After an extensive application/visit/decision process, I cataloged some of the information I gathered about the graduate school application process. Toward the end of grad school, I took another look at this document. Overall the advice seems sound in hindsight, but I’ve inserted some important edits and revisions. I’m guessing there will be something here for everyone applying to PhD programs in engineering, and for those considering other types of programs (master’s, or degrees in the physical sciences or humanities) to a lesser extent.

Preparation

Ideally, you began thinking about graduate school during your freshman or sophomore year. You chose to take hard courses, you got involved in research (and obtain funding for your work), and you performed well. If you are lacking in one of these areas, it will hurt your application. This is not because graduate programs expect you to possess a specific set of skills, but simply because you need to demonstrate exceptional interest and ability in academics to be considered at top programs. There are incompetent people accepted and competent people rejected from good graduate programs; admissions committees make imperfect decisions using imperfect information. Your goal should be to give them no reason to throw your application away.

A successful application contains the following components, in (roughly) increasing order of importance. Items 1 through 4 are used principally to screen the majority of applicants. Items 5 and 6 are used to decide among the remaining applicants.

School-specific application

Each school’s application is a little different, and it takes a surprising amount of time to enter the basic information into each. Budget upwards of one hour per school to sign up and fill in this information.

GRE Scores

Everyone has to take the general GRE, and it’s fairly inconsequential. Buy a study book and read it simply so you know the logistics of the exam (what material will be covered, what types of questions will be asked, how many of each type of section you will see. Chances are you’ll get a perfect 800 on the quantitative portion), since 20% of EE grad students do. The verbal is trickier, and your time is best spent learning vocabulary. If your programs require a particular subject test (e.g. Mathematics, Chemistry), you ought to study hard for it; scoring well will result in automatic or early admission to many programs and increase the likelihood that you will be awarded a fellowship. Note that the subject tests are only administered a few times per year, and that desirable test times for the general exam will fill up well in advance of the test date, so schedule your exam as early as possible. Ideally, you would take the subject test over the summer and the general exam in early fall. Note that if you take the general exam after November, your score will not be considered for the NSF or NDSEG fellowships.

Official Transcripts

Order these early, and if they’re cheap, just send them to every school and fellowship you’re considering. It’s not worth being rejected due to late transcripts. As I mentioned earlier, you ought to have taken hard courses and performed well.

Curriculum Vitae

If you’ve been applying for scholarships and research grants, you should already have this document. Be more explicit and technical about any research or project experience. Give full citations for any publications, including a URL if possible. Specifically highlight any peer-reviewed publications. LaTeX it for extra points.

Personal Statement

If you’ve been applying for scholarships and grants, you should be relatively proficient at writing these. It’s best if you revise this document multiple times, and get many people to read it. It’s also best if you tailor the document to each program you apply to (EDIT: specifically mention individual faculty and research projects you are particularly interested in). It won’t break your application if you don’t do these things, but if you have any other weak points in your application, then you must have an extremely polished personal statement that addresses your weaknesses. Your statement should be no longer than 2 pages and should address the following.

• What area do you want to get a PhD in?
• Why do you want to get a PhD in that area?
• Why do you want to get a PhD at this school?
• What makes you particularly well-suited and well-prepared to study in that field at that school?

Letters of Recommendation

This is easily the most important component of your application. You should have three letters from faculty who know you well. Ideally, you have a letter from three leading professors in your field whom you have written papers with. More realistically, you have one letter from a professor you have written a paper with, one letter from a professor or post-doc you have completed a project with (through a class or through research), and one letter from a professor you have had substantial interactions with (many classes, relevant extracurricular activities, teaching assistantship). It will severely hurt your application if you do not have three strong letters, so diversify your efforts as an undergraduate:

• talk to many professors
• get involved in a variety of projects
• pursue collaborations with your current research
• apply for summer research programs at other universities

If you are applying to sufficiently varied programs (e.g. some Computer Engineering and some Bioengineering), it may be worthwhile to get different collections of professors to submit letters to different programs.

