A prospective grad student (PGS) reader asked me to post this query, which is timely for many applicants. I will also share some of the brief comments I responded with, but I’d like to invite the academic blogosphere to chime in with their wisdom. When you comment, please state what field you’re in (broad is fine), because a lot of advice is field-specific.
————-
Dear Prof. Phlox,
If you don’t mind, I have a blog request. I wanted to hear your opinion, if possible your advice, on how to write a Statement of Purpose (SoP) for grad school applications in STEM (particularly physics). This in the larger interest of the academic community of students writing the statement and faculty having to read them.
There is a lot of contradicting advice out there and it has been quite an unnerving task to separate the actual information from the noise. While it is clear that it is expected of us to discuss our future career/research goals in the light of work we have done so far, some aspects like the right amount of details remain very vague.
I am given to understand that with no test scores in the application package, the importance of this single piece of document could go up exponentially in the application package. What do the faculty expect from an average student sans any major achievements? By an average student I mean, someone with good grades(A/A-) in required courses, a couple of REU stints to know how the academic research works and may be an undergrad thesis. In my case, being from small private college I have only had 2-3 months at a time while working on summer internships, this meant I couldn’t get to publish any of the work I did. Would I still be considered a competitive applicant?
Best,
Prospective Grad Student
——-
Hi PGS,
In my opinion, the SOP is supposed to convey something about your personality that the letters of recommendation presumably cannot. What you find interesting and why, how you like to work. Basically, this is an opportunity to talk about what you liked in school and why, what drew you to certain topics, what made you decide to pursue graduate school, what you enjoyed when you worked on research. Anything that makes you stand out as an individual with respect to others with a similar background is also beneficial. The SOP is a professional document but it is one that should help convince members of the graduate recruitment committee/prospective advisor “This kid sounds like they would fit right here / in my group.”
Best of luck!
Xykademiqz
————-
A few follow-up questions:
PGS: In general, is it important to convey why one is attracted to science?
X: Yes, if you can articulate it succinctly and without resorting to cliches.
PGS: For example, I want to work in [theoretical physics subfield]. I have a couple of reasons: (1) I like working with calculations, both analytical and numerical as they are my strengths. (2) This could sound naive but I like to do theory over experiment as it gives an opportunity to explore research ideas on my own, to a reasonable extent in my academic career. The bottom line I just see myself in academia because of my curious nature and willingness to dedicate a significant part of my life to the science that matters to me.
X: The way you conveyed it above is fine. It’s clear and to the point; that’s always ideal. Don’t be too colloquial or rude, but being informal is not bad in the US, and it’s also OK to be confident. This is a personal statement, so your personality should come across.
PGS: Do faculty like to hear an account of previous research even if it is not directly related to my intended direction of research?
X: Absolutely. Especially the nontechnical parts. Did you like working alone/with others? How you were advised (hands on/off; lots or few meetings; lots of freedom or frequent checkins?) especially if you liked the style. What was something you didn’t expect/that surprised you? What was something cool you learned?
PGS: What are some of the things faculty hate to see in these statements?(Apart from cliched childhood science fantasies)
X: Yes, cliched openings like childhood science fantasies are definite eyeroll inducers. I also hate anything that is obviously meant to pull at heartstrings or showcase the applicant as some sort of otherworldly creature with no thoughts, feelings, or experiences beyond science; anything that smells of pretense or manipulation; anything that reads like it was written by a 50-y.o. professional editor (because it was). Honesty, clarity, and specificity go a long way.
PGS: Should students discuss some of the ongoing research in the labs that they’d like to work?
X: Absolutely, but not too specifically because published papers are generally 2-3 years behind what is currently being done. Focus on finding 2-3 professors whose general area of research interests you and mention them by name in the SOP (very important!). You should also email them and in those emails you can indeed mention some specific papers from their group (ideally you’ve tried to read them). But in the SOP their general area of research should naturally flow out of your described interests; for example, you want to work in [theory subfield], spent the whole SOP talking about it and maybe specifically mentioned some subtopics, then obviously people working on those topics should be listed in the SOP.
—————
Academic blogosphere, do your thing! Let’s help new applicants craft their best SOPs this admission cycle! When you comment, please state what field you’re in (broad is fine), because a lot of advice is field-specific.
xykademiqz 
There is not one right answer. This is going to depend very much on the discipline (certain fields have certain and very different conventions) and sometimes the specific department. But, as former director of my Ph.D. program, my biggest advice is, “do you due diligence, find out what the department/university expects- it’s probably clearly stated- and follow that.”
