It's not really about the money, though. The number of qualified, published academics in CS is just not high enough to meet new academic demands. I'm currently a student at an Ivy League school with an extremely reputable CS department, and we just hired 3 new professors who were considered big wins for the department (we have also seen a quadrupling of CS majors in the last 4 years). The thing is, we sent out something like 12 offers. Schools across the country are trying to scale their CS departments as elastically as they can (Yale, for instance, just launched an initiative), but the number of qualified professors is relatively inelastic. Of course, not all teaching jobs require a PhD (intro classes and applications classes, for example), but there are so many other more theoretical classes that do require an experienced professor.
agreed. it has been over 15 years since I graduated from a pretty good CS program in an engineering school, and we definitely had math professors teaching many core courses that were math focused (e.g. http://www.cs.cornell.edu/courses/cs2800/2015sp/). and professors from other engineering disciplines taught courses across (e.g. OR/IE profs doing stats).
this might have been possible at my school and not possible elsewhere based on how the different departments are focused / rewarded / isolated. more math profs might have been interested in getting CS experience back then.
More persons w/ only Masters but not PhDs should be used to teach the intro classes.
Although I was in EE at Stanford, I took a bunch of CS classes, and hands down Nick Parlante (an instructor, not a professor) was the best actual teacher that I had. CS108 and CS193i
Plus it is possibly that these persons actually make better teachers as opposed to quite a few professors, who are more preoccupied with research, tenure, grad students, etc.
I agree, this would be an effective and efficient solution. Also I think introductory, and eventually upper year courses could be taught more interactively and autonomously. A human teacher isn't actually required for most material. I think a new paradigm in teaching methods for CS will naturally emerge. I'm thinking of an interactive learning tutorial and practice guide that's comprehensive enough to out perform the best human teachers.
I hear you. There is a distinctly different skill set between "Pushing the boundary of human knowledge" and "Teaching an intro class well." Universities and tenure have historically been built on this being a high correlation. Reality is that sometimes it happens, sometimes not. Universities need to decouple the teaching from the research. Some of that is leveraging the best pure teachers available.
At my grad school, the joke was that Nobels were more important for admissions than academics.
I hear at Berkeley some of the intro classes can reach 1,200. More students registered then there are seats. Students are forced to skip class and watch recorded lectures online.
If you need one-on-one instruction in an intro to CS class, you may want to consider a different major, or take a semester off and learn to code before enrolling in a CS program.
“We’re all teaching three times as many students as we were six years ago, and we don’t have any more of us and any more money.”
This seems very strange. Are universities still so far in the dark ages of management accounting that they don't give departments more money if they have more students taking their classes? I understand it won't be purely proportional (Philosophy departments may need to have small classes and require subsidies) but if a department grows 3X, shouldn't the resources at least grow somewhat? Say 2X or 2.5X? If not, it's a recipe for disaster.
13 comments
[ 3.0 ms ] story [ 38.9 ms ] threadThat's pathetic. It's not like Stanford doesn't have the money.
Also what's wrong with stats teachers teaching machine learning?
CS hasn't been an S for that long, so I think you can rely on these other fields somewhat.
this might have been possible at my school and not possible elsewhere based on how the different departments are focused / rewarded / isolated. more math profs might have been interested in getting CS experience back then.
Although I was in EE at Stanford, I took a bunch of CS classes, and hands down Nick Parlante (an instructor, not a professor) was the best actual teacher that I had. CS108 and CS193i
Plus it is possibly that these persons actually make better teachers as opposed to quite a few professors, who are more preoccupied with research, tenure, grad students, etc.
At my grad school, the joke was that Nobels were more important for admissions than academics.
Part 1 -> http://www.stanforddaily.com/2015/04/15/a-look-at-stanford-c...
This seems very strange. Are universities still so far in the dark ages of management accounting that they don't give departments more money if they have more students taking their classes? I understand it won't be purely proportional (Philosophy departments may need to have small classes and require subsidies) but if a department grows 3X, shouldn't the resources at least grow somewhat? Say 2X or 2.5X? If not, it's a recipe for disaster.