Qualifying Exam Season

Gordon Watts reminds me that the start of a new academic year means more than just the arrivial of a new crop of freshmen. For grad students, it’s qualifying exam season.

For those not in the know, “qualifying exams” are a common feature of most Ph.D. programs. These are big, comprehensive tests that all students have to take at the end of the required course work. They usually come at either the end of the first year, or the start of the second year, and you have to pass the test in order to continue in the program.

And, of course, the tests aren’t exactly easy. As Gordon puts it:

The common saying is “You’ll never know more physics than you do for this test.” It is totally correct. Henry, Toby, and I tried to help two students working on a qual problem from several years ago. Toby and Henry are senior professors. I’m getting up there. None of us (except maybe Toby) could look at the problem and solve it right off.

Calling this a high-stress situation would be an understatement. It’s basically the physics equivalent of the Bar Exam, or the MCAT’s. The actual Ph.D. defense is a cakewalk in comparison– if they let you schedule a defense, they’re pretty much going to give you the degree. With qualifying exams, they actually fail a good number of people– at Maryland, they flunked half of each class as a matter of policy.

(Painful qualifier story below the fold…)

My own qualifying exams were a thoroughly miserable experience. It’s the one time in my life that I’ve really been hit by test panic– the sort of situation where you freeze up completely, and just can’t answer anything.

The qualifying exams at Maryland were split over two days– in my program, the first day was spent on thermodynamics, statistical mechanics, and “quantum chemistry,” which really meant “molecular physics.” The second day was all quantum, all the time. There were three questions each day.

The first two questions of the first day were pretty easy. The QChem question was a slight variation on something that had been on the final for the class, and the StatMech question was relatively easy. And then came the Thermo question. Which asked about something we hadn’t covered at all in the class.

While that’s technically fair game– we knew we were responsible for material outside the actual lectures– it’s more than a little disconcerting. Worse yet, it wasnt a question in any of the areas that we knew to expect different material. Over the summer,w e had spent hours going over old qualifying exams, and not once in the six or eight years’ worth of old problems had “fugacity” come up.

I was absolutely baffled. I couldn’t see how to do part a) of the question, and couldn;t even figure out how to write the pathetic little partial credit essay for part b) (“If I knew the answer for part a), I would take it and do the following…”). I had nothing, and left the room wobbly.

I went home, and spent the evening studying quantum mechanics in a very fragile state of mind. And came back the next day for the second part of the test.

When I opened the booklet, I had no idea how to do the first question, and right then, I was finished. My mind totally went blank, and I was basically unfit to do more than doodle in the margins from that point on. The second question was equally opaque, and while I knew where to begin with the third, I was so rattled that I flubbed that one, too.

I left the exam room absolutely shattered. And, of course, fifteen minutes later, I knew exactl what I should’ve done for the first two problems…

That was probably the low point of my graduate career. I remember going to see a movie that night, and making the unwise choice of Natural Born Killers. Not only is it not a good movie, but it’s not a good movie to see in the state that I was in– I basically sat there thinking “I could shave my head and go on a multi-state killing spree…”

Of course, it worked out in the end. Because I failed the qwritten qualifier, I had to take an oral exam to make up for it, which wound up being scheduled for two days before Thanksgiving. I crushed the orals, to the point where one of the professors on the examining committee asked “How did you get this question wrong on the written test, anyway?”

I’m really, really glad to know that, whatever else I may do with my life, I’ll never have to take Ph.D. qualifying exams again. Best wishes to all those who are facing their own quals in the next few weeks.

Comments

  1. #1 Rob Knop
    August 24, 2006

    At Caltech, we had a couple of 6-problem written quals — one classical, one quantum. We were given three tries to pass them, so most people don’t really study much the first time around. I studied some, but not a lot, for the first test. I squeaked by the classical, and didn’t squeak by the quantum; quantum got nailed the second time.

    My later-to-be-housemate, however, managed to fail them all three times. The first time was because he didn’t bother to study. The other two times was becuase he was doing so well in his research that he didn’t take enough time away to study. His advisor went to bat for him, they worked something out, and now he’s tenured at MIT. Even qual disasters don’t necessarily spell the end, it seems.

    What was most painful for me was the oral qualifier. I went into it nervous, and came out of it feeling like a wreck. I passed, but when I had left the room for the deliberations, I thougth that I had been a clear failure. I guess it wasn’t so bad as all of that, but it really was a painful experience. There were some questions I didn’t answer that I should have been able to answer. It was a mess, because I had my presentation ready– but a committee member asked a question early on that I would answer later, so I pulled my slides out of order to answer, and never recovered from that orgainzational mistake.

