The Grad School Application Process

I'm teaching our senior major seminar this term, which means that once a week, I'm giving hour-long talks on topics of interest to senior physics majors. This week's was "How to Pick and Apply to a Graduate School."

I've probably written this basic stuff up about three times already, but I'm too lazy to look for it, and this particular presentation was slightly different than anything I may have put on the web in the past. And I might as well wring another post out of the topic, while it's fresh in my mind...

There are several steps to the grad school application process, but the most important part of the process comes at the very beginning. Step Zero: Do you really want to go to graduate school?

(Continued after the cut...)

There are lots of bad reasons to go to graduate school. Chief among them is inertia-- college seniors have been in school for something like seventeen years, and from that perspective, it can seem like going to school is the only thing they know how to do. Grad school seems like a continuation of that process, and so some people drift in that direction.

This is a terrible reason to go to graduate school. It's just postponing the inevitable process of getting a job, and the reward isn't worth the hassle. If you're thinking about grad school just because you don't know what else to do, find some other ideas. Get a job for a while. Take a year off and travel. Do something other than being a college student for a little while, and then see if you feel like going to graduate school.

Another weirdly common idea is that there's good money to be made by going to grad school. I'm not sure where this comes from, but it's not accurate. Don't get me wrong-- small-school academics are at the low end of the Ph.D. pay scale, and we're comfortably middle-class. But you're not going to get fabulously wealthy by getting a doctorate.

There's one and only one good reason to go to graduate school: You should go to graduate school if you really enjoy doing research, and want to make a career of it. Grad school is designed to train you to do research in a particular field of science, and not much else, and a Ph.D. is pretty much a prerequisite for a career as a researcher. You can do other things with the degree-- plenty of Ph.D. physicists work as engineers, or programmers, or financial analysts-- but there are easier ways to get those jobs, ways that don't involve graduate school.

The only thing you need a Ph.D. for is a research career, and that's the only reason to think about getting a Ph.D. in science.

Step One: Determine your research interests. By this, I don't mean that you should plan your thesis project before you settle on a grad school, but a lot of the school search process depends on what sort of thing you want to do. Anything you can do to clarify your interest in research-- and you are interested in research, as we established in Step Zero-- will make your life easier down the line.

"Determine your interest" in this context can be broken down into two big questions. The first is simply "Broad or narrow?" Are you interested in a range of different fields, or is there one specific field that really grabs you? If you have broad interests, then you're going to want to shoot for schools that are good at a bunch of different things. If you have narrower interests, then you're going to want to make sure that the schools you apply to have strong programs in your desired field.

The other big question is "Theory or experiment?" This can be a tough one to answer, because many students have a skewed idea of what those categories mean. Some students seem to think that theoretical research is just like doing problems for class, and the experimental research is just like the stuff you do in lab classes. They're wrong, on both counts. This is an area where actually doing some research will help. If you work on an experimental project as an undergrad, and really like it, you might be an experimentalist; if you hate it, you might think about theory.

(In a purely strategic sense, I'm told that graduate programs in physics get many more applications from students saying they want to be theorists than they get from students saying they want to be experimentalists, so there may be some slight advantage to saying you want to do experimental work. I doubt the gain is big enough to be worth lying about, though.)

Step Two: Investigate graduate schools. There are lots of ways to look at grad programs. US News helpfully provides rankings of graduate programs (I'm not willing to pay to read them, but you might be). A site called has a handy online ranking generator. And Google is your friend when it comes to finding information about various colleges and universities.

And, of course, you can always use the old-fashioned technique of actually talking to people. If you're a college student, you're surrounded by faculty, all of whom have gone to graduate schools, and will have opinions and advice to offer. Ask them. If you're in an active department, there's also probably a colloquium series of some sort bringing in speakers from other institutions-- talk to them, if possible. Many of them will be actively recruiting students, and will be happy to talk about their school. And if you're completely desperate, there are even cranks running weblogs, who will happily hold forth on just about anything.

Step Three: Apply to lots of places. The meat of Step Three is really the application process, but it's worth emphasizing that you should apply to a large-ish number of schools, to give yourself a range of options. Yeah, there's usually an application fee, but it's better to spend some extra money and have choices than to save money on applications, and not have any options.