For example, in my department and in my field, you’re expected to have a specific thesis project and advisor in mind from the get-go, and for a realistic shot at admission it should be stated in your letter (“I plan to work on the Whatsaexperiment with Professor X, who I’ve discussed my potential thesis topic with, and she has stated she’d like to support me on her Department of Energy grant”), and the applicant should have been in contact with the prospective advisor n advance, and at least have some agreement at the match-up and project. A letter expounding on the wonder of physics as the foundation of the universe and no indication of having had any contact yet with any professors, much less any indication of what specific subfield or topic or project the student wants to work on, will not get the applicant far- although we get lots of applicant statements like that. Likewise, a statement saying “I plan to work on Whatsatheory with Dr. B.” without Dr. B having had any commmunications or discussions with the applicant- is a red flag. This is very well spelled out on our department graduate program’s web page under a section called “prospective applicants”- yet many applicants seem to have never read it. Likewise, we get applications from students wanting to work with professors who have long since retired or gone elsewhere… ???
But other universities, departments, and/or STEM fields operate very differently! Some have all grad students do “rotations” through different research groups or labs in their first year and basically prohibit starting out with a specific advisor and project till the first year rotation is complete. It’s generally explained in the department’s or program’s graduate catalog or web page, so, do your advance research!!! (Sometimes it’s different within the same department: some biology departments have their biomedical/cell biology grad students all spend the first year doing lab rotations, while their ecology grad students are expected to start out in a specific lab with a specific project, decided in advance by consultaton with advisor before acceptance, from day one!)
Thus, my biggest advice is, do your advance research on a particular department- they will often rather explicitly state what they’re looking for in applications- and in most STEM fields, nowadays, find a professor you want to work with as at least a prospective advisor, be in touch with them well in advance, and get their advice on how the process works at that university or program!
Econ:
Don’t start with “Since X, I have always wanted to…” Don’t be cutesy in your intro– this is not an undergraduate college admissions essay. You don’t get extra points for metaphors. Instead make it clear in general the direction you want to study (what kind of research are you interested in) and why (it’s fine to have a personal touch here but also fine not to– you can be interested because someone you love died of X or you can be interested because it’s a fascinating topic), and also demonstrate that you understand research design (causation), that you’ve had research experience, and you have some ideas about what kind of projects you’d like to do in the future even if they’re not exactly feasible right now.
In econ we don’t contact current professors directly… there’s some old boy’s network in which a professor who likes you might put in a good word for you with a friend at a grad program, but it’s not necessary. We don’t really have labs, and the prof you want to work with might be on their way to a completely different school at any point in time since there’s so much movement among the top people. (What is becoming increasingly necessary to get into a top 5 or even top 10 program is a research-heavy pre-doc.)
I’m in physics (and I do theory). My department does not usually accept students who have only done REUs, but we have accepted such candidates occasionally. We certainly regularly accept people without publications, and this is not seen as a negative.
A note: I would guess that (at least in my department) the lack of test scores will increase weight on recommendation letters, not SoPs.
I feel that often domestic students remember the college application process, with its wishy-washy essays about everything, and that that leads them astray. The people reading your application are looking for evidence that you have potential to be a good researcher, which means experience, knowledge, and interest. Xyk’s point about conveying why you want to do science succinctly and without cliches is good, I think. I also very much agree with her point about childhood stories.
Broadly, I think your statement needs three things: 1) Why you want to do science (and why you want to do the specific science you want to do), 2) Your history (esp regarding research), and 3) What you want to do (including specific names of people you want to work with).
Ideally, the meat of your SoP should be #2 above. You want to show that you’ve done substantive research, and talk about your previous work showing that you understand the context of your work (ie, why it matters, how it relates to the literature). Describe exactly what you contributed, especially if you contributed new ideas to the project.
Regarding #3, my advice is to keep it short and simple. One paragraph at the end, describing broadly what you want to do, including a specific list of potential advisors. Be sure that the potential advisors do the thing you say you want to do!! Many people write SoPs stating that they’re interested in working with Professor C doing X, when he hasn’t worked on X in 10 years. Professors’ websites are not good ways to figure this out, as they are often not updated, but talking to people you know in the field is (ie, the advisors for your REUS, or emailing the professors directly expressing interest). It is not a red flag in my department if you have not contacted the people you would like to work with, but it can only help you to reach out to them.
