Your humble bloggers here at Cosmic Variance have spent quite a bit of accumulated time in academic and research settings — in fact, my guess is that none of us have spent an entire year away from such a setting since the age of about six or so. That’s a lot of accumulated wisdom right there, and it’s about time we started sharing it. Since it’s that time of year when applications are being sent off to graduate schools, I thought I would start off by letting everyone in on the secret to how to get accepted everywhere you apply. Of course I can only speak for physics/astronomy departments, but the basic lessons should be widely applicable. [Update: see also Choosing a Graduate School and How to Be a Good Graduate Student.]
So, here goes: have great grades, perfect GRE scores, significant research experience, and off-scale letters of recommendation. Any questions?
If, perhaps, it’s a bit too late to put that plan into action, here are some personal answers to questions that come up during the process. Co-bloggers (and anyone else) are free to chime in with their own take on these complicated issues. Keep in mind that every person is different, as is every grad school — in fact, specific schools might behave quite differently from year to year as different people serve on the admissions committee. Don’t sink your sense of self-worth into how you do on these applications; there’s a strong random component in the decisions, and there are a very large number of good schools where you can have a fun and successful graduate career.
- What do graduate school admissions committees look at?
Everything. Keep in mind that, unlike being admitted to college (undergrad), at the grad school level the admissions are done by individual departments, with committees comprised of faculty members with different kinds of expertise, and often students as well. They’ll look at your whole application, and in my experience they really take the responsibility seriously, poring over a huge number of applications to make some hard decisions. Still, it’s well-known that careful examination of a thick file of papers is no substitute for five minutes of talking to someone, which schools usually don’t have the luxury of doing, so decisions are always somewhat fickle. - Even my personal essay?
Well, okay. I wouldn’t sweat the personal essay; in my experience it doesn’t have too much impact. Let’s put it this way: an incredibly good essay could help you, but a bad essay won’t do too much harm (unless it’s really bad). To a good approximation, all these essays sound alike after a while; it’s quite difficult to be original and inspiring in that format. - Are GRE scores important?
Yes. At least, in the following sense: while bad GRE’s won’t kill your chances, good GRE’s make it much easier to admit you. (We’re speaking of the Physics GRE, of course; the general tests are completely irrelevant.) It stands to reason: given two applicants from similar schools with similar grades and interests, there’s no reason for a department to choose the student with lower GRE scores. At the same time, you can certainly overcome sub-par GRE’s by being outstanding in other areas; this is particularly true for students who want to do experiment. I know at Chicago that we let in students with quite a range of scores. - What about research experience?
Research can be a big help, although it’s by no means absolutely necessary. These days it seems that more and more undergrads are doing research, to the point where it begins to look unusual when people haven’t done any. There is some danger that people think you must want to keep on doing the kind of research that you did as an undergrad, although I wouldn’t worry too much about that. Mostly it shows some initiative and passion for the field. It can be very difficult to do theoretical research as an undergrad, but that’s okay; even if you eventually want to be a string theorist, it’s still great experience to do some experimental work as an undergrad (in fact, perhaps it’s especially useful). - How do I get good letters of recommendation?
It’s more important to have letters from people who know you well than from people who are well-known themselves. One of the best side benefits of doing research is that you can get your supervisor (who hopefully has interacted with you quite a bit) to write letters for you. It’s really hard to write a good letter for a student who you only know because they took one class from you a year or two ago. Over the course of your undergrad career, you should find some way to strike up a personal relationship with one or more faculty members, if only to sit in their office now and then and ask some physics questions. Then they can write a much more personal and effective letter. Of course, if you are just a bad person who annoys everyone, it would be just as well to stay hidden. (Kidding!) - Is it true that the standards are different for theorists and experimenters?
Typically, yes, although it might be different from place to place. Because a lot of undergrads haven’t been exposed to a wide range of physics research, a large number of them want to be Richard Feynman or Stephen Hawking or Ed Witten. Which is great, since we need more people like that. But even more, we need really good experimenters. Generally the ratio of applicants to available slots is appreciably larger for theorists than for experimenters, and schools do take this into account. Also, of course, the standards are a little different: GRE’s count more for prospective theorists, and research experience counts more for prospective experimenters. And let’s be honest: many schools will accept more prospective theorists than they can possible find advisors for, in the hopes of steering them into experiment once they arrive. - So should I claim to be interested in experiment, even if I’m not?
No. Think about it: given that schools already tend to accept more students who want to do theory than they can take care of, what are your chances of getting a good advisor if you sneak into a department under false pretenses and have to compete with others who came in with better preparation? It makes much more sense to go someplace where they really want you for who you are, and work hard to flourish once you get there. - Do I need to know exactly what I will specialize in?
