This is a guide for those preparing to apply to, applying to, interviewing for, and choosing science PhD programs and fellowships, primarily in the US. It assumes that you have already decided to apply to grad school and are willing to put forth a bit more effort than a few Google searches and coin flips.

Why should we listen to you?
If you are considering basing your entire grad school application process on a single obscure blog post, then perhaps a life in research is not your best option. I highly encourage you to seek advice from a range of professors, administrators, researchers, and grad students, especially those who know you and can offer more personal advice. The lessons and suggestions below are merely anecdotes drawn from my own experiences applying to grad school during Fall 2010/Spring 2011.

For full disclosure, I applied to a hodgepodge of programs that would support research in theoretical neuroscience (which translated to Neuroscience, Biophysics, Physics, and Applied Math programs), as well as just about every fellowship for which I was eligible. With the generous guidance of about a dozen professors and grad students, I managed to snag a Churchill Scholarship to spend one year at the University of Cambridge and a Department of Energy Computational Sciences Graduate Fellowship (DOE CSGF), a National Science Foundation Graduate Research Fellowship, and a Hertz Fellowship to support my PhD studies thereafter. I was also fortunate enough to receive offers of admission from the majority of schools I applied to but deferred or declined them all to spend a year in the UK. (During my year in the UK, I reapplied to a few schools and ended up going to Princeton for its physics program.)

My goal in writing up this guide is not to help you design your college experience, craft the perfect resume, and con your way into grad school, but rather (1) to leverage the obscene amount of time I spent gathering information on grad schools and fellowships to save you time and (2) to help you make a reasonably well-informed decision.

The Process

1. Identify faculty, schools, and programs (in that order)
2. Contact faculty and students
3. Apply for fellowships
5. Worry
6. Interview and/or visit
7. Worry
8. Decide

1. Identify faculty, schools, and programs (in that order)

Why in that order?
You are going to grad school to learn how to do research (or to escape a poor job market, but my recommendation stands); finding an appropriate research advisor is one of the most challenging and important parts of ensuring that you have a good experience. By focusing on schools or programs first, you may become distracted by far less important details, such as coursework, impressive web pages, and ambitious program names.

Are you saying schools and programs do not matter?
Not quite. A school defines where you will live and who you will interact with, and a program defines your first-year course requirements and may restrict who you can work with. My point is that working with a great advisor at a good school is, for most people, going to be a better experience than working with a mediocre advisor at a great university.

What criteria should I consider in looking for faculty?
Interesting research - Obviously, you need to be interested in their research.

Well-connected - Since your advisor will likely be your primary liaison for establishing collaborations during grad school and finding jobs afterwards, you might want to find someone who is well-connected and collaborates frequently in your research community. Frequent invites to speak at seminars and conferences, a diversity of co-authors on publications, and frequent citations of his/her research (in journals, not USA Today) are all good signs.

Advisement style - Advisors vary widely in how closely they interact with students. Some will assign their students projects and meet with them for a couple of hours multiple times a week. Others prefer to let their students struggle a bit to define their own research problem and may not require any meetings, serving only as a source of advice a couple of times per month or semester. Where your optimal advisor falls on this spectrum of involvement depends on your own personality and preferences. Emails to grad students, interviews, and visits (discussed below) will help you determine the advisement style of the faculty in whom you are interested.

Availability - Your advisor also needs to have the time and funding to take on students. At a minimum, they need to be willing to advise you, but depending on your preferences for interaction, you may want to consider how much time they actually have for their students. A quick glance at their website may reveal whether they are taking on students but if not, a brief email asking this question is absolutely acceptable. As for determining how much time they actually have, emails to grad students, interviews, and visits are again helpful and will be discussed below.

Great explainer - Another helpful quality in an advisor is the ability to clearly explain complicated ideas. Not all great researchers make great mentors, and having an advisor who is a great explainer can help you learn a lot more during your time in grad school. Looking up talks and papers by a professor, as well as interviews and visits, can help you infer how well they explain their work.

Personality - Finally, you need to get along with your advisor. Choosing an advisor because you thought they were “funny” or because they shared your interest in crocheting may sound absurd, but you will be spending between four and seven (or more) years working with them. Due to the nature of scientific research, you will inevitably run into many failures, dead ends, and frustrations, and overcoming such obstacles will be far easier if you have an advisor you enjoy working with. Again, not all great researchers make great mentors.