The application process isn't too different from the college application process-- you have to provide transcripts, and test scores, and write an essay, and all that stuff. Some elements that may not be completely obvious:

First of all, you're going to need recommendation letters from faculty. Make sure there are faculty who know you well enough to write a good letter, and ask them in advance. Going to a faculty member and saying "Um, Dr. X, can you write me a recommendation for Harvard? By the way, the deadline is tomorrow..." will get you a letter that starts off "Student Name is the sort of jackass who waits until the last second to ask for a recommendation...," and you don't want that.

(OK, it's not that bad, but you won't get as good a letter as you would if you gave them more notice...)

Also, don't be afraid to ask advice about your statements. Show a draft to your faculty advisors, tell them where you're applying, and ask what they think. And for the love of God, listen to their advice-- if they say "You should apply to better schools," apply to some better school. If they say "These schools are kind of a stretch," apply to some lower-tier schools, too.

If you get the chance, it probably doesn't hurt to visit the schools you're interested in at this stage. I don't know that I'd fly cross-country on your own money to check out a school on the opposite coast, but if you're applying to Harvard, and you're going to be in Boston anyway, swing through Cambridge. Take a tour, talk to whoever's around. It can't hurt, and might help, either in making up your mind about applying, or by fixing your name in the mind of the people who will read your application.

Step Four: Acceptance. Assuming you're accepted to at least one school, how do you decide what to do?

The one absolutely essential item here is to visit the school. Many schools will actually pay for you to come out and see the place, and you should take them up on it. It gives you a chance to see the campus, visit some research groups, and get a sense of what you're in for. Even if you only get accepted by one school, visit before you accept their offer-- if you go there and hate the place, you can always get a job for a year, and try another round of applications later. Better that than going to a place where you'll be miserable.

When you visit, make sure you talk to some students. You'll definitely get to talk to some faculty, but the students are the important ones. They know the quality-of-life stuff that will really matter to you, and they're less likely to try to sell you a line of bullshit. If they're miserable, they'll tell you. If they're happy, they'll tell you that, too.

When you talk to people, find out about departmental policies, which vary quite a bit. Are there qualifying exams or not? What are the pass rates on the qualifiers? Is there a limit to the length of a graduate career? What is the funding situation like? Are there generally research assistantships available, or will you be expected to teach to make ends meet?

A critical thing at this stage is a sort of analogue of Yog's Law: You Do Not Pay to Go to Grad School in Science. If they're not offering you a tuition waiver and a TA job at the very least, don't go. You're not going to make big money while in graduate school, but you shouldn't have to pay for the privilege. Better to be paid for flipping burgers and apply again next year, than to shell out good money for the right to work problems from Jackson.

Another thing you might want to do is to try to figure out where you stand relative to the rest of the students. How do your test scores compare? How does your undergraduate preparation compare? If you're way at the upper end of the distribution, you could probably get into a better school, if you're way at the low end, you may be in for a struggle. This is a little dicier, though, as that information can be a little hard to get.

That's a quick and crude sketch of how to go about the grad school application process. If you look around a bit, I'm sure you can find no end of other advice (some of it written by me), much of it broadly similar to what I've written here, and much of it contradictory. This is what I'm offering this week-- thoughts and comments are welcome.

More like this

It's that time of year again, when eager undergraduates start thinking about their futures, including the possibility of graduate school. This inevitably leads to emails of the form "Hi, Professor, could you write recommendations for me for these nine schools? And by the way, they're due Friday.…
Sean Carroll is offering more unsolicted advice (though it is in response to a comment, which makes it borderline solicited...), this time about choosing an undergraduate school. He breaks the options down into four categories, with two small errors that I'll correct in copying the list over here:…
This is a repost from my old blog, from a year and a half ago. But it's time for academic positions to be advertised - if they haven't been frozen due to budget cuts. So, some old advice on getting a job, while my own job is keeping me especially busy. So. You want a job, do you? At an…
As a newly minted Associate Professor, I sort of feel like I ought to say something about the recent tenure discussions. These were kicked off by Rob Knop's recent despairing post (though it should be noted that Rob's been worried about this for a while), and most of the discussion has taken place…

When should undergrads start thinking about grad schools?

When should undergrads start thinking about grad schools?

As soon as possible.
Well, ok, I can be more specific than that.

