The comments to my previous post, which addressed the statement of purpose for grad school and miscellaneous admissions criteria in several disciplines, are detailed and very interesting, and I recommend that you go read them all.

Here, I want to answer (in my trademark meandering way; sorry folks!) lyra211’s question from her comment: “Xyk, I’m interested in the part of the question you didn’t answer — namely, would a student like this student (good grades, a couple REUs, and undergrad thesis, but no first-author publications) be a competitive applicant?” (Also note Erica’s response to the same question.)

I advise graduate students from three physical-science STEM departments: one a basic science, two applied. My school is a large public research university, and most physical science STEM departments are ranked in the mid-teens, with some among top 10. So, nothing to sneeze at, but not as fancy a place as Erica’s, for example.

This means we get plenty of good applicants, but often cannot recruit the very obvious cream of the crop, especially from among the domestic pool (which, alas, also means that few of the NSF GRFP and similar fellowships go to us, but that’s a story for another day).

I find it that I can recruit top-notch international applicants, good as anyone in their generation, because they often don’t look as good on paper as similarly able and usually highly pedigreed domestic applicants. International students often don’t have super high GRE General scores (except GRE Quantitative, and once upon a time GRE Analytic before it was replaced with the useless analytic writing) and they don’t have super highly polished statement of purpose.

Here’s an example. I recruited a student a few years ago from a country with a very strong national selection program before college entrance. This student was in a national gifted program, then at the nation’s top university, had great GRE Quant and GRE Subject test scores, and was still going to be rejected because their analytic writing was below some stupid university cutoff. In that sense, it may be good we are dispensing with test scores, because some careless bureaucrat won’t have the ability to auto-reject a great student. This student had research experience and extremely strong, eloquent support from their undergraduate research advisor. I decided to bring them on because of my good past experiences with students from that school and the ringing endorsement from research advisor. This took some doing because the student had already been rejected, but I made some noise, we got them “unrejected,” and I was not disappointed. This student is probably the most talented one I’ve ever had and did some amazing work.

Alas, relying on past experiences with a certain institution can also backfire. Example: My first graduate student was phenomenal, joined my group after their original one didn’t pan out. Happy with their performance, I recruited from the same school a person who looked identical on paper (similar GPA, letter writers, the works). That new student was a dud in multiple ways. No drive, no initiative, completely inactive in the absence of very specific, finely granulated instructions, doing only the bare minimum. I am now very reluctant to recruit from that particular milieu again, because I now know I can’t trust the letter writers and I can’t trust the grades to have predictive value as far as research excellence is concerned.

Overall, I wish we had kept the scores. They are not everything, but they are something. I used to look at GRE Quantitative (I expect maximum score from international applicants, a bit lower is OK for domestic). Very high verbal scores from international applicants seldom indicate actual facility with the language, and I ignore it. I do, however, look at TOEFL scores, which are much better predictors of language skills. Not all departments I recruit from require GRE Physics, but if they do, or if the students provided it anyway, I expect more than a decent score from international applicants and decent from domestic. I have never met a student with a low (I mean really low) GRE Physics score who didn’t have serious gaps in their knowledge. (Ask me how I know. Ugh.)

---

Now, how have I recruited thus far and what will I do in the future? I recruit A LOT of international applicants because my chances of getting a top international candidate are higher than for a similar domestic applicant. The fraction of domestic applicants who are good enough (in terms of math and physics) for what I need them to do AND who do not get snapped up by fancier places is not high. So I focus my recruiting efforts heavily on great international students. I also keep an eye out for excellent hidden domestic gems, students who show great potential but don’t look like superstars on paper, maybe because their GPA isn’t perfect or they went to a small school or similar. This has been hit or miss, I must admit.

International applicants: There are several countries (and a couple of schools in each of these countries) from which I have recruited in the past and every student has been awesome. I am happy to recruit from these countries and schools again because I feel I now understand the scales/meaning of their grades, I trust the national weed-out process, and I have other students in the group who can help me read between the lines in the application. Basically, I have a few pipelines and plan to keep using them. Strong letters of recommendation and research experience are always a plus. One strong letter from a research advisor who knows a student well and can really attest to their abilities and personality goes a long, LONG way. And, of course, there is a lot you can tell about a person after you’ve videoconferenced with them.

There are other countries from which I recruited in the past and it has been hit or miss. In times of limited funds, plentiful applicants, and no tests, I will give high priority to those coming from the school systems and universities I know and trust.

Domestic applicants: Things are trickier here, at least for me. The best of the best who are also clued into how academia works (so also went to top schools) do not come here. An additional issue is that I apparently cannot charm the domestic applicants who come to open house; they probably run away from my accent if they don’t from my research area, and are more likely to go to American-born advisors; this was such a pronounced effect that for many years I didn’t partake in the open house because I could never recruit anyone from there, so why waste time. The last couple of years I did partake but didn’t make any offers because the students who were domestic and interested in my field were too weak; they came from PUIs but missed some important coursework or their grades in the courses I looked at were not as high as I like them to be, and the GRE Subject scores were way too low.

Among my current and graduated domestic students, some of the great ones came from: a fancy private SLAC (looked great on paper, I got this student by dumb luck of their previous advisor not working out); a public non-elite PUI (didn’t look that great on paper except for phenomenal research letters; also by luck as they moved to my group from another one; student missed lots of coursework from undergrad, but super talented and intelligent, so ramping up very quickly), and a public R2 (looked decent but not flashy on paper, was great at research).