These criteria of course reflect my own particular preferences, and your own may differ. I strongly recommend drafting your own list before contacting professors or visiting schools, so that you have some idea of what questions you should be asking and what you should be looking for. Update your list as you learn more throughout the application and visiting process.

What criteria should I consider in looking for schools and programs?
Livability - If you grew up in Orange County and will need to be rushed to the hospital if a snowflake impinges upon your sensitive tanned skin, then perhaps the University of Chicago is not your best option. If you were raised climbing trees and having picnics in Oregon, then think carefully before applying to NYU. If you are a diehard rock climber, then perhaps you should pass on the University of Kansas. At a minimum, you should be sure that you can lead a moderately satisfying life in any city containing a grad school to which you apply.

Community - Your school and program more or less define the people you will collaborate, eat lunch, and socialize with for several years. The people in your program are usually the first that you get to know (since you will likely take classes with them for at least a year or two), though how close-knit these communities remain over the next few years varies greatly by school and discipline (mathematicians, for example, typically shun human contact and live in hermit caves after year 3). I have had one professor tell me that the single factor that contributed most to his positive experience in grad school was the enthusiasm and motivation of the students around him; it is much easier to deal with the trials and tribulations of PhD research when you are constantly reminded of how much fun science can be. In addition, your colleagues in grad school will likely become your collaborators afterwards, so surrounding yourself with people who do good work may serve you well in the future. Visiting your program and talking to students ahead of time is the best way to assess whether the community will be a good fit for you.

Multiple advisors - While you may enter grad school gung ho on working with a particular professor, your interests may shift, that professor’s interests may shift, funding may be an issue, or you simply may find that you two do not get along. To prevent yourself from needing to switch schools (or worse, work on something you are not interested in), look for schools and programs that have more than one faculty member you can see yourself working with. Also, if your interests span multiple disciplines, you may need multiple advisors with different areas of expertise anyways.

Advisor flexibility - Make sure that your program gives you the flexibility to work with any of the multiple advisors in which you are interested, regardless of their department affiliations. In my experience, this was never an issue, though some more traditional departments may emphasize “training their own.”

Coursework - This criterion is listed last for a reason. While long lists of interesting courses make for impressive websites, it is your advisor, research, and colleagues that will be far more important in the long run. In addition, many (most?) grad students end up teaching themselves and each other skills as needed during research. Do not be distracted by the red herring of coursework. That said, make sure that you can tolerate the course requirements for any program you apply to.

Just as for faculty criteria, this is my own biased list, and you should draft your own as well.

How should I find faculty, schools, and programs?
Ask professors and/or grad students who know you well and/or are knowledgeable about your area of interest for suggestions. Browse review articles, textbooks, and papers and note faculty whose research interests you; follow up by browsing their websites. Diligently scanning the complete faculty profiles for schools and departments in which you might be interested also works, though exhaustive search is linear in problem size (in other words, its time-consuming). As for programs, do not feel as though you must apply to only programs with a particular name; it is more important that you find a program that will give you the flexibility and support to do the research you want to do with the people you want to do it with (again, I personally applied to programs with titles ranging from “Physics” to “Neuroscience” to “Applied Math”). Keep a detailed list of faculty, schools, and programs as you go. For perspective, the Google Doc I used for this purpose outputs a 53-page PDF. Gird thy loins!

2. Contact faculty and students

[GASP] You can email professors?!

How should I email professors and what should I ask them?

Why would I want to email grad students?

3-4. Apply for fellowships and grad schools

What are fellowships?
Funding, mostly distributed by the government though also by several private organizations, that will follow you to whichever school you choose.

Why should I apply for fellowships?
Preparation - Fellowship deadlines are typically earlier than those for grad schools, and many of the essays, recommendation letters, and other application materials that you prepare can be reused. Inevitably, you will start putting together your grad school applications too late, and the forced preparation of fellowship applications will save you from missing deadlines.

Grad schools care - Many grad programs require their current students to apply for fellowships, and if you can say that you have done so before even applying, you may impress them with your apparent competence.

Academic freedom - There is a slight chance that you might actually win a fellowship and if so, you will then have a reliable source of funding that follows you to whatever university and research group that you choose. This means that funding is no longer an obstacle to your working with a particular group. It means that your research will not be tied to a particular grant so that you have far more freedom in selecting a thesis topic. It also frees you from taking on teaching assistantships for funding, thus giving you more time to focus on your research. You may still decide that you want to teach or your program may still require a semester or two of teaching in order for you to graduate, but you now have flexibility. Lastly, if you are wait-listed or rejected from a school that you are still keen on attending, a fellowship may encourage the admissions committee to reconsider, as you are now essentially free labor.

