Why are some departments changing the way they do comps but others refuse to do so? How do you evaluate students objectively if you have them? Did some of you ever considered implementing a portfolio system in your department?
(probably feeding the trolls but I might get some useful insights too. What can I say, I'm an optimist).
Hey Greffey, what's a portfolio system?
You know more about Political Science literature when you walk out of comps than at any other time in your career.
^Sure, keep telling yourself that, you third year grad student who just passed comps!
^ He just stated the truth. His career will end two years from now when he goes on the market. Maybe sooner.
Is there a difference?
^ You mean outcomes will be as bad? The written stuff we get is terrible and orals worst.
No way you know more right after the comps than later after your become a professor. That is, if you are productive and publish, teach grad seminars, etc.
Maybe I thought I knew more back then, but looking back it's pretty clear I know more now about the literature than back then.
^Agreed.
Comps are probably a waste of time.
Comps are useless. Professors who like them are ego-driven blowhards, everyone else knows they're **** but is too lazy to actually do something.
^ Is there any evidence for that, or is that just your impression?
I found comps useful, insofar as I read a bunch of stuff I wouldn't have otherwise read, and I've used some of it occasionally in teaching and research.
Thanks for inspiring a blog post: http://saideman.blogspot.ca/2012/04/purposes-of-comprehensive-exams.html My take, in a nutshell, is, yes, they are for more than just hazing.
I think I was more broadly read after comps than at any time I've been in my career afterwards. I'm not saying I was a more polished scholar right after comps, but as the research agenda has narrowed over the years, I've, for the most part, left behind my 2nd and 3rd exam fields and only kept up with them enough to stay abreast of particular methods for publication purposes. Given the rigors of teaching, service, etc I do not have the time or inclination to stay caught up on everything that's happening in each of my examination fields as I did when I knew I was going to be tested over the material.
Comps are supposed to prove that you can teach a graduate seminar on the topic. Nothing more, nothing less.
I agree with Steve's post. I would add that from the faculty's perspective one of the things we want from comprehensive exams is the ability to certify that students are prepared to teach a broad introductory course in the examination field.
Duke has made an interesting change in this process that I think I like. We now have written qualifying exams at the end of the spring semester of the second year. If the students pass these exams then they spend the summer working on a prelim paper instead of reading for comps. The paper is supposed to be a research effort of potentially publishable quality that the students must defend during the fall semester of their third year. Then students must defend a dissertation prospectus during the spring semester of their third year. The prelim paper can be used as the foundation for the prospectus and/or can be revised and sent out for publication.
I like this system because it means that students do not spend (waste?) the summer after their second year reading for comps and instead they have a paper they can send to conferences and/or journals by the middle of their third year.
^seems like a more productive system yet prelims can turn into pseudo comps as we confuse the purpose of both and just write bad questions. Here at least...
Qualifying exams at the end of the second year is a good idea, as long as the coursework is designed to prepare students for the exams. At some places the exam curriculum and the courses offered are not systematically linked, meaning students need extra time to brush up on exam material over summer.
You know the field much more broadly right after comps than you do later, especially in those second and third fields. Once you start teaching and researching, you will know a lot about those areas, but you'll lose track of what you aren't using. Passing comps at least provides the department some assurance that you'll know where to start when prepping your first few classes and won't just teach your research.
Comps are both learning experience and hazing. Yes, people are probably more broadly read as a result of comps (and that is a good thing). But, the reason we don't see more reforms towards evaluative tasks more on point to being a political scientist is hazing.
The oral component of comps is pure hazing. Totally demoralizing yet useless experience.
I actually don't agree that the oral defense is hazing. As academics we are often judged by our presentations to groups and especially on our ability to respond during Q&A. This is true at conferences and especially job talks. So students are well served to practice. This is especially true for the defense of the prelim paper in the new prelim system at Duke.
^ This time I get to agree with Chris. While the oral comp feels like high stakes, they are good practice for the job talk which are truly high stakes.
I never did oral comps, and I would have hated it, but I agree that it's useful. With written comps or research you can talk around gaps in your knowledge. While this is a useful skill, there will be times in an interview, at a conference, at a talk, or even just among colleagues when someone will ask you how your work relates to Waltz's (or Huntington's, or Aristotle's) well know claim that (whatever). I'd rather my students practice looking stupid with me now than have them doing it later, in public.
My take on comps is that they are useful for somethings, but far from perfect. It's true that you never have a *broader* knowledge of political science than the day you take comps, and for that's very important. It's especially important for those who will go on to teaching-oriented careers, but everyone benefits from a broad grounding in multiple literatures - it improves your own research as well as your ability to understand others' research. It also means you can engage more people in more conversations about more things, and this last thing is especially important on the job market - never underestimate the "can I have an interesting lunch with this guy/gal? factor. Furthermore, there is some diagnostic value to seeing which students are able to do well and who simply can't hack it - it saves everyone several years of painful toil on dissertations which everyone (except maybe the student) knows aren't going anywhere.
On the other hand, there's little-to-no correspondence between taking comps and the process of doing research, and so they're bad at teaching students how to do that part of the job. Along the same lines, comps do little to encourage depth of knowledge in a micro-targeted literature, even "focus field" comps (or whatever they're called at your school). So I understand why some schools are moving to models which place more (or even all) emphasis on field papers.
- Associate prof.
Top 6 here. The purpose of the comprehensive exam is to provide faculty with a non-confrontational and "objective" mechanism to get the worst students out of the program. No more, no less.
It seems to me if the goal of comps is to weed out weak students then they are a failed institution. I have been doing this for over 15 years now, and I would say that about 5 IR students have failed exams during that time at Harvard and Duke. So that is a pretty weak screening mechanism. Yes, we are trying to select good students for admissions, so you would expect few failures 2 or 3 years later, but the truth is that grad admissions is a low information business.
Comps are much better at giving students some fairly honest feedback about their strengths and weaknesses than they are at actually forcing people out of the program.
BTW - top 6? Really? Are you sure it is not top 7 or top 5? What is your confidence interval on that estimate? ;-)
My experience at different top-ten schools than Gelpi's is similar--maybe one student gets weeded out every 3 or 4 years, but that person was probably about to drop out anyway. But I think more should be weeded. If we believe that we are providing students incentives to do something important by preparing, then a little more motivation wouldn't hurt. And if being in the bottom 10% is a strong negative signal of their potential, then we should encourage them not to waste another 3-7 years of their time (and ours) trying to get a PhD rather than finding a career that suits them better.
Oral exams have very little in common with job talks. Your committee comes into the oral with prepared questions designed to tear you a new one and to leave you feeling intellectually inadequate. In fact, the better your written answers, the more grueling your oral.
A job talk, by contrast, involves you talking about the thing you know better than anyone else: your research. You spend a long time preparing your talk. And you do several practice runs, with practice questions, to ready yourself against tough and, more important, out-in-left-field questions.
Crucially, you have more prep time than the questioners do for a job talk. For oral comps, the opposite is the case. And for me, that makes them very different experiences.
I agree with ^ this post. That is part of the reason we shifted the oral defenses at Duke to be defenses of a prelim paper and kept the written portion of the comprehensive (lit based) exam.
I also agree with ^^ this post. We do need to find more systematic ways to prevent students from sinking years and years into a degree that is very unlikely to lead them to professional success. A bit more courage by those of us administering comps might not be a bad way to do that.