You need to think about what specific schools you're going to apply to sometime in the fall of your senior year (assuming you plan to go straight to grad school, anyway). The application deadlines are usually in February/ March or thereabouts, and the deadlines for NSF fellowships (if you stand a chance at such things) are in November. You should at least be thinking about grad schools in the fall.

If you might be interested in grad school geenrally, there are a few things you ought to start doing even earlier. For one, make sure that there are several faculty who know you, and have a positive impression of you, because you're going to need letters of recommendation. This means something a bit more than getting an A in their class-- they need to know something about you as a person as well.

It's also a good idea to get involved in research projects as early as possible. Not only does this enhance your application, it also helps you learn whether you're really itnerested enough in research to be considering graduate school. Look into the summer research options available at your school, and at various other schools and national labs.

"There's one and only one good reason to go to graduate school: You should go to graduate school if you really enjoy doing research, and want to make a career of it."

Careful... I would add to this - you really enjoy TEACHING and wish to make a career out of it. I have a position at a Small Liberal Arts College that requires me to do very little in the way of research, but a position like that definitely still requires a PhD.

Chad Orzel wrote:

"It's also a good idea to get involved in research projects as early as possible. Not only does this enhance your application, it also helps you learn whether you're really itnerested enough in research to be considering graduate school."

I'm a biologist, not a physicist, but I STRONGLY that last point. If you think you might be interested in a research career, the most important thing you can do (IMHO) is start doing independent research projects by your Junior year of college.

Real-world scientific research is quite different from what you do in a lab class. You really need to do a project that involves multiple experiments over an extended period of time (months). That's the only way to get a decent sense of whether you'll really like research. If so, great - you'll be confident in your career choice when you apply to grad schools the following year. If not, no problem - you'll have time to reconsider whether grad school is for you.

"A critical thing at this stage is a sort of analogue of Yog's Law: You Do Not Pay to Go to Grad School in Science."

Under no circumstances should anybody pay to go to a Ph.D. program in anything. For a master's, only if it leads to guaranteed professional opportunities.

If a Ph.D. meant an almost certain tenured academic position, it'd still be far too expensive to pay for one out of your own pocket. It'd be like buying a house. Given that the job market for a Ph.D. stinks, and will do so for the rest of our lives, it's a far worse investment.

In addition, the 'conventional wisdom' is that, when a program does not offer support, the program is letting you know that they think of you as a very marginal candidate.

Back in the day (I don't know if this is current, although I suspect it is), Yog's Law for Grad School was a lot harder to show fidelity to if you were on the humanities side of the fence rather than the science side...

Even so, I do wish someone (maybe Yog--who I didn't, and don't, actually know) had pounded the point into my head until I said "fine, I won't go, I'll get a job doing X". Ah well, bygones....

By Trent Goulding (not verified) on 22 Sep 2006 #permalink

Don't forget the very useful AIP factsheets on different grad schools. There you can discover many interesting things such as how competitive departments are in terms of number of applications received and number offered admission. It also includes information about average GRE scores and funding. For condensed matter types, you may notice that you have a better chance getting into applied physics than physics and can often work with the same faculty. See for example the data from Stanford:
Applied physics: 46 accepted from 161
Physics: 76 from 485

You can search for departments here:

I'd add two things: first, get realistic information about the job market in your field. The academic job market in many fields is depressingly bad, but that's not true of all fields; and some fields have lots of jobs for people with PhDs who don't want to go into academics. It's better to make decisions with open eyes.

Second, find out what the statistics for your field are for job placement and time to degree, and check the schools you're accepted to (or even applying to) to compare their statistics. Look for the real statistics, and not how fast departments say they expect students to finish.

Rather than thinking simply about the job that you want, think about the lifestyle that you want to have and whether that job will support it.

Do you want to live in a big city or small town? What cities do you prefer? You should realize that certain specialities could limit your options when it comes to where you live.

How materialistic are you? I don't mean this in a bad way, but if you like designer clothes, nice cars, and expensive restaurants, academia may not be the place for you.

By My 2 Cents (not verified) on 22 Sep 2006 #permalink

I got into grad school by the "random chain of coincidences" method, more or less. I needed a project for my Master's degree and since I didn't want to do it at the CompSci department I walked over to the department of theoretical philosophy, which had/has a research group in cognitive science. It turned out one guy had a computer-science related thing that needed to be done anyway so I did.