    Ah well. Water under the bridge.

    The defense, as you say, was much, much nicer than that. Not only because on who isn’t going to pass won’t be let in, but because at that point, I really was the one in the room who knew my topic a lot better than anybody else. I had already started my post-doc before I gave my defense, and I’d given a seminar on the topic at the new location– so I’d even had practice. I had gotten better at doing presentations by then, and I had been living and breathing the thesis material. I wasn’t really all that nervous, and the whole thing went quite well. (Of course, there was another tragic tale a few months before that, but I’ll leave that for another day….)

    -Rob

  2. #2 Kate Nepveu
    August 24, 2006

    “Fugacity”? Isn’t that a measure of how badly you have to be dressed before you appear on Go Fug Yourself?

    Oddly, my recollection of the bar–and Chad, you may have seen it differently–was that studying for it was two months of deepest darkest hell [*], but that the test itself was not that bad. It probably helped that I didn’t study at all the nights before (the bar is also two days), instead watching silly movies (and I read a silly book during lunch, one of the Dortmunder novels). I suspect most of it has to do with the sheer number of questions–I have completely blocked out how many essays we had to write, but it was several, and the Multistate day is something like 200 multiple choice questions. With that many, if you freeze up on one, it’s a lot easier to go on to another.

    [*] I spent more nights than I care to think of trying to calm down enough to sleep by staring across the room at the red LED clock, counting off seconds in my head and seeing how close I could get to an accurate minute. Except for the nights where it was so damn hot that I slept in the living room, which had a ceiling fan. I will never live without central air conditioning again.

  3. #3 Kate Nepveu
    August 24, 2006

    Oh yes, the other test-taking advice: never check your answers after you’ve handed in your test. Certainly not after the first day of a two-day test, but generally just don’t.

  4. #4 Chad Orzel
    August 24, 2006

    “Fugacity”? Isn’t that a measure of how badly you have to be dressed before you appear on Go Fug Yourself?

    Sadly, no.
    According to the textbook for my graduate thermo class:

    “…in the classical limit, the fugacity is the ratio of the ‘thermal volume’ (lambda_T)^3 to the volume per particle (of a single spin orientation) V/(N/g_0). The system is in the quantum regime if the thermal volume is larger than the actual volume per particle (of a single spin orientation) either by virtue of large N or by virtue of low T (and consequently of large lambda_T).”

    HTH. HAND.

  5. #5 kstrna
    August 24, 2006

    Our biochem/biophysics qualifying was set up differently. Not a two day affair nor was there a written exam. Two months before exam time, you submit your thesis proposal and a proposal for an additional research topic unrelated to your thesis. Before this, during the summer after your first year, you develop your thesis ideas and review your course work & the literature to develop the non-thesis proposal. The dept. selects a qualifying committee based on what you submit. You then write up both proposals like you were applying for a grant. You read, reread and review during this time, meeting with your committee members and others to figure out if your ideas are sensible or not. 6 weeks later you hand in those written proposals. Your committee accepts or rejects them. If revisions are asked for, they have to be made before your oral exam date. You spend then next 3 weeks reviewing your course work and rereading all the important papers. Then you have your oral exam where you are presenting your proposals and defending them as your qualifying committee is asking questions left and right testing your ability to think, design experiments and your understanding of basic knowledge in biochemistry & biophysics. Few outright fail as those who are looking unprepared (usually a student who were not biochem/mol bio/etc. majors in undergrad) are encouraged by their committee to move their exam to the spring term so they have more time to get up to speed. Usually conditional passes are given (ranging from retesting on one proposal to rewriting on section to having research article discussions with one of your committee members for a term).

    It is basically two months that are the worst of your graduate career. You don’t have time to do research as all available time is to be devoted to writing, reading and discussing. By the time the oral exam date rolls around you are ready for it all to be done one way or another. Since the questions can range over many topics and techniques, grad students study way more than they can be asked in the time given. Which is the point of the qualifying exam, to make sure we have a strong foundation which we can build on and use to move into any subfield down the line.

