

Cooperstock's battle with McGill began last spring, after two students in the bottom quarter of his 500-level AI class complained he had awarded them grades of C-plus - a pass for undergraduates, but a failing mark for students in master's and PhD programs. At McGill, 500-level courses are open to both undergraduate and graduate students.

Engineering dean John Gruzleski said yesterday he boosted the grades to a passing B because of an inconsistency in university guidelines for grading between the engineering department, where the professor sets his own number-to-letter scale, and graduate studies, where the marking system is defined in the course calendar.

Since Cooperstock never told the class precisely what his rating system was, Gruzleski said graduate students - the two who protested and third who had also received a failing mark - had good reason to assume they'd be governed by criteria used in graduate studies. When Cooperstock refused to budge, Gruzleski says, he intervened "out of fairness to the students," whose lives and careers are at stake.

"If he'd wanted to fail them, he should have given them an F."

He admits "hybrid" courses can be awkward to administer. An academic committee has been asked to shore up the gap between the two marking systems - a discrepancy dating back to the early 1970s but never challenged before.

Cooperstock contends McGill has sent the wrong message, inviting lazy students who know their grades won't be improved by the standard "reread" process to take advantage of an ambiguity to get a pass.

As a teacher, he's always reminding engineering students their decisions will affect human lives.

A pass is the university's "stamp of approval, our way of saying they've been verified and they're good to go. If I'm recruiting and I know someone has a C-plus, I may still hire them, but maybe that's not the person I'm going to want to put in the most safety-critical position."

In a letter posted on his Web site, Cooperstock said: "Academic integrity ... is being seriously eroded by our catering to students who complain the loudest. . . . I wonder what sort of legal liability the university would be exposed to should similar grading decisions be made in the faculty of medicine, allowing students to become practising doctors without demonstrating ample knowledge of medicine?"

Cooperstock's Web site is