Invariably, each PhD program requires the completion of a number of PhD courses and invariably at the end the students complain about the waste of time they supposed. That was also my feeling when I did mine (quite a few years ago already I’m afraid).
I believe it’s time to rethink this requirement. Maybe generic courses on writing & reviewing papers, doing good presentations, technical English,… make sense (though the last two should be already part of any undergrad programme) but that’s it. All the other advanced courses on X are useless. If your PhD is on X you’ll learn that anyway (and probably you already know it). If your PhD is not on Y, then at this moment of your career you couldn’t care less about Y.
Let’s save the time of our PhD Students (and corresponding lecturers) and money (both from the government and the students who share the cost of the courses). We will all be happier.
I actually learned a lot from my PhD courses, especially the seminar-style ones. I did not dislike taking them—but I did dislike the facts that (a) they were mandatory, and (b) there was a breadth requirement. Those two things often spoiled the fun.
I enjoyed mine, but only had a couple to take at UofT. More importantly, I learned about different teaching styles (ranging from small seminar to large lecture), formed a peer group, learned about different styles of presentation for assignments/essays, learned about some things I would never look into myself, and got to complain about professors a lot :-). Also, I lost any fear I might have had about learning new things, and I suspect my ability to improvise (on the fly) improved as a result.
I’m of two minds about courses. The UK system doesn’t have them, and PhD degrees normally take 3.5 years in total here; compare that with the Canadian system and the 5.5 years. However, I can’t help but think that the average UK PhD student comes out with less breadth than the average Canadian PhD student as a result.
I also think that, given the push from funding bodies for researchers to work in an interdisciplinary way, we need CS researchers who are able to learn new things quickly, and who have broad understanding of the field. I suspect (but can’t prove) that PhD students who’ve taken some courses are more likely to meet these requirements.