For lay readers, comprehensive exams are sometimes known as candidacy exams and are what most North American PhD programs require their students to pass before being allowed to call themselves PhD candidates instead of students. The remaining stage of the PhD after this is generally the completion and successful oral defence of the dissertation. There is no universal standard of comp exam, and departments and universities set their own criteria. In some cases, the student has an option to pick one among several forms of examination. In others, the department has a single set type. Some universities, such as those in the UK or Australia, often don't have any sort of comp exam. North American PhD programs also tend to require the completion of graduate course work before comps are attempted and research undertaken. A PhD in my department requires the student complete 10 courses (think 2 to 3 years of full time study). An analogous department across campus where students can do the same research requires three, often shared with their cousins in my neck of the woods. The standard required for comps differs too. My department keeps their Phd students in house completing course work (a whole other post is need to deal with the issue of masters labour like me) for 3.3 times longer than the TWO comparable departments on campus where a PhD student could do the exact same research.
At the end of the day the PhD is awarded largely based on the successful completion and defence of an original (as much as that's possible now) piece of research, the dissertation, not a gatekeeping exam no one will ever ask you about and which most PhDs I know have, like most students and exams, long forgotten. However, that does not stop people like Dr. Lukacs and others from making grand arguments about the reputation of schools and rigour demanded by comp exams and the apparent "worthlessness" of a PhD from UofM or whatever school has a bit of a hiccup. I mean, the mathematician in question DID do very well with their disseration and had an above average publication record coming out of his PhD: How much by those realworld standards was the failed comp actually worth? Yes, I can seriously see a hiring committee (and I've been part of one) turning him down because "well, despite your outstanding publication record for a new PhD and your excellent references, you failed one of your comps 6 years ago so we're not going to hire you." Sure. Although UofT might, but that's another story. Is there an actual pedagogical value to comprehensive exam? If so, is it universal across disciplines? Hmmm...there's an education dissertation in there somewhere.
I've lost track of how many people have told me that I'd never get hired with an Aussie or British PhD because they don't require coursework or comps or whatever like Canadian or US doctoral study , nevermind the really outstanding research that gets produced in those apparently inferior institutions and the scholars in my field under which I would be very fortunate to study. Or what my CV looks like, you know the things that matter, like refereed publications, after a PhD. Oxford or Cambridge are apparently acceptable. Maybe LSE. Seriously, I hear this line frequently.
2. The University of Alberta. A math professor is asked to resign after he asked his students to protest the departmental lowering of class average GPA from 2.16 to a positively evil 1.79 on a 4.0 scale, and a 29% failure rate in a first year course.
Back in May, Kovalyov received an email from an associate chair in the Department of Mathematical and Statistical Sciences informing him that grades for his first year math course had been lowered, resulting in a change in class average from 2.16 to 1.79 on a 4.0 scale. Other sections of the same course had averages that ranged from 2.13 to 2.95, according to documentation obtained by Maclean’s. The math professor says that he had already failed over 20 per cent of the class before these changes were approved. University guidelines suggest an approximate mean average of 2.62 for first-year courses, with only six per cent of the class failing. When faculty services officer David McNeilly, who is also responsible for reviewing final grades, first proposed the changes to Kovalyov in April, he explained in an email that the department’s proposed grades for Kovalyov’s class were “more generous than the typical exam cutoffs.” He also pointed out that in Fall 2009, the department failed 29 per cent of students in one section of the class. “In particular, we are being consistent,” McNeilly wrote. Kovalyov responded to McNeilly that if so many extra students deserve to fail, then they should never have passed and received credit for math courses in the previous semester. “If we were consistent, all these students would have never made [it] through” the prerequisites, he wrote.
I hate bell curves. Profs will tell you that your grade may change between the final they send you and what appears on your transcript after the senate or whatever body is deemed responsible, reviews course marks; they simply can't have too high an average. The awarding of final marks has much less to do with the quality of work from the class, and much more to do with the idea of conforming grades to a bell curve. Yes, by and large, a bell curve will emerge in grading around a particular mid-range average, usually, in my experience in the 65%/C to 75%/B range (although 75%/B is usually considered too high). The final grade is thus an administrative decision, not an academic one, based on the idea that a class should conform to a curve and the curve conform to specific letter grades and somewhere down the logic chain that the work from the students rate would rate accordingly or something.
Rank and file profs do not want to explain to the dean why their grades are too high but, according to colleagues, it isn't usually a problem to have high failure rates in the courses so there is downward pressure on grading which has nothing to do with quality of instruction as it is reflected in student work, as the UofA case shows. Having worked as a TA, however, I have seen dramatic improvement in students as they address comments in their work and start to 'get' university. An outstanding instructor, one with a gift for teaching, can do wonders for improving the quality of work from lagging students and thus raise overall averages. God forbid that undergrads actually learn something from their instructors and consequently show improvement in their marks.
But you'll get some profs and deans wax on about the need to maintain rigour and standards and the bell curve is the master of all that, but in reality they're playing a numbers game, not a learning game. The final grade curve is adjusted by people who have nothing to do with the course or students in question. The students, out of pocket thousands of dollars a year usually via debt, come to play the latter. Some faculty will complain to no end about the seemingly perpetual declining quality of incoming students, yet when students show marked improvement, there is pressure to keep them 'in their place. C student first year, C student 3rd year: why would an instructor even bother to grade by 3rd year? Just look at the student's past transcripts and give them a suitable number or letter that fits within their GPA. After all, the university wants them to remain within their grade-class as too much larnin' upsets their perfectly mowed little grade-mounds and deans get the vapours.
It's a Wonderland here at times. Any thoughts on my little rant? I might have more soon.