What are the major fellowships?
In increasing order of benefits, there are the Ford, NSF, NDSEG, DOE CSGF, and Hertz fellowships. Each involves some combination of tuition, living stipend, research/travel grants, research internships, and conferences, lasts for 3-5 years, and may only be used at schools in the US. I will abstain from listing the benefits of each program, as they may change year to year. Check their websites for up-to-date info. In my case, I missed the deadline for the Ford, applied for the NSF, NDSEG, CSGF, and Hertz, received the NSF and CSGF, and made it to the final round of interviews for the Hertz (more on this experience below), so I will offer comments on each of the NSF, NDSEG, CSGF, and Hertz.

What can you tell me about applying for the NSF?

What can you tell me about applying for the NDSEG?
The most unique feature of the NDSEG is that your application must pique the interest of a Department of Defense research office. This does not mean that you necessarily need to be building the Death Star or mind control devices. In fact, every NDSEG fellow that I contacted mentioned that they made no special efforts to reference military applications of their work when they applied. Also, while the NDSEG site mentions that you can browse the websites of the various research offices to see what kinds of projects they find interesting, I would advise against it for two reasons. One, no NDSEG fellow I contacted mentioned any perceived benefit in doing so. Two, government websites are notoriously difficult to navigate and out of date. It seems the best strategy is to apply just as you would for any other fellowship and hope for the best.

What can you tell me about applying for the CSGF?
The CSGF stands for “Computational Sciences Graduate Fellowship”, but “Computational Sciences” is broadly defined as any research leveraging computers and/or mathematics, thus you need not have a server farm in your basement to qualify. In addition to the very generous financial benefits, the CSGF is unique in requiring a summer internship at a national lab. This feature can either be viewed as a necessary obstacle to receiving your funding or as a great opportunity to try out a non-thesis-related research project at one of 17 stellar labs across the United States.

What can you tell me about applying for the Hertz?
Not only is the Hertz the most competitive (about a dozen superstars pick one up each year), lucrative (5 years of stipend and tuition), and prestigious fellowship in the US, it is also by far the most fun to apply for, as it includes two rounds of intense interviews.

After the initial applications in October, about a quarter of applicants are invited to interview. If you are among this lucky bunch, you will receive an email reminiscent of a James Bond movie, requesting that you don a suit, meet at a hotel, and:

After receiving this message, I was preparing to be doing push-ups and integrals at the behest of a military officer. In reality, although the interviewers asked impeccably sharp questions, the conversation was quite casual. (I was, admittedly, slightly disappointed that I did not get to do push-ups and integrals.) My question topics ranged from my personal motivations (e.g. “How did you become interested in neuroscience?”, “Where do you see yourself in ten years?”) to my previous research (e.g. “Tell us what you did in this project.”) to creative puzzles (e.g. “If you had a friend on the other side of the moon and were banned from launching anything into space, how would you communicate with him?”). In all cases, the interviewers (there are typically two) continually peppered me with sub-questions, ranging from clarifications on specifics (e.g. “And about how large is a neuron?”) to speculations on broader issues (e.g. “And what implications do you think that has for the field of neuroscience?”). Note that you will almost certainly be stumped by one or more of the questions asked and that this does not necessarily disqualify you. I stumbled on at least two questions in the first round and still managed to limp my way to the second round. (The second round interview is very similar.) Personally, I applied for the Hertz mainly because the interviews sounded like fun and I was not disappointed. If you have a similarly sadistic sense of pleasure, then this is the fellowship for you!

I just got a first-round interview! Any advice?
Know your research well, not just the technical details but also the broader significance. Be able to explain why your projects were important and exactly what role you played in each one. One question I was unprepared for was: “What is the most creative idea you have had in your research?” Another was: “What is the most interesting new tool available to scientists in your field?” You should be able to come up with better answers to these than “I vectorized my Matlab code” and “Google Plus”.

