Advice For Physics Majors (a serious article, for once)

Furthering the pain part II — Physics grad school

How do I make sure they pick me?

So you’ve narrowed it down to Yale, Harvard, and MTDS (‘Merican Truck Driving School). What do you need to do to get in? Let me go over the most important parts of your application, in order of their importance. Again, this is only from my personal experience — I’m not and never have been on any application committee. This info is a summary of advice given to me, and observations I made, during the application process.

1.  Courses you’ve taken, and grades you had in them

The most important part of your background. You have to have a solid undergrad curriculum in physics and math, and have a good GPA in the courses. Doing a really hard curriculum with lots of advanced courses won’t do you any good if you got D’s in all of them — the schools are really looking for the A students.

2.  Research experience

You don’t necessarily have to have anything published, but some kind of senior thesis, or extended independent research, or the like is close to mandatory. Publishing a paper, or being a coauthor on someone else’s paper, is pretty common as an undergrad for students going to the top schools. Try to get involved in someone’s lab for a few hours a week, roughly the same time commitment as another class (but don’t let it drag down your work in your regular classes!). I made the mistake of being too intimidated to ask professors for a job, and then once I got up the gumption, I made a poor choice of research topics for my skills. Generally, experimental work is going to be a lot easier to do as an undergrad, since you don’t yet have the advanced physics and math background to handle theory. I chose to try theory, but never really got anywhere, and wound up quitting my project about halfway through my senior year. I really didn’t have anything to show for it, beyond doing some graduate level homework problems — not really anywhere near a research project. The friends of mine who did become successful in undergraduate research joined a lab back in their sophomore years, and worked on research projects in those same labs through their senior year. Note that it isn’t necessary, though it is helpful, to pick a topic that you’ll be doing in grad school — the demonstration that you have the tenacity and work ethic to complete an independent research project is more important than getting a head-start in your chosen field. I couldn’t find anyone at my college who worked in my field of interest, so I was perhaps not as motivated to do research as I should have been. Looking back, I now know that research in any area would have helped me.

And of course research is good for you too — it’s probably going to be your first real job, a job not involving big deep-fryers or washing plates. Even aside from the specific research area, working in a lab teaches you what a job is like — how to solve problems that aren’t answered in the back of the book (in fact most likely haven’t been answered anywhere), how to work on a project with other people, and maybe some lab or computer skills you won’t pick up in a classroom. Most of all it gives you the chance to see what grad school is really going to be like — if you don’t try it out first, how do you know you’re going to like research? If you’ve got the time, give research the time commitment equivalent to about another physics class, and just count on making that commitment for your entire senior year (or earlier if you can).

3.  GRE scores — physics subject test

This test of course measures you against all the other schmoes vying to get into your dream school, and so is your chance to completely flatten them. In my case, I started dabbling around with some practice tests in the spring of my junior year, started studying for real later in the summer, and was studying a few hours a night that fall quarter of my senior year. I took the test twice, once in October and once in December of the senior year fall quarter — both tests were in time to get to graduate schools, but late enough that I was able to take the last few classes I needed to fill in my test preparation. Oddly enough, I scored better on the December test, even though I had studied more intensely for the October test — perhaps the extra time allowed physics to get a better “foothold” in my brain. I considered studying for the test to be equivalent to another class, in time-commitment. (And of course independent research is another “equivalent class”, as I mentioned above, so budget your time wisely!).

By far the best studying tool is practice exams — all the way down to taking tests under the same conditions as the real tests, under the same time constraints. Practice tests for physics are hard to find, so conserve any real tests you find — don’t look at them until you’re really ready, and then take them under real testing conditions, so you know how well you’re doing. I found that the GRE company’s own review book, which has a few older physics tests, was the closest to the real test. There were one or two other companies putting out example test collections when I was studying, but they tended to have errors here and there and weren’t as good reflections of the true test difficulty level.There is a surprising amount of “basic” physics on the test, the kind of simpler topics and formulas you learn in your freshman physics course. Stuff like the Rayleigh criterion, physics of fluids, lenses and other simple optics — the more basic, surface-skimming of a wide range of topics. Because of this, someone who really knew the freshman intro-to-physics course inside and out could do pretty well on the GRE. And because of this, you can do yourself a favor by going back to that freshman book and memorizing some of those equations you thought you’d never see again.

Old homework problems are your next best friend, and since real practice exams are really scarce, you’ll probably spend most of your study time redoing old homework, or doing more problems in your old textbooks. (You do keep your old textbooks, don’t you?) You should know this all too well by now, but reading and understanding a physics book does not mean you can pass the test — the only surefire way to study for the test is to practice with similar problems, as much as you can stand. (I’ve mentioned this before, in talking about homework being the best way to study for physics tests — still applies here, and will still apply in grad school.)