My experience with non-elite PUIs, especially very small campuses, is that students often miss certain fundamental coursework because their faculty are small and the department might not offer a whole roster of courses that someone in a similar major at a larger school might have taken. This is a real phenomenon, I have seen it half a dozen times at least. This also makes the students score less on GRE Subject (because there is material they never covered) and the  students might have issues when starting grad school because they first need to catch up. One of my weakest domestic students was from a small public PUI, and they had pervasive gaps in preparation that never really closed. 

I am not against PUIs by any stretch, and there have been some great applicants with plenty of research experience from various PUIs. However, it is true that not all PUIs are made equal in terms of the curriculum they offer (R1s and R2s tend to be much more uniform in their course offerings, perhaps because they have larger faculty rosters). So I look at students’ transcripts in detail to make sure they had the needed coursework and did well in key courses. 

---

I have one excellent undergraduate who will have been with me for 2+ years by the time they go to grad school next year. They have a near-perfect GPA and will have a couple of papers and a recommendation letter from me that is stronger than an enraged Bruce Banner. What they will also have is a polished statement of purpose (read and commented on and edited by a group mate and by me) and two hours of me sitting with them, us looking at faculty rosters of several of their top choice schools, deciding on 3-4 names of faculty well aligned with their interests, whom they will list on their statement of purpose and contact individually. You do not get this sort of help after a passing engagement with a research group. But if you are all in, what you put in will pay this type of dividend.

Applicants, I cannot stress this enough: Contact the professors who do what you are interested in via email. Write a brief cover letter (that’s the body of your email) and attach your CV, test scores (if applicable), unofficial transcript, a statement of purpose if you have it ready, and any papers you’ve published (definitely not necessary, but if you’ve got them, flaunt them). Put all of this in one or at most 2-3 PDFs (do not send anything that needs to be unpacked; no one will unpack random zip or rar files from unknown senders) and attach them to your email. I always look at the cover letter (there are clues in the cover letter that the person is carpet-bombing faculty or otherwise isn’t suited for my group), but if that initial pass looks good, I will look at the rest. I look at grades in math and physics courses. I look for indicators that this person can do research, such as info about it in their personal statement. If they’ve completed the application, I will go look at their letters; if the application is not in yet, I will tell the applicant to let me know once the application is complete so I can go look. Applicants, you need to have at least one letter from someone who knows you really well, ideally after substantive research.

Letters aside, a serious research experience before the PhD is paramount because it will show the student whether research is something they would actually enjoy doing for several years. Doing well in coursework is weakly correlated with enjoyment of research. Research requires a very different mindset, a confidence in one’s own logic, an ability to work on loosely framed long-term projects, being self-propelled and able to withstand setbacks without giving up.

---

Before I forget: In some departments (one of the three I am affiliated with) I don’t get to see any applicants that the grad recruitment committee hasn’t approved; students are all admitted to the department, on department’s nickel, and only join research groups after the first year if they pass the qualifying exam; that’s when PhD advisors pick up the research-assistantship slack, or not, it’s not a big deal, because TA-ships are always abundant. In the other two departments I am affiliated with, students are admitted directly to a research group, the advisor is on the hook for funding from day one until the end of time, and TA-ships exist but using them is frowned upon.

---

So, as to lyra211’s question, what I think is important, especially without tests:

• Good grades, and especially good grades in select courses relative to the subspecialty go a long way as far as I am concerned.
• Research experience, not just pro forma which so many students engage in (perhaps leftover from the whole collecting-random-extracurriculars mindset of applying to college, which is also misguided) but the real kind, where a student pursues a genuine interest and works on something for multiple semesters and/or summers. (Some REU experiences suffice here, others don’t. It depends on the student and their interest and match with the REU.)
• At least one strong (strong = positive but also very student-specific and detailed) letter of recommendation. (Again, some REU experiences produce great experiences and strong letters, others don’t. Depends on the student and their match with the REU.)
• A well-articulated interest in a subspecialty (what they want to pursue and why) and targeting places with a number of faculty in that specific subspecialty.
• Direct contact with prospective advisors via email.

I recently had a videochat with a former undergrad who’d been out of school for a couple of years working, and is thinking of going back to grad school. While on the surface he sounded like he knew what he wanted, was energetic and confident, and he had actually gone to school here, I probed a little bit and he could not really tell me who the people were here who really did what he wanted to do. So I told him to do some research and figure out with whom he’d like to work in an ideal world, and then we would talk again.

Applicants, you have to know what you want, otherwise how will you know what to ask for? Maybe you didn’t know what you wanted to do when you applied to college. Grad school (here I primarily mean getting a research degree like a PhD; professional schools and professional MS degrees are different) is a time of specialization and learning to do research professionally, which means at a high-enough level that someone will pay you to do it. A PhD is not for the undecided. Ask yourself what you are really interested in. It cannot be everything. When I ask prospective students if they are interested in theory or experiment, many say both. I tell them that, if there’s a chance they could be happy in a lab, they should go to a lab and do experiments, because the options after an experimental PhD are more numerous. I need people who are strongly drawn to projects with a lot of math, who get excited by mathematical models of the world.

Prospective students, deep down, you know what you want to do. It doesn’t have to be super specific, but you should have a subarea well defined enough so that you can find 3-5 (or more) faculty well matched to your interests in each of the places you apply to.

And don’t all apply just to big-name coastal schools. They can’t take you all, and there are great universities in Flyover Country where you can have an excellent PhD experience.

---

Phew, that was a long one!
Blogosphere, what did I forget to talk about? Thoughts, shouts, murmurs?