Most importantly, be prepared to be humbled. The interviewers will probably be nice, but they will also probably push you to and beyond your intellectual limits. That is their way of testing you, so do not feel ashamed if you cannot answer some of their questions. What you should not do is remain silent for long periods of time, mumble, or bullshit your way through an answer. Instead, vocalize your thought process, and be honest and precise about any uncertainties you have. The interviewers want to see how you think, and they cannot do so if you sit there daydreaming.

Finally, remain confident. Treat the interview like a game, and I am sure you will have lots of fun. I truly enjoyed my interviews and would do it again in a second.

Sounds great! Can I receive all of these fellowships?
Up until 2010, it was possible to accept multiple fellowships and use them one after another. Beginning in 2011, however, the US government decided to spread its funding and no single student can accept more than one scholarship from a federal source (which includes the Ford, NSF, NDSEG, and CSGF). Since the Hertz is awarded by a private organization, it might be possible to accept it along with a federal fellowship, but this is between you and your funding sources. In any case, if you win a Hertz, you should have cured cancer and disproved P=NP by your fifth year, so a second fellowship should not be necessary.

What are applications like in general?
Applications typically ask for a mixture of essays (e.g. personal statement, summary of previous research), recommendation letters (typically three, though as many as five), general GRE scores, transcripts (sometimes official, sometimes unofficial), and personal info (address and such). Some may ask for a subject GRE score in your area of focus. Grad schools have application fees (typically $60-$120), but fellowships do not.

How should I prepare to apply for fellowships and grad schools?

How should I prepare for the GREs?
If you are a native English speaker, all you need to do for the general GRE is to take a couple of practice tests ahead of time to understand the format. ETS (the inept company that administers the GREs) offers test prep software that allows you to simulate the riveting experience of using a computer in the early 1980s… and taking the painfully primitive test in an official GRE test center. There are also plenty of practice books available from companies like the Princeton Review that offer additional practice tests. The test is similar in nature to the SATs and your score is probably not particularly important, so you should not spend too much time preparing for or worrying about the test. If you are not a native English speaker, I hear the test is considerably more difficult. Sorry.

The subject tests, on the other hand, are far more challenging and do require practice. The only subject tests I have had experience with are physics and math. As for math, it is rumored that universities do not particularly care how well you did on the subject test, though they require it anyways. Physics programs, on the other hand, worship the physics GRE, and rumor has it that you need a 900+ score for top programs to even look at your application. Fortunately, there are four old exams floating around the interwebs as well as a fantastic website discussing various approaches to solving the individual problems. I recommend taking each of the (timed) tests on your own and reviewing the web solutions afterwards to fill in the gaps in your learning (I took each twice over a period of 6 weeks and did fine on the exam). Ignore prep books for the physics test; they are a waste of money, as the practice exams and web solutions offer plenty of preparation.

Who should I ask to write my recommendation letters?
Ideally, a Nobel Prize winner who swears that you are the spawn of Isaac Newton and Mother Teresa and will revolutionize your field before quals. Realistically, at least two professors who have overseen your research and an academic advisor who can speak about your coursework, participation in academic programs, and general background and personality. Having recommendations from a professor outside of your university (e.g. from a summer research program) or one who plays poker and shares oscilloscopes with faculty on the review committee (yes, scientists are humans too) are considered pluses. However, the party line is that you are better off with a good letter from an unknown professor who knows you well than a mediocre letter from a Nobel Prize winner who met you once at a seminar.

How should I write my personal statement?

Fellowships, on the other hand, are slightly different. They seem to be slightly less interested in your interests and motivations and more interested in your previous and proposed research. Several fellowships, including the NSF, require project proposals, and in these cases, feel free to go into far more detail than in your grad school applications.

Will I be hanged, drawn, and quartered for breaching the page or word limits for essays?
I doubt any one actually counts words in your essay, although an extra page may raise eyebrows. Streamlining your essays for readability over poetic flourish is a good idea, but do not stress over a few extra words.

How many schools should I apply to?
As many as you see yourself potentially striving at. I know students who applied to as few as 3 and as many as 15 (and pre-meds are known to commonly apply to 20 or more schools), though the average is probably between 5 and 10. I personally applied to 13 schools, a number at which I arrived by applying to every school which I could plausibly see as my top choice after visiting. Reasons to limit the number of schools to which you apply include applications fees (typically $60-$120/school), time spent applying (though this decreases rapidly since you can recycle essays), and difficulty scheduling visits in the spring (if you are fortunate enough to be accepted). Reasons to apply to more schools include increasing the probability that you get in somewhere, increasing the number of visits you get to do in the spring, and allowing for a “borderline” school to surprise you during a visit. Your own optimum school number will depend on how many schools fit your interests, your pre-application confidence in your admission likelihood and school preferences, your financial status, and your time for and interest in spring visits.

