Friday, November 14, 2008

Proposal Process

I've been working (in a very very preliminary way that mostly involves writing a paper for SoCG) on my proposal. Also (full disclosure) I'm one of the grad student representatives to my department's education committee, which has been discussing changing our process. So I've been thinking a lot about the process, how it works, whether it's good, whether it does what it means to, etc. Which leads me to ask - what's your proposal process like? Does it get students through the system in a reasonable amount of time (reasonable is not a code for short, I actually mean reasonable)? Do students enter the writing process/ leave the school knowing "what they should?"

I'm rather frustrated with the process at Maryland. (The qualifying exams process, on the other hand, I liked.) It consists of two sections which take place at the same time - the reading list and the research talk (with accompanying write-up). The reading list is a list of 30 papers in your broad area (for me, that means Theory) which are divided into three subareas of ten papers each (for me, this might include 10 papers "outside" of Theory in application areas). The proposal itself consists of about an hour long talk followed by a Q&A section. The Q&A section is divided into two parts - the part with the open audience and the closed section. The open section is like the end of usual research talks, and the closed section is (as far as I know having not had one myself) an oral exam with your committee members. During the oral exam, committee members can choose to talk about the research talk you just gave or quiz you on the reading list. The research talk is generally considered to be the main portion of the proposal, and most people at Maryland wait until they're about a year away from graduating before proposing.

I'm fine with these two components, but I don't think that having them at the same time works well. I actually think it sounds great to spend a lot of time thinking deeply about papers in Theory while not trying to immediately use them in research - just appreciating the papers as they are. But trying to do this while also preparing a research talk (and working on creating the results for that talk) means that instead of spending the effort on the papers that they deserve, most people rush through them or pick papers they've already read in their very narrow research area. Which also means that it's "better" to have committee members who won't ask about the papers. In a passing conversation with Bill recently about whether he'd be on my committee he made the point of mentioning that he does indeed ask questions about the papers, and that often students rule him out for this reason. Which I took as yet another symptom that the process doesn't work the way it should. Yet it doesn't make sense to hold myself to a different standard either. So what's the ethically and professionally best course of action that still allows a quick proposal? And yes, Bill, I'd be happy to have your opinion too.


Anonymous said...

I am sure Professor Mount told you that the unwritten rule is that 1) You need a paper accepted somewhere before you can propose and 2) unless you are completely clueless about the papers on your reading list, you shall pass. So, just keep working on the SoCG paper, and after that is accepted, form your reading list.

GASARCH said...

Anon 1 is assuming that there is a uniform standard. This is not true.
There is some flexibility,
which is both good and bad.
On the GOOD side: not all profs insist you have a PAPER. Some (including me) insist that you have A RESULT. I am not going to penalize a student because some conf committee didn't choose their paper, or some journal is dragging their feet.

Also- the reading list varies some, but I have been on committees where
the reading list is quizzed on quite a bit and even some where the student is told NO that was
NOT sufficient, in which case we may REDO that part of it at a later point. Ironically that is sort-of what Sorelle wants--- separate exams.
And on that I agree with her.

Bender said...

As the third UMD person to comment on this post, let me say that I also think the exams should be separate, but it's not obvious how to do it. It makes sense to me that the reading exam should come first, perhaps up to a year or two before (and perhaps serve as a way to encourage "weaker" students to reconsider their degree, if the department would desire such a thing). But it needs to be tailored to the proposal topic, which may not be finalized that early on. So maybe spreading the two exams over 3-6 months would make more sense.

And as for "spending a lot of time thinking deeply about papers in [my area]", many would say grad students should be doing that all the time. :-P

sorelle said...

bender - I guess I see the reading list as ideally more broad than just my narrow area (which is usually what I spend my time reading papers in) and so it seems fine to me to do it first and not have it connected to the proposal topic other than tangentially. Presumably you'll be reading the more specific papers anyway...

Bender said...

Reading lists that I have seen have had both breadth (seminal papers in the area) and depth (several papers relating to the proposed topic). With 30 papers, they can afford to have both. ;)

I fear that a system based on an exam that is broad and general could turn into something like quals. It could also let a student appear to make progress without narrowing their focus, if you were concerned about such a thing.

sorelle said...

Hmmm. I guess I understand the worry about heading towards something like quals, but I don't think that means that the reading list and research talk can't be severed, just that there probably need to be some other guidelines to accompany the new version of the reading list oral exam. Any suggestions?