Not really, although in certain circumstances it can help. Professors like to know that someone is interested in their own area of research, and might push a little harder to accept someone whose interest overlaps with their work; on the other hand, most people understand that you don’t know everything after three and a half years of being an undergraduate, and it can take time to choose a specialization. In particular, at most American physics departments (unlike other countries and some other disciplines), it is generally not expected that you need to know ahead of time who your advisor will be when you arrive, or which “group” you will work in. - Should I contact faculty members individually if I’m interested in their research?
That depends, mostly on whether the person you are contemplating contacting is desperate for more grad students, or is overwhelmed with too many requests as it is. In popular areas (ahem, like theoretical particle physics, string theory, and cosmology), there are generally more applicants than departments have advisors for. In that case, most people who receive random emails from undergraduates will just urge them to wait for the admissions process to take its course; remember that it’s a zero-sum game, and for everyone who gets in there’s someone else who doesn’t, and it would be a little unfair to penalize those applicants who didn’t contact faculty members personally. On the other hand, if you have reason to believe that someone you’re interested in working with is trying to get more students, or if you think your case is somehow unique and requires a bit of attention, feel free to email the appropriate faculty member with a polite inquiry. The worst that can happen is that you get a brush-off; I can’t imagine it would actually hurt your chances. - Is my life over if I don’t get into my top grad school?
Yes. Well, only if you let it be. The truth is, how you do in grad school and beyond (including how you do on the postdoc and faculty job market) depends much more on you than it does on where you go to school. In the next episode of “unsolicited advice,” we’ll think about how to actually choose where to go, including how to get the most out of visiting different schools.
Actually this episode was not completely unsolicited; thanks to Philip Tanedo for suggesting we share some of our invaluable insight. See, sometimes we really do listen.
I give up. I’ll say it one more time clearly and then I’m done. In my experience of faculty hiring in theoretical physics at a research university, which is quite a lot recently, Ph.D. institution is essentially irrelevant. Who you are, what you’ve done and what your reputation is are the key points. This is actual data. Take it or leave it.
To be honest, I think this is nonsense.
Firstly, it is simply untrue that a CV gets looks at for 30 seconds before being tossed. I have served on several faculty search committees, each of which dealt with > 100 applicants. In each case, the committee divided the list so that each application was read closely by two or three people, and the whole committee went through the whole list. You are undoubtedly right that a PhD from Harvard looks good but it makes epsilon’s worth of difference in the long run — at the faculty level, the key issue is what the person has produced, and what their letters say about them (and the letters are more important if you work in a big collaboration, since we will want to know what *you* did and a good letter will explain your own contribution to some large project.) Hiring at the faculty level is the single most important thing an academic department does, and a surprising amount of work goes into identifying and recruiting candidates.
Secondly, the initial cut is to draw up a long list, and that will often have 20 or 30 names. If it is not self-evident you should be on that list, then your chances of making the short-list are very slim indeed.
Thirdly, as to the case you cite of the person who graduated with a couple of papers and then had an unproductive post-doc at the more famous place: that person is in deep pickle. And it is not because they don’t have the Harvard/Stanford/Princeton seal of approval on their diploma, it is because their publication list does nothing to convince a search committee that they will be able to set up an independent research program. Moreover, reading between the lines of what you said, the new job involved mastering cutting edge techniques (presumably) they not picked up at their not-quite Tier 1 institution, which suggests that the students graduating from the Tier 1 place are in better shape when they go on the job market. This may be unfair, but it is an excellent example of the feedback mechanism which ensures that people graduating from top-drawer places are more likely to have done cutting edge work than their colleagues at Tier 2 places, independently of any differences in their raw talent.
Over the years I’ve noticed that people who did not get an assistant professor job after 2 or 3 postdocs, are quick to say that it’s only people with PhDs from tier 1 universities (ie. Harvard, Princeton, etc …) who get all the jobs. They essentially blame all their “failures” on the fact that they couldn’t get into Harvard or Princeton for grad school, when there’s other obvious reasons for their failure to get an assistant professor job.
On a closer examination, frequently the main reason for these folks’ failures was that they did not publish many papers other people were interested in. They don’t want to admit that hardly anybody (or nobody for that matter) was interested in the papers they published. So instead of blaming themselves for their failures, it’s always easier to blame it on the pedigree of the grad school they went to.
It’s human nature to not want to take responsibility for one’s actions and decisions, and instead blame it on something and/or somebody else outside of one’s self.
Can’t speak for my co-bloggers, but I’m clearly only 29! In fact, Hektor Bim of #58 guessed wrong on every single detail describing me….
In ideal world everyone would be evaluated on their science instead of where they got their degrees.
In real world this becomes increasingly difficult with so much science done.