  6. #6 Jeff F.
    August 24, 2006

    Thanks for making me remember my quals, Chad! Back when I was a grad student at Yale, we had to take two sets of qualifier exams. The first, the “pre-lim” exam, was a 6 hour/6 written question exam given on one day after the end of your first semester. Questions would cover material from our recent Mechanics, E&M, and Quantum classes plus anything else at undergraduate level was fair game. The latter part sucked because it was like studying for the physics GREs again (i.e. relearning everything you learned as an undergrad). Worse off, this test was given during the second week of your second sememster. For me, this meant cramming in studying over our winter break because I sure had no free time during the previous semester. Barely passing it, I felt like crap and defeated by this exam.

    This whole ordeal was repeated again the following year, after our third semester and again with only a few weeks to study. This qualifier exam covered advanced topics taught in grad school, from particle to atomic to GR and everything else in between. The morning session was 3 mandatory questions to be answered over 3 hours, followed by 3 of 5 questions in the afternoon session. This time around, I did much better, but was mentally exhausted afterwards.

    But by passing these qual exams, you knew you were almost home free. Before you could defend your thesis you first had to take an oral exam on a selected paper. Thanks to my advisor’s view of the administration he got me around this silly requirement so I could concentrate on lab work. It really was silly because no one ever fails this exam, especially if you already have had experience at giving talks.

    Then came the thesis defense and as Chad and Rob mentioned, this really was a cakewalk (with cake and champaign afterwards too)!

    Needless to say, Yale’s exam requirements were a little too much when you factor in their class and teaching requirments. I really hope they’ve changed their system by now.

  7. #7 Julie Stahlhut
    August 24, 2006

    I really lucked out in more ways than one. I’ve always been a good exam-taker, and a very calm one. My written candidacy exam was based on six different core areas — evolution, population ecology, prokaryotic and eukaryotic cell biology, and animal and plant physiology.

    I’d taken all of these courses and aced them except for animal phys; I’d spent a year in an entomology program at another university and got an A in insect phys there, so my committee placed me out of having to take animal phys (all of my research was going to be on insects). At the time, I thought I was going to be able to choose four out of six questions — no problem. Make this around March; Animal Phys was going to be offered in the seven-week spring session, and at this point there was going to be no need to take it.

    This is where the two “oops”-es happened. One was that the department decided on fairly short notice to require candidates to take the exam in all six areas. The other was that I found out that I needed surgery for a benign tumor that wasn’t guaranteed to stay benign if not removed, that the only time my surgeon could schedule it was right in the middle of spring session, and that I should plan for a minimum of two weeks out of commission.

    After a flurry of frantic e-mails to various department members, we came up with a solution. Instead of having to register for Animal Phys, I was permitted to audit it. I missed the middle of the session to have surgery, but spent a lot of time afterwards on the couch with a bunch of good books. (Including, of course, my Animal Phys text and notes.)

    Passed all six questions, had a couple of drinks afterwards with a fellow comp-taker, and continued the celebration (“Six questions, six passes, no cancer!”) by going to an ecology lecture that night. :-) (Actually, I didn’t get formal confirmation that I’d passed until later that summer, but I got that good news on my birthday, which I’m sure contributed to another celebration.)

  8. #8 Pam
    August 24, 2006

    never check your answers after you’ve handed in your test.

    Never say never. Part of my candidacy exam was a 1-hour oral (that took place a few days after the 3-day written exam), the point of which was to give students a chance to redeem themselves over questions they flubbed on the written. So, like, checking answers after the test was vital!

    And it’s true what Chad said: it is cheering to know that, whatever else I do in my life, I will never have to take another physics Ph.D. candidacy exam.

  9. #9 Simple Country Physicist
    August 24, 2006

    I think we all have some embarassing story to tell about quals and/or defense, if not both. Mine arises from taking the required coursework from a professor other than the one who wrote the exam. What was actually almost a freebie I saw as something more substantive. Happily the writing professor was so taken by my approach to the problem that he offered to pass me if I would collborate with him on a paper on how I saw the problem.

  10. #10 Stephen
    August 24, 2006

    I just finished my quals half an hour ago….i’m starting to blog the experience at http://www.astroh.org.

    my brain hurts and/or is spilling out my ears

  11. #11 Craig
    August 24, 2006

    Our qualifiers (Materials Engineering) were all oral, and designed more to make sure you could stand up under pressure and have the fundamental cross-disciplinary knowledge than to weed people out. (Unlike the neighboring ECE program, with a 40% failure rate, Materials tries to eliminate people during the application process.) The end result is akin to Kate’s bar exam experience – studying is hell but the actual exam wasreasonable.