After graduation I started a job, but as chance would have it, my advisor called me to talk about making my master's project into a publication at a point where I was really frustrated at my current employment (a stupid thing involving a broken coffee pot and lack of project oversight), and by the way, the department has just gotten money for a couple of PhD student positions and would I be interested?

I did not go because it was a good career choice. I did not go because I wanted to teach (at that point, the research group did not do any undergraduate teaching at all). I did it because those months doing my master's project had been the most frustrating, stressful but fun time I'd had in my life, and doing a PhD promised to extend it to five years. Which it did.

As a side note, PhD students In Sweden are today almost always paid employees. The deal is that you get five years employment out of which 20% of the time you're available for non-research work - usually teaching, but also administrative tasks, technical support or similar. That works pretty well, especially since quite a few people want to start families at that time in their life and being a regular employee means you have the right to paid parental leave, sick days and so on just like any other employee.

Rather than thinking simply about the job that you want, think about the lifestyle that you want to have and whether that job will support it.

Do you want to live in a big city or small town? What cities do you prefer? You should realize that certain specialities could limit your options when it comes to where you live.

That one is a big one. It just isn't how passionate you are about doing research. It is how passionate you are about doing research relative to other activities in life, other pursuits. Many I know who have left science to pursue other things either dropping out of grad school or after they received their PhDs loved research. The problem was they had many other loves. Whereas others I know who aren't as passionate about research but don't have as many other passions tend to enjoy grad school more and tend to be the one's who want to pursue being research faculty.

You do need to figure out who you are and what you generally want in life. If you are not sure, take a year or two to figure it out.

Looking at the job market is good but don't limit yourself to looking at research faculty positions or faculty/research positions in general As others have noted, not all PhD students can become faculty at research universities. Anyone who sells you that is selling a pyramid scheme. Ask for information about what students from the grad program end up doing 1, 5, 10 years out. If they don't have that information then you should question if that is the program for you. Look at the choices available to you. Do more than one of the options interest you? If not, then grad school is probably not for you. You can love research based on undergrad and working but after 5 years of grad school your opinion might change. Do you have options then? Or will you be stuck? Many of my classmates get recruited into high paying jobs as consultants (still long hours) for consulting firms and many times end up working for companies that had been their clients so there are other options out there. A good grad experience should give you skills beyond lab ones. The most important of which are analytical skills. Those are very valuable.

Look at the research of many faculty members in program. You may love the research of one faculty member on paper but actually being in lab doing the research is a different ball game. Give yourself options. Plus you never know, the person just might not be the right person do be your mentor.

I think most of this is generally true in other sciences.

Like qetzal, I want to emphasize the idea of doing research early.

Once in grad school, you'll only take a handful of classes, and they'll be over in the first 1.5-2 years. Most of your time (and it's a lot - long hours and for an average of around 6 years) will be spent doing research, so you'd better make sure you like it.

What hasn't, I don't think, been mentioned yet is that a good amount of research looks really good on your application. The impression of the schools will be:
* You've got a head start, so you'll be productive sooner (remember, they're paying you, so they want a return on their investment).
* You've seen what research is like and still want to do it, so you'll be less likely to leave (again, they're paying you, and they don't particularly want a student whose going to cost money but not stick around long enough to produce results).
If your research before grad school has actually been productive, I'm sure that's all the better, but they don't necessarily expect you to have accomplished much, just to have gotten your feet wet.

This is great advice. I would only add the following to the string of insightful postings above. If you enjoy science and/or engineering and would like other options besides the PhD-track graduate programs, I would urge you to investigate the Professional Science Masters Programs ( available at a number of schools in a myriad of fields. These programs offer a stimulating, high-value career track for those interested in science and engeineering careers outside of academe.

I have a different question. I'm finishing a PhD in Materials Science in December and would appreciate some advice on transitioning into astrophysics-related work. During my somewhat protracted graduate course, I've come to realize my talents are more theoretical, scholarly and presentational than experimental/hands-on. I think I'd make a great astrophysicist, but an unexcelling Materials Scientist, despite my doctorate. One skill I've developed great expertise in is finite element analysis (FEA). Plus I'm am a talented technical writer. Could I use these skills to transition into Astrophysics? Must I obtain another PhD?

By Ian Nieves (not verified) on 04 Nov 2007 #permalink