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? That description frankly applies to most of our top graduating seniors at my PUI, and their grad school acceptance rates have tanked over the last ~4 years — used to be that a straight-A student (maybe an A- here or there) with a couple of REUs and a couple of years of research experience in our department culminating in a thesis could expect to get into about half of the 8-10 grad schools they applied to (and rightly so, I think — I did all my training at Ivies and R1’s, and our top students are absolutely on par with the top students at those places). Now it seems like if they apply to 12-15 schools, they’re lucky to get into more than one, and sometimes they don’t get in anywhere. From talking to my friends at R1s, it sounds like their application numbers are through the roof (probably due to a combination of larger numbers of majors AND the death spiral of higher application pressure leading students to submit more applications each). But on top of that, they mostly assign “bonus” points to students with first-author papers during the admissions process, but in such a high-pressure environment, those “bonus” points end up making the difference, and effectively a first-author paper is starting to be a prerequisite for admission to graduate school (which is bonkers, IMHO — part of the general insane arms race of everything in our society, and akin to some of the college admissions trends whereby 17-year-olds take multivariable calculus and linear algebra but don’t know how to call and make themselves a dentist appointment). I’m curious what it’s like at your state R1. We don’t see PhD admissions from the other end, so it’s hard to keep up with current trends and advise our students properly.
Ecology – my field/program at a state R1 is one where the admissions committee has very little input and the statement of purpose is basically written for the prospective advisor. I barely even look at applications from students I haven’t already spoken to (this happens rarely). I usually have a few applicants that I’m really excited about based on good conversations in the fall, but the personal statements and recommendation letters are what I look at to make the final call in terms of offering admission. I personally value hearing a student’s ‘voice’ in the statement – I want to get to know them and feel like they are a good fit. The weakest statements for me are ones that just regurgitate the CV and don’t tell me why they want to do a PhD.
It would be very rare for an undergraduate applicant to have a publication of any sort in my field, and I don’t think having a pub vs. having a thesis is any different in my mind. I would interpret having a pub as meaning that the student had an advisor with the time and desire to get it published. Ultimately, I’m just interested in whether the student has the experience to know what research is like so that they’re making an informed decision. On the other hand, if the student has an MSc and no publication, I would consider that a red flag.
This year has been rather extraordinary in terms of number of applicants for me – I don’t know if that’s COVID or just a blip, but I have ~10 applicants this year where normally I would expect ~3.
Bumping the answers to the last 2 Q&A. The number of cliche openings & broad statements where it’s not clear what sub-area the student wants to study are very difficult to evaluate.
Also, typos and formatting! So many essays are ugly/have weird spacing. Make sure you check it over MANY times!
Finally, I think it’s quite crazy to expect students to have a first-author (or any author!) publication. That is very dependent on luck at the undergrad level, more-so than talent.
lyra211, I will write more tomorrow about this.
@Lyra211 – I’m at pretty good university, definitely top 10 in physics. Top 2-3 in some subfields.
Roughly half of our accepted students are international, and somewhat different standards apply there.
Of the domestic students we heavily consider, virtually all of them have straight As and nearly perfect test scores (exceptions for people with truly extraordinary research backgrounds, or other extraordinary extenuating circumstances). This leaves us with 5-10x too many candidates. At this point, all we’re looking for is potential to be an excellent researcher. A vs A-, who cares. But a recc letter where someone says that the student was truly doing independent work – this matters a lot. I really don’t think publications enter that much, except that usually if you have a first author publication the corresponding recc letter is very strong. It’s hard for someone only doing 6-8 weeks of research in a lab to get their feet under them, let alone think creatively about the work. Undergrads at my institution give group meeting talks, participate in journal clubs, and think hard about their research project over the course of years. It’s hard to compete with the research maturity that that gives you.
A force in the opposite direction, however: we’re looking for the best, and the fifth best student from Harvard is just not that interesting. Maybe brilliant, but someone from a small college with incredible recc letters would do better.
So: yes, I think a student like the one above could be competitive. Ballparking, I think about half of our domestic accepted students have no publications. Many rejected students do have first author publications. Annually ~2-3 accepted students are from small liberal arts schools, and did all of their research through REUs. That’s not terribly high, but perhaps it’s higher than you are imagining?
I haven’t been around long enough to speak to changes.
Thanks for that response — it’s helpful. I’m in a physics-adjacent field that has a somewhat different culture, but I suspect that most of what you say applies as well. I just wanted to point to one thing that you said:
“Undergrads at my institution give group meeting talks, participate in journal clubs, and think hard about their research project over the course of years. It’s hard to compete with the research maturity that that gives you.”
Undergrads at PUIs do all those things too. Not all PUIs, of course, but at the top liberal arts colleges, absolutely. The main difference is that you have PhD students and we don’t. I think that’s a mixed bag, frankly: there are absolutely benefits to undergraduates from being around PhD students and seeing those research behaviors modeled at the next level (it’s one of the main reasons we encourage our students to go off and do REUs!). On the other hand, since as I’ve mentioned I did all my pre-faculty work at Ivies and R1s, I can confidently say that our undergrads frequently take *more* ownership of the research, and are *more* deeply involved than typical undergrads at R1s, because they are getting all of the faculty attention and the seniors are very conscious that they are the experts and the leaders within the research group. The buck stops with them (OK, really with me). When I do publish with students (which I do on average 1-2x per year), it’s always with undergrads as the first author; it’s just that it generally takes a year or even two post-graduation to get the paper accepted because there’s just not enough time for them to finish a thesis and go through the whole publication process with me during their time as an undergrad. So I actually worry that this (mis)perception that undergrads at R1s are somehow automatically better at research, or more immersed in a research environment, is incorrectly damaging to students at PUIs. It’s not aimed at you specifically; it’s just that I heard an echo of that common attitude in your comment.
