What would I tell someone asking me for advice on how to pick a graduate school ? Based on my last twenty years spent as a physics graduate student (5), postdoc (5) and faculty (10), I feel that I can at least offer my two cents worth.
Disclaimer: I am not discussing here the “right and wrong reasons to go to graduate school” , focusing instead on the one decision that all graduate applicants (who have already come to the conclusion that graduate school is the right option for them) have to make, namely: where ?
This decision will have serious, long-term repercussions on their careers. What follows is drawn from personal experience, and therefore may be in some respects more relevant to potential physics graduate students; however, it is my sense that some of the considerations may well be generally applicable.
Finances. I know, I know, this ought not be the overriding criterion, but I am placing it at the top of the list anyway, because it is an issue of the utmost importance, one that is often not given due consideration, neither by applicants nor by departments and universities. It is fully understood that graduate school is not meant to be a money making proposition, but graduate students should be allowed to focus on their education, without worrying all the time about their finances, much less having to take a part-time job to make ends meet. Applicants should especially watch out for the following, when considering an offer of a teaching or research assistantship, concomitantly with their admission into the graduate program at a given institution:
1) What is the take home pay ? The stipend may be clearly indicated in the offer letter, but what about tuition and fees ? Are they waived ? A clear, unambiguous statement to that effect should be requested. If students are responsible for their tuition and fees, it is imperative to find out to how much they amount annually, for that can significantly reduce take home pay, especially for out-of-state and foreign students who may not establish readily local residence.
2) Is the take home pay adequate, considering the local cost of living ? Is subsidized university housing available ? If not, how much should one expect to pay for rent ? If living off campus has to be considered, is public transportation available, or is a car needed ? Does the university offer some form of comprehensive health coverage ?
These are all appropriate, legitimate questions that applicants should not be afraid of asking of graduate chairs, requesting written responses whenever appropriate. For departments and universities to be less than 100% forthcoming on financial matters is unacceptable. Graduate students should not be, and are not in it for the money, but they should not be expected to skip meals either.
Moreover, the inability of a graduate program to provide a decent financial package to its own students may be a sign of seriously limited resources across the board, also affecting, e.g., office space, computing infrastructure, equipment, conference travel funds and so on.
If the take home pay is insufficient, and if nobody else can foot the bill for you, I strongly recommend against trying to “make do”. Taking a side job is a bad idea, as graduate school is not meant to be a part-time endeavor. Look elsewhere instead.
Program effectiveness. Clearly, this is the most important assessment to be made; applicants must be willing to spend some time researching graduate programs under consideration, as well as individual prospective supervisors. The fundamental question that one should be asking is “Is this graduate program going to help me achieve my career goals ?”. I think that “customer satisfaction”, is by far the most reliable indicator. In other words, how successful have program graduates been at securing the type of career opportunities that applicants wish for themselves ? What fraction thereof go on to high profile postdoctoral appointments, eventually landing faculty jobs or employment at research laboratories ? How well do those who choose to pursue activities outside academia or research ?
Similar considerations apply with respect to the faculty under whose supervision an applicant wishes to do doctoral research work. How many of this person’s past advisees are professionally at the place where the applicant wishes to see him/herself down the road ?
Departments, programs and individual researchers with a good track record of advising and placement, will be willing and eager to share these data with potential applicants; normally a web page or a program brochure will spell them out, but there is only so much one can learn in this way. Talking to past students is a far better way to get to the bottom line. If these people occupy positions of prominence in their fields, in academia but also in other venues, it will be generally easy to locate them. If they are altogether happy with their graduate experience, they will be glad to share their opinions with a prospective student. If, on the other hand, they have dropped out of sight, that is a bad sign in and of itself.
Academic strength of a program (size does matter). A large graduate program with a good size faculty (for a physics graduate program, ideally at least 50, most of whom active in research ) and many graduate students is a more vibrant and intellectually stimulating place than one with few faculty and even fewer students, and will make for a more enjoyable, rewarding and fruitful graduate experience altogether. In particular, I have personally come to regard interaction with fellow graduate students as a crucial part of the educational process at the graduate level. A graduate student spending most of the time alone when not talking to the direct supervisor, will not only get bored much more, but almost certainly will also end up learning less than one who has the opportunity of discussing and exchanging ideas with other graduate students, both engaged in the same field of research but also in other areas.
The size of the faculty is important for a number of reasons. Firstly, different research areas will be adequately represented, offering a broad spectrum of choices to those students initially undecided about their research focus, but also allowing (through colloquia and seminars) those who know from the beginning where their research interest lies, to keep abreast of developments in other fields (or even to change their mind…). Secondly, in a large department a student doing research work in one particular field will not have just his/her own graduate advisor as a reference, but will be able to talk to other faculty in the same area, as well as postdoctoral associates; this will be beneficial not only intellectually, but also in practical terms, as these people will serve as professional references for the student, later on. Finally, in the sciences the presence of a critical mass of faculty engaged in similar research is typically instrumental to bring in the funding required to establish initiatives such as focused research centers (e.g., in materials science), whose presence will normally provide additional sources of support for students.
Ranking of the graduate program. Yes, I am talking about rankings such as those published annually by US News and World Report. Although they are often labeled as biased, bogus and unfair (the charges may have some validity), dismissing them altogether, or ignoring them, is silly. Simply put, these rankings reflect the scholarly hierarchy in a specific field, as perceived by those who are typically in charge of hiring decisions. Whether that perception is accurate or not (and we can debate that ad infinitum — I happen to think that it generally is) is ultimately really immaterial to the future of a PhD graduate . In physics, for example (as noted by Incoherent Ponderer), PhD graduates at the top 30 institutions account for almost 90% of faculty hires in the US; I doubt very much if the same is not true in other disciplines. A prospective student, especially one with academic career ambitions, should be at least aware of this state of affairs. That is not to say that one cannot receive an excellent graduate education at an institution whose program is not prominently ranked, and still enjoy a reasonably successful career later on. However, that scenario is less likely, as pedigree does matter. This is why, everything else being equal, I have always recommended my students applying to graduate school to give preference to departments enjoying a more favorable ranking; only in exceptional circumstances would I recommend anyone to apply to a program that is not consistently ranked within, say, the first 40 nationwide, and if at all possible I think it is better to try to get at least into one of the top 20.
 See, for instance, here, or here.
 The definition of a “research-active” faculty may be debatable. My personal observation is that research activity correlates quite well (not perfectly) with supervision of graduate students. So, in a department in which most faculty are research active, only a fraction thereof will not be supervising any graduate student at any given time.
 In fact, rankings based on “reputation” only, often the target of the most vitriolic criticism, may well be the most relevant of all, from the viewpoint of a prospective applicant.