Monday, May 12, 2014

How should qualifying exams and courses work

In my last post about what Every Theory Grad Student Should know some more general questions were raised:

Qualifying exams vs Course Requirements.
Why do we have either?

1)  To correct mistakes that admissions made. That is, some students that were admitted are not going to finish and better to get them out of the system sooner rather than later. 

2) To make sure students are well rounded.

When do exams work:
In Math or Physics there is a well defined notion of basic stuff that all grad students should know. That's why, for the most part, every Math prof can teach every ugrad math course. That a Logician Teaches Linear Algebra does not surprise anyone.

CON for CS exam:
In CS... we have not settled on an `every CS grad student must know...' I suspect we haven't because the field changes fast AND  our field is just so diverse. Hence  NOT  every CS prof could PASS every ugrad CS course. In short- having a qualifying exam may be problematic since we don't quite know what  the basics are

CON for  exams:
You are getting rid of people. Are they the right people?

Would you believe that someone we  kicked out of the program had four papers in STOC! Would you believe it! Four papers in STOC! No. Oh. Would you believe 2 papers in ICALP? No. How about a failed proof that P=NP?

 Where do you draw the line for `she's so good we should let her get a PhD even though they failed the qualifying exams' This doesn't happen in math so much since they don't start research their first year. Again, the difference is the age of the field and the prereq needed to get up to speed.

PRO for exams:
You get rid of people early before they waste too much time.

PRO for courses: If you take Blah courses in blah blah areas then even if you try to game the system some, you do KNOW something.

CAVEAT about courses: If the course grades actually matter for the requirements then the professor has to give out real HW and real exams. This forces people to learn the material-- even grad students need some encouragement. But  it may be bad for the learning environment.



  1. Can you please clarify what you mean by getting rid of people ? And no, it seems to go against the very principle of promoting talent when you kick someone out that has laid down a foundation and/or progressed the field through his STOC contributions. How many people did we come across that have passed the qualifying exams, did extremely well in exams yet at the end of the day produced little to nothing in terms of originality in their career ? Or marginally interesting papers ? Yet we kicked out talent that we should have promoted.

  2. By getting rid of people I mean that the INTENT is to get people out of the program who would otherwise spend a lot of time in the program and NOT get a PhD, so it saves both them and you time. Why did they get in?- admissions committees aren't perfect.

    If someone DID have four STOC papers then I may well argue that even though he failed all of his qualifying exams, even the theory one, we should find a way to waive the requirements.

    If someone DID have two ICALP papers that his advisor wrote most of then less likely to argue the case.

    If someone DID have a failed proof that P=NP then would not argue the case.

    But this points to a problem with an exam system--- you end up making exceptions,
    some of them quite reasonable, but where does it end?

  3. I agree with you on the 2nd point wholeheartedly perhaps even 3rd point ICALP papers (dominantly written by supervisor) perhaps even then, u could ease the requirements ?

    If someone did have a failed proof that P=NP then how would we find out ? (It usually takes up too much time to track down what is going on ...).

    At the end of the day grad school is not for everyone. (So, there's a sense in making exceptions ... )

  4. You ask whether the student with 4 STOC papers should have been allowed to continue. What about the student who really wants to pursue research in TCS, but only has 1 paper to show, and can't, or simply doesn't want to, learn these arcane topics they have little interest in? Isn't it towards the end of the PhD stage that students are expected to pick up the speed at which they publish after all?

    It is in everyone's better interest to root out students without the right intent and aptitude. But, can these exams really measure their ability to produce a good thesis? Do they test students for these skills? Standard exams tend to favour the hard working students over the ones capable of original thought, whereas its the latter that can truly advance research. Open book, and more generally, take home exams are closer to research, but not widely administered.

    Exams are also very impersonal, objective and skewed, in strict contrast to PhD research. For instance, in a TCS problem, where you simply need to reduce the given problem to some known one, the reduction itself may be non-obvious; you get the intuition, you get full marks, and naught otherwise. In research, this intuition needn't be obtained in a restrictive time frame and through their personal interactions, the professor would know not to write the student off for the bad grade.


    PhD students are typically expected to pass their qualifying exams within 12-18 months from their joining the program. Why not skip the coursework, work with their supervisor, try and build up a rapport, and finally decide together if they want to continue (assuming, of course, that the student is already familiar with the area and has done previous coursework)? This way, you evaluate their ability to research by doing research, rather than through cousework, while rooting out the undeserving ones who got through admissions; even avoid the scenario where a student passes the qualifying exam and then can't work with the supervisor. And the best part? The truly deserving students are much less likely to get turned away from the discipline.

    As far as making sure the students are "well rounded" goes, I think PhD students have the maturity to understand the benefits of widening their knowledge base to open up newer avenues for research. Typically grad courses are designed to consolidate the base for those pursuing research in the subject area; this level of depth is unnecessary and taxing for the students trying to get a general idea about the subject, turning them away. Auditing such courses, perhaps even with compulsory attendance, without having to prepare for the exams, is an ideal compromise indeed!

    I have no illusions that these proposals have not been made before and that they are going to be make any change this time. These matters would require consensus between the different disciplines, and repressive administrations are all too common in Academia. But I did have the fortune of learning at institution where the administration did not interfere in matters it didn't absolutely have to, and it was good. In the end, one can only hope the system will become less hostile for future generations.