PUIs have trouble keeping up with the rate of publication at R1s because we don’t have long-timescale PhD students to fuel the publication machine, and we have to train each student from the ground up… the quality is just as good, but the timescale is longer. I suspect that’s why so many of our students do better on the second go-round of PhD applications these days, when they’ve had time to get their thesis research submitted or even accepted. It’s the same student who was applying a year earlier; the only difference is the presence or absence of a first-author publication in hand (generally during the year between they are doing something science-adjacent, but not enough to really boost the quality of their application, particularly since they apply for the second time just a few months after graduation). The real heart-wrencher is that too often our FGLI, URM, etc. students are the ones who can’t afford to take two cracks at grad school admissions, and who leak from the pipeline before they have a chance to get that first-author publication on the books.
@lyra211
I suspect that for programs like mine where the faculty advisor is really the ‘decider’ as to which student(s) are accepted, it depends *a lot* on the background of the individual faculty mentor. Probably that’s also true of the make-up of admissions committees, which often change every year or two.
My own academic background is from a PUI, so I know the quality of attention and research training that undergrads get there. I also really like students from R2 state schools who have REU/thesis experience because I suspect they get similar attention with fewer resources and opportunities, and that impresses me. Having been a grad student and postdoc at an Ivy, they don’t impress me much.
I also went straight to PhD from undergrad, whereas master’s are common in my field and many of my colleagues will not accept PhD without MSc. I prefer them straight out of undergrad.
I guess the best advice I could give to prospective students is to make sure to talk to your prospective advisor and get their encouragement to apply, then get as much advice from your mentors as you can and take the pieces that work best for you.
I’m at an R1 university (AAU even) which has an unusually high ratio of undergrads to grads, and I’m in the department that gets the most funding (per faculty or absolutely) of any in the school of engineering. We’re also considered the name-brand school in a few niche fields (including what half my department does), but treated like an R2 for many other fields (in some cases deservedly so). The top students graduating from our department at least often come out with first-author publications (but our department is unusual—with 8% of the school of engineering’s students, they get about 80% of the awards).
Our department does lab rotations and contacting faculty ahead of time is usually a waste of everyone’s time—unless you already have something to contribute to a collaboration. I particularly dislike the email sent via mailchimp—it is clear that those students are just spamming.
We do get a lot of top applicants, but we also know that everyone we are willing to accept will also get offers from rich places like Stanford and MIT, so our two main concerns are 1) can they do research? and 2) do they really want to come here? We cannot pay what the rich schools do—our best fellowship is smaller than the least offer that Stanford, MIT, or a dozen other schools would offer.
We have eliminated the GRE requirement (there was no really relevant subject GRE and the general GRE just tests whether students have had an adequate high-school eduction—though it was useful for weeding out the bottom third of the pool). Almost all the weight ends up on the letters of recommendation and the students’ description of their prior research experience. We don’t expect prior publication (though it certainly helps)—a recommender who is excited about the student’s research projects and describes the student’s contributions to them in loving detail carries a lot of weight.
This discussion seems to me to be about a month too late—our grad application pool closed on Dec 1, and the applicants’ files have been assigned to faculty to review. If specific faculty were named in the statement of purpose, they will generally get assigned to review the application.
This year we’ll be accepting very few PhD applicants because our TA funding was cut by 4 students and we did not get a 6-student training grant renewed (I understand that the problem was that a progress report was not filed—by the faculty member who had demanded a full course relief every year because of the workload of filing the progress reports). We normally take in about 10 students a year, but with funding for 10 fewer students, we’ll probably be accepting only enough to take in 3 or 4. MS admissions will probably go up to compensate, as we don’t have to provide funds for them (and an MS in our field is highly employable, so arguably we should have been taking more MS students all along).
I’ve been rambling all over here—the bottom-line is that the advice others have been giving is sound: check out what the target department’s practices are and whether there are multiple faculty there you want to work with, do research as an undergrad, write passionately about the research you have done and hope to do, propose topics that the department actually does, get good recommendation letters by being a productive lab-group member, and make it clear to your first choice school that they are your first choice. Even if your application is great, you may get rejected, because of external factors (loss of a grant, too large a cohort the previous year, … )..