Where can I find more additional resources on applying for fellowships and grad schools?

Advice from other previous applications, including:

My own application essays, including:

Additional tips from the NSF (though I found these to generalize to other applications) on:

5. Worry

Now that you have submitted all of your applications, it is time to sit in front of your email inbox, hitting refresh on your browser 24 hours a day for the next couple of months.

6. Interview and/or visit

Wait, I have to interview?
This varies by discipline. Some programs require in-person interviews, others may conduct phone or Skype interviews, and still others may admit/reject students based on their applications alone. In the latter two cases, accepted students may be invited to visit before making a decision. For some reason, interviews and invited visits seem to be more common in the life sciences than in math and physics. This may either reflect the cultures of the fields (biologists are usually slightly more gregarious than mathematicians) or the amount of available funding (as a general rule, the more applied departments usually have more money).

Should I be nervous?
Not at all. Interviews and visits are the best part of applying to grad schools! Be prepared for several weeks of:

• one-on-one time with faculty to discuss their current research, potential projects for yourself, or just about anything else
• plenty of time to talk to current grad students
• lab tours
• the opportunity to meet other students from all over the country and world with similar interests
• ...and (surprise!) free travel, meals, booze, and entertainment!

That’s right - spring visits are basically extended science parties. For both interviews and visits, programs typically plan a mix of one-on-one meetings with faculty (30 minutes to one hour), info sessions, lab tours, faculty talks, student poster sessions, lunches, dinners, and receptions with students and/or faculty, and entertainment, including everything from parties to hikes to aquarium visits. “Interviews” are as much recruitment as they are actual interviews, so prepare to courted. Also, in my experience, any reasonable expenses will be reimbursed, from baggage fees to taxis to lunch to coffee. I even heard one rumor of a group of students who ran up a \$2000 wine bill during dinner and managed to get comped, though I recommend against trying this yourself.

Does that mean I am guaranteed admission if I am invited to interview?
Not quite, though the odds are in your favor. Most programs claim not to have a particular quota in mind, but rumor has it that typically somewhere between 50% and 75% of interviewees are accepted. Note that more students are accepted than desired to enroll since programs anticipate that not all students will accept their offers. Note also that a post-interview rejection does not necessarily mean that you were a disappointment in person. Interviews are as much about judging your own fit and interest in the program, and faculty may have sensed that your interests were better served elsewhere.

How should I prepare?
Publish three Nature papers, lose 20 pounds, and memorize all of Shakespeare’s sonnets.

What should I ask faculty during my interview/visit?

While you can also discuss research with them, grad students are your go-to resource for questions on local living and entertainment, coursework, and what it is actually like to work with particular professors. In addition, I found the following three questions to be very helpful: (1) what is the worst thing about your school/program?, (2) what was the biggest surprise?, and (3) where else did you consider going and why did you choose to come here? See the school and program criteria I posted above for some other ideas.

What if I do not get to meet with a faculty member who I am really interested in speaking with?
Many faculty may be traveling or otherwise unavailable during interview/visit weekend, while others are simply over-requested by prospective students and cannot meet with everyone. If someone you requested is missing from your interview schedule, try asking the program administrator if you can replace a less desirable interview or fit in some one-on-one time elsewhere. For example, you might meet with a professor at a reception, over a meal, or during a lab tour. You can also speak with current grad students advised by that professor or other prospective students who interviewed with that professor. If you still feel neglected at the end of your visit, feel free to email a professor to ask questions, set up a phone discussion, or plan to meet at an upcoming conference.

What do I do if I was accepted but not offered a visit and would like to make one?
Let the program director know that you are interested in visiting and, in some cases, they may offer to help pay for a visit (the only school to which I applied that did not offer visits was willing to do this). Even if they do not have the funds to do so, they will likely be willing to set up meetings with professors and students if you can cover travel costs. If you are in this situation with regards to a school you are very interested in, I highly recommend investing in a visit. Choosing a grad school is a major commitment and you do not want to condemn yourself to several years working with an advisor you do not get along with on a project you do not care about in a city you despise simply because you were too cheap to splurge on a visit.

7. Decide