    Except, of course, for the classmate who was given the exam for the wrong subfield by a confused professor – his test was pretty damn stressful.

  12. #12 Chris
    August 24, 2006

    I’m in the middle of studying for mine as we speak *grumble*grumble*

    Ours are bit different than what you described. We’re required to write a grant proposal on a topic other than what we work on in lab. We resarch and write the proposal and then do an oral defense of it. It’s designed to test our ability to do real science, rather than just our ability to memorize facts.

    Some days, I think I’d prefer the rote memorization . . .

  13. #13 Jonathan Dursi
    August 24, 2006

    My colleague Ryan Scranton had particularly vivid analogies to describe the qual experience which he shared at some length with the incoming grad students over beers for the next couple of years after our own grillings. Typically we’d tell war stories about the miserable, horrible experience just as that time was coming for the next crop. It wasn’t that we were *trying* to demoralize them, you understand, we were just reliving the experience aloud.

    I’m probably not alone here by saying that for well over a year afterwards I would occasionally have dreams of suddenly being thrust into the exam being completely unprepared — having forgotten the test was that day or some such. What’s weird is that I had no such dreams before the actual test.

  14. #14 Jeff F.
    August 24, 2006

    We’re required to write a grant proposal on a topic other than what we work on in lab.

    I’m kinda jealous, really. I believe that the qualifying exam in physics is a messed up process that serves no other purpose than to further torture students who’ve already proven their worth. I wish instead we had something like you described, because it’s actually useful. Grant writing is one area of training that we don’t typically get in physics (or at least in the group that I worked in) until your actually a faculty member now needing said grant writing practice.

  15. #15 Kelly Sheppard
    August 24, 2006

    Writing the two grant proposals (thesis and non-thesis) for my qualifying experience really got me to think about the experiments I was purposing and the process of writing a grant. We had workshops and met with faculty members in small groups to review what we were writing that helped as well. A good committee is a must. The chair of mine has me come into his office and we discussed the ins and outs of enzyme kinetics and he had me justify what I thought were the key papers in my subfield. Others had committees who were there to torture said students.

  16. #16 GF
    August 24, 2006

    My nerves were shot before I even showed up for my qual. partly it was because I was a (somewhat) older student, having not just come out of undergrand, with something of an inferiority complex. Mainly it was due to something that happened while I was getting there. The qualifying exam at Cornell Applied Physics is a two day written exam that comes at the end of winter break after the third semester, which puts it during the coldest part of the Ithaca winter. Normally, I think of winter in central NY as being pretty mild, since previously I had always lived in MN or WI, where it is consistently colder. That day, however, it was -40 F outside. On top of that, at the time, my wife and I did not live in Ithaca, rather we lived in fair Whitney Point (due to a commuting split) which is about 35 mi away. When I left that morning, I got about a third of the way there before my car overheated. I had to pull over to let the engine cool down, all the while cursing. I already had a serious case of nerves over the qual, but this about pushed me over the edge. After about 10 minutes I restarted the engine and tried to keep going. A few more miles down the road (Richford, actually) the needle started edging back into the red zone. Curiously (and thankfully) it all of a sudden started cooling down before I had to pull over again, and I was able to get the rest of the way there. It turns out my coolant had frozen in the radiator, and it wasn’t until the entire engine block warmed up for long enough that it could flow. Anyway, I did make it to the test on time, and although I seriously sweated it, I passed them all. (Whew!)

  17. #17 agm
    August 25, 2006

    No quals here, just an MS to defend and submit for publication/do up as a thesis as proof of being good enough to try the big one. Rumor is that they decided the classes were brutal enough that you’d fail out anyways if you couldn’t hack it.

  18. #18 Charles Kuffner
    August 25, 2006

    Well, as someone who did not make it through grad school (my subject was math), my experience was a little weird. Rice had three math quals (algebra, analysis, topology) which you needed to complete by the middle of your third year. (There was also an advanced test, in your chosen subject, to be done before you started on your thesis. I have no idea how stressful that was.) I took and passed the algebra test at the start of year two. By the time I sat for the analysis test a year later, I and everybody else knew that grad school was a bad fit for me and that my days were numbered. As such, I didn’t prep very much for the test, and was rather surprised to learn that I’d come close to passing it – close enough, I believe, that had I still been a serious student, I’d have gotten a do-over. I had this strange mixture of being disappointed at failing and yet relieved for not having succeeded, which meant my exit path, which I was ready to take, was clear.