The GRE general isn’t all that important, though everybody requires it. I studied about a week or so for it, did a few practice tests, took the real test once, and left it at that. If your math score is abnormally low, work to get it up, but this test should really be the easy hoop to jump through. Most everyone going on to physics grad school will get close to perfect on the GRE math, and you probably will too.

4.  Recommendations from teachers

Make sure you develop some good relationships with a few teachers, who will be willing to vouch for your thunderous physics skills. (It’s just like high school again, isn’t it?) A research advisor would be a great choice, but other teachers whom you’ve taken several courses from are also typical. If you get one from someone outside of physics, there better be a very good reason — the vast majority of the recommendations are going to be from physics professors. (A math teacher or research adviser outside of physics is of course okay — what I mean is, you can’t get your pastor, scoutmaster, or mailman to write something, as they can’t really attest to your physics skill.) I gathered four recommendations — most schools ask for three, but there is the occasional school or fellowship that requires four.

The recommendation process is a pretty decent amount of work for your professors, so buy ’em a beer (or something) afterward to show your appreciation. Almost every school is going to ask you to give a form from their application to your recommenders, and have them fill it out and mail it off to the school themselves. I didn’t want to do this, considering it to be a bit too much to ask of them. Instead, I asked my recommenders to give me 20 or so copies of their letter to me, in sealed envelopes with their signatures across the flap — and I didn’t open them to read them! It’s very important that the schools know you have not read your reviewers’ letters — the letters can be considered much more candid and genuine if it is known you haven’t been reading them. When I filled out a particular application, I just dropped in a copy of each letter into the envelope — even though this was against the “rules” of some of the schools, I didn’t get any complaints from the schools, and I felt it was more considerate to my recommenders.

5.  Research essay

For most schools, and for all fellowship applications, you will be expected to write an essay explaining yourself, what you hope to do with your research, that kind of thing. This essay should really be written like an academic research paper, proposal-like — use the same kind of academic, impersonal voice you’d use to write a publishable research paper, or your senior thesis. Mine was a bit too informal, and too vague. I didn’t know all that much about “neurophysics” at the time, and so what I wrote was pretty silly, looking back — it would have been pretty obvious to anyone in the research area, and probably everyone else on the review board, that I really didn’t know what I was talking about. The essays that seemed to be the most successful that I’ve seen were very specific, describing a research area in detail and proposing legitimate, reasonable research topics — what you’re really doing is giving a first stab at what you want to do for your graduate research, so what you describe should be on the level of a PhD thesis. My essay was too broad, describing the field in general and where I thought it was going — it described lots more research than I could possibly have done for a PhD, and didn’t describe anything in enough detail. (And it was completely wrong about the field, but that’s another lesson…)

So what do you do if you’ve been doing research in one area, but you plan on switching to a completely new area and you don’t know enough to write a coherent essay on the new area? Well, you’ve got two options. First, you can do an intense study session in the new area, try to catch up, and do the best you can. This is your best bet, but it is a real gamble — you may wind up like me, clueless about your proposed area, writing an essay that must have made the reviewers spit their coffee out their well-educated noses. Your other option is a bit sneaky — go ahead and write an essay proposing to continue the research you’ve been doing, which you know enough about to write a coherent detailed research proposal. Even if you wouldn’t touch that topic again with a slide rule, you are best prepared to write your essay in that area, and may give yourself the best shot to show your skills by doing this. This is also a gamble, because you are going to have to discuss your research interests when you go to interview. But it may be a particularly effective strategy for fellowship applications, where it doesn’t matter so much whether you plan on continuing in that area.

When writing an essay for a particular school, try to tailor it towards that school. If you write all about your interest in the physics of soap bubbles to a school with only accelerator jockeys, the school will know you didn’t do your homework and wonder why you are applying. Of course if you’ve chosen the right school, their research will reflect your interests — so mention that in the essay, bringing up their research by name. A particularly impressive strategy is to contact the professors beforehand, and if they seem interested, ask them what kind of research you’d be doing in their lab. Then you can mention the research, and the fact that they are interested in you, in the essay.

The application process

In a lot of ways, you might find the whole process to be a lot like applying for colleges, and it is. Most of the same lessons apply. You’ve got to spend some effort in undergrad to collect an academic record that will be impressive to the schools (although extracurriculars don’t matter for zilch now — only your physics- related record will be considered). You can probably aim to apply to 5-10 schools. I applied to 13, which is on the high end. You’ll collect a lot of color pamphlets from schools. You’ll have to stay extremely organized to remember deadlines and application requirements. You’ll spend a lot of time at a computer, looking at the schools’ web pages and working on your applications (many schools and fellowships can take electronic applications these days). Hey, it could be worse — have you heard about the Hell On Earth that is applying to medical school?