I still think publication record (quality and quantity) is the best criterium.
Recommendation letters underwent so much inflation it’s hard to say whether the letter is really really positive or simply “really positive”, which I guess stands for negative. When was the last time someone got a neutral or negative letter?
Publication (and citation) record however is very field-specific. All things equal I would rather prefer to hire a postdoc with Harvard PhD than Kansas State (no offense to Kansas people) PhD. The difference is substantial to overcome some difference in terms of publication record, even though only to some degree.
I think there’s some truth to #78 – people like to blame it on outside circumstances, but people with Top 10 school degrees (and I am one of them) – you can’t deny it helps you get your foot in the door.
They did blind studies where professors were shown fake identical CVs with either female or male first name and when asked to guess the level of professional position (assistant professor, postdoc, etc.) – people put males higher than females. I am pretty sure if the same study was asked to evaluate identical publication records with different PhD granting institution, people would rank Tier 1 school CVs much higher than, say, Tier 3 school.
I’d like to believe committees were completely blind to that, but they are not. Just like some “hot” fields of science get preference over lesser known fields – even though science may be of much higher standard.
I agree 100%, Ponderer (#80). In fact, this seems so self-evident to me that PhD school is not a neutral factor that I have a hard time grasping the objections. Maybe the only way to reconcile Mark’s “actual data” (#76) and my own experience, which is also “actual data”, is to conclude that it depends on what institution is doing the hiring.
regarding (#81)
another way to reconcile Marks “actual data” could involve one of you looking at the world through “rose colored glasses”
ksh95, it’s either “rose colored” or “anti-rose colored” (“purple-colored”?) glasses, depending on your perspective.
Mark, I realize you are frustrated with people contradicting your experience at Syracuse, but looking at rumor mill pages it does seem there’s a correlation with degree institution – offers to people from top places with relatively modest research CV are quite commonplace. I think this is because often decisions are made based on potential, rather than past record – and how do you measure potential?
I am not saying that faculty search committees sort applications into “Tier 1” and “Tier 2+” piles, or that they would bring up this issue out loud – like most stereotypes those could be quite more subtle. I think it’s great when department tries to isolate good science regardless of where it’s coming from, but how would we ever know that it’s not happening on subconcious level?
Three new faculty hired at Syracuse this year are all theorists and hold PhD’s from Harvard, Berkeley and Colorado Boulder. All three are clearly outstanding schools, but perhaps it’s self-selection like Sean argues? Who knows…
Hi Ponderer. I guess it seems like self selection to me because I never see any evidence of any other way this comes into the discussion. I was happy to refer to experience I’ve had, but obviously won’t get into any details of the actual searches at my school.
Thanks for your comments.
Fair enough, JoAnne. Sorry about that.
JC is mostly right. In most cases, the faults are in ourselves. People work in the wrong subfield, get stuck in a project that is going nowhere, fail to get enough PRLs, annoy their advisor unnecessarily, etc. Most of these things can be foreseen and avoided with iron discipline and great focus and persistence.
But this happens to people at Harvard and Princeton (and other Tier 1 schools) too. It’s not like getting into a top ten grad program guarantees you an academic job. Far from it. Even in the top ten schools, your chances are still 50% or less, sometimes much less, depending on your advisor and subfield. (This is based on “actual data” by the way.)
The problem is that is isn’t much fun to be engaged in an activity where the slightest misstep takes you out of contention. There isn’t much of a margin for error. Make a mistake, and the wave takes you away.
That’s why people should think a lot about grad school and have some serious strategies for where they will go and what they will do. The price of mistakes is high, if you want a research position.
Sean, Mark, JoAnne & all others engaged in this conversation:
I’m finding these observations fascinating because I’m at the stage of planning for grad school apps. The tangent involving postdocs & faculty appointments is still interesting from my self oriented point of view in part because of the recurring, but unspecified, references to the top ten grad programs. I am not in a research environment & so have no personal data on which to base such a list. Would any of you care to offer your personal top ten? I realize that no department is uniformly strong in all disciplines, so let me suggest crafting a list according to your personal criteria. I’ll offer my own list, but please be tolerant of my limitations. The list is not objectively determined because I am inexperienced in this field (so much so that I can think of only eight). So, here it is in approximate order (I don’t have confidence to include universities outside the U.S.):
1. Cal Tech
2. MIT
3. Princeton
4. Stanford
5. Harvard
6. Berkeley
7. Chicago
8. U of Illinois
Chicago is unique – the only one not associated with an engineering school (as far as I know). Probably impossible to move very far up the list without this potential synergy.