    Who knows what might have happened had I wanted to pass the analysis test? Of course, the topology test was supposed to be the real killer, so it’s just as well that I never got to find out. I’m happy to have left that road not taken behind.

  19. #19 Bill Hooker
    August 25, 2006

    We’re required to write a grant proposal on a topic other than what we work on in lab.

    That’s the system to which grad students here are subjected, and it drives me nuts. Why in the name of all that’s efficient not have them write a grant for what they ARE doing in the lab? The only answer I ever get back is, “because their advisors would write it for them”. Well, bullshit. Most of them wouldn’t, and it wouldn’t be difficult to catch most of the ones who did. Besides which, you’ve given this person tenure, now you’re saying you cannot trust them to do the right thing by their grad students? But you’re still happy to let them take on students in the first place? WTF?

  20. #20 kstrna
    August 26, 2006

    That’s the system to which grad students here are subjected, and it drives me nuts. Why in the name of all that’s efficient not have them write a grant for what they ARE doing in the lab? The only answer I ever get back is, “because their advisors would write it for them”. Well, bullshit. Most of them wouldn’t, and it wouldn’t be difficult to catch most of the ones who did.
    **************************************
    Bill in my program we write an anti-thesis as we call it to show breadth of knowledge. On the extremes they want to make sure someone in a molecular genetics lab knows something about biophysics/structural biology and vice versa. Since the boundaries between the subfields are failing it is less clear cut but you still have to show you can think about designing techniques and questions outside of those addressed in current lab. Of course in our department we also have to write up the thesis and defend it at the oral exam along with the anti-thesis. There are some professors who do help their students write the thesis proposal. Some do not because the rules state they can’t and others don’t because they are too busy. For the first group, it does create an advantage on the writing section. For the oral exam though the student is on his/her own as advisors are not in the room so no matter what a student needs to know what they are talking about.

  21. #21 Jordin Kare
    August 26, 2006

    Ah, qualifying exam memories. As I recall, when I took them at Berkeley, it was 4 3-hour exams (mechanics, E&M, thermo, quantum?), over 2 days, with 2 chances to take them. I thought I’d done well on all except the quantum physics exam, which I’d blown. When the results were posted, I’d done near the top of the class on the quantum exam — they’d made it a bit hard that year….

    But my *most* vivid memory from the written exams had nothing to do with the content. I’d recently been given a (for the time) rather nice pocket clock/calculator which included a countdown timer. I’d set it to show the time left in the exam, but I set it a bit short, and I hadn’t realized that the countdown timer triggered the alarm. So there’s a room full of incredibly tense grad students, heads down, no sound but scratching pencils and occasional hairs being torn out, and this stupid alarm starts going BEEP! BEEP! BEEP!….and I couldn’t get it to shut off! I hadn’t used the alarm function before, and had no idea what button to push.

    I finally sat on it.

    That didn’t actually stop the alarm, but muffled it enough that everyone stopped glaring at me and went back to work. But I’m amazed I didn’t cause at least one heart attack.

    Berkeley also had oral qualifying exams — one for classical and one for quantum — an hour at the blackboard with 3 professors asking questions. In the classical exam,
    Frank Crawford asked me: “How does air drag vary with velocity?” which I answered correctly (as I’m sure most students did) after a bit of thought: F = const*v^2.

    Then he pulled out a stack of cup-shaped coffee filters and said, “Do an experiment to test your theory.” I suspect a lot of students were completely flummoxed at that point; when I responded right away (by stacking up 4 filters and showing they fell twice as fast as one) he said, suspiciously, “Have you done this before?”

    I ended up working in the same group as Frank Crawford at LBL, and he was my kind of scientist; he invented new musical instruments (the corrugahorn) and published an article titled “The Hot Chocolate Effect” in Am. J. Physics.

  22. #22 Scott Spiegelberg
    August 28, 2006

    I’ve written up my qualifying exam experience here. Music theory is a different bag from physics, but I always enjoy pointing out to my brother, the PhD in ChemE from MIT, that I used harder math in my dissertation than he did in his.

  23. #23 Scott Spiegelberg
    August 28, 2006

    I’ve written up my qualifying exam experience here. Music theory is a different bag from physics, but I always enjoy pointing out to my brother, the PhD in ChemE from MIT, that I used harder math in my dissertation than he did in his.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.