Emile — are you interested in condensed matter, or astrophysics, or particle physics, or what? Also, experiment or theory? The answer to which schools are best is strongly dependent on that. If you’re not sure, it’s probably best to focus on schools with very large departments with many strengths (e.g. Berkeley), so that you can be happy there no matter what you decide to specialize in. If you think synergy with engineering is relevant, it sounds like you’re probably interested in more applied fields….
I don’t really think any fine-grained ranking is particularly useful, especially given how things can vary by subfield. All the schools you list are quite good, however. You could probably also include UCSB and Cornell, off the top of my head.
Anon-
Theory is more appealing & my current interests are General Relativity & Cosmology with particle physics and condensed matter as runner-ups.
Most of the schools you name are fairly strong in cosmology, as are UCSB and Cornell, which Aaron mentioned. On the other hand, they are not all so good at particle physics. Princeton and Chicago, for instance, have a number of very good string theorists (and have good particle experimentalists) but are not so strong at particle theory. (If by “particle physics” you include string theory, then all of the schools you name are strong.) I can’t really comment on condensed matter. There are some other quite good state schools you might want to consider as well (Washington, Wisconsin, Michigan come to mind, there are probably others). I suggest you either try to narrow your focus and look into the work of specific professors you might be interested in, or if you don’t want to decide on a specialization, try to stick with larger departments where you have flexibility.
all of those schools are very good. Urbana, Cornell and Harvard are particularly good in condensed matter – both experimental and theoretical.
Also keep in mind that Berkeley and MIT have huge departments, while others (Caltech, Harvard, Princeton) are relatively small. This cuts both ways, obviously.
Hmm, my top 8 had Michigan instead of Illinois, although I might have Illinois in the top dozen.
It’s easy to come up with the top 7 or 8 … not so easy to do the next 10.
Weren’t we promised a future blog entry on how to choose a grad school?
In case anyone is still reading this far, and is worrying about the whether instead of the how, a Brian Weatherson discussed grad school at Crooked Timber, and I have some thoughts on his post here. Weatherson was talking about philosophy, and I’m a biochemist, but from reading Sean’s advice and the comments, it seems like the generalizations apply. My bottom line:
The question of how the school you go to affects your long-term job prospects is a complex multivariate problem. But I think that Mark is absolutely right that PhD institution per se is significant at the search committee stage. If I’m looking for a quick triage, it’s based on pubs. Pedigree influences perception of quality of those pubs, but for that, postdoc gets weighted more than PhD.
As an outsider (I studied materials science), I have always been intrigued by the lengthy post-doc dance that physics (and may be biology too) seems to require before landing a faculty position. Many people have already written about it here.
In most engineering subjects, on the other hand, one post-doc is the norm. If one has worked in ‘hot areas’ even that is not necessary for landing a faculty position. Quite a few of my contemporaries became assistant professors immediately after their Ph.D.
I have always wondered what made physics choose the culture of extended postdocs.
Any insights?
Abi (#94)
As I understand it, this practice started sometime in the early era of the development of quantum physics (1920s). The principle driving force (my conjecture) was due to the fact that physics graduate level training was in an upheaval caused by all the new physics – not only QM but also relativity (as a consequence of which Maxwell theory became a necessary part of the curriculum) as well as statistical mecahnics. So there was a conversion from the late 19th century course of study to the new 20th century curriculum, coupled with an explosion of research into many new areas which are commonplace today. I believe it became clear that PhD training was insufficient for mastery of the field in general, and much less so for mastery of one’s specialization. Private institutions funded a network of postdoctoral studies opportunities throughout Western Europe. The concept grew from there and has taken hold in the US with the added impetus of the last few decades when PhD supply has been far in excess of faculty-position demand, causing a need for a holding pattern (again, my conjecture). No doubt there is also the advantage accruing to search committees of seeing evidence of an applicant’s ability to initiate and sustain a viable research program, which would tend to enhance acceptance of this practice.
I think you are correct – in fact among friends physicists who recently got hired as junior faculty, disproportionate number did so by joining EE or Mat Sci departments, rather than Physics – even though their degrees were in physics with a little sprinkle of engineering. Just heard today from a friend who has joined a faculty in biomedical engineering at one of the top schools, despite physics training with no biomed background – he was doing some ultrafast laser physics that has only tangential relevance to biomedical problems.
I think part of the reason is that those departments are undergoing tremendous expansions at many schools. It’s probably somewhat easier to get funded working on more applied problems. Let’s face it – we have been facing overproduction of Physics PhDs for many decades now.
Pingback: Cycle Quark » Getting into Graduate School in Physics
Hey, how about some advice on how to choose which grad school to attend once you’ve been accepted? It’s getting to the time when we start to hear back from schools!
This assume we get accepted to more than one, of course 🙂
Pingback: Often in Error...
Pingback: Unsolicited advice, Part Deux: Choosing a grad school | Cosmic Variance