Applying

Selecting Programs

You should apply to as many programs as you can afford. I applied to 10, I had a friend who applied to 11, and I’ve met a person who applied to 20. I think 8 is a good number to aim for. To select the programs, call upon your cumulative experience in your field. Who are the leaders in your field? Where have breakthroughs occurred in the past? What programs are declining? What programs are up-and-coming? If you are applying to PhD programs in engineering, then four of the top five programs in your specific field are probably at MIT, Caltech, Berkeley, and Stanford, so plan to apply to those schools. Even if you currently don’t think you would want to go there, it’s worthwhile to apply simply to visit, because these schools have almost certainly generated and harbored some of the preeminent thinkers in your discipline. It’s worth your time to have conversations with these people.

At this point, you may know what other programs you should apply to. If not, consider your letter writers. Where did they earn their PhD? Where have they sent students recently? Where are their colleagues? Where have they recently given invited talks? Apply the same reasoning to other top researchers in your field. Give preference to the careers of younger faculty, as the programs that were good several decades ago may have declined in quality.

As a last effort, talk to other people. It’s best if you enter these conversations with a good idea about which programs you want to apply to, so you can avoid repetitive conversations (“Have you considered MIT?”; “MIT? No, I’ve never even heard of it. Is it a good school?”). Current graduate students can give you good personal information (how much each program pays, whether the neighboring city is fun to live in, how hard the students are expected to work) while faculty can give you good career information (who sends students to good faculty positions, who is close to retirement, who is likely to move, who is headed for administration, who has stopped going to conferences, who has stopped taking students). However, you should realize that any information you obtain second-hand is inherently biased. If you talk to enough people, you will surely be given conflicting information.

Note that you will have one (or at most two) faculty (co-)advisor(s) for your PhD. Consequently, you should select programs that have at least one faculty member you could see yourself working with. If you want to apply to a particular school but don’t see faculty with compatible research interests in your department, check the other departments.

Selecting Fellowships

If you are an engineer, you should apply for the NSF, NDSEG, and possibly the Hertz. The NSF and NDSEG applications are essentially identical, except the NSF is due approximately one month before the NDSEG and generally two weeks before the first round of school applications. Fill out these applications even if you don’t have a strong research statement: the essays can be modified to become your personal statement for grad apps, and the process will make you more articulate when you talk to faculty during visit weekends.

A word about the NSF

It was not clear to me when I read the essay requirements, but “Broader Impacts” should constitute roughly half your proposal. In my feedback from the NSF, I received top marks for the content of my research proposal, but poor ratings for my Broader Impacts statement. If you want the fellowship, you will need to present strong arguments that your research will benefit (a) minority groups, (b) younger students, (c) society at large, and (d) the scientific community in general. (EDIT: I reworked my essay by devoting a much larger portion to “Broader Impacts” and received the fellowship as a first-year graduate student.)

A word about the Hertz

I would not recommend applying for the Hertz unless you are extremely confident in your abilities as a researcher. I gave a terrible interview for the Hertz, and it was the most demoralizing part of my graduate application experience. You are probably capable of earning a PhD at a top program, but if you botch an interview with a brilliant scientist, you may will be discouraged about your prospects for doing graduate-level research. Since a significant portion of a PhD is earned through sheer determination, this is an unnecessarily challenging experience to have as you embark on graduate study.

Logistics

Applying to graduate programs is expensive, particularly if you are simultaneously paying tuition and living expenses. Anticipate paying $100 per school (including the application fee, official transcripts, and GRE score reporting), in addition to $150 per GRE exam.

Visiting

With any luck, you will be admitted to many of the programs you apply to. It is unlikely you will be admitted to every program, as there are many qualified applicants competing for a small number of positions. But if you are admitted, you are typically guaranteed generous funding and invited to travel to the school on a (maybe partially) subsidized visit weekend. These trips are your best opportunity to gather gut-level information about the school, so you should attend as many trips as is feasible. There are certain kinds of information you can spend time gathering during a visit that will be more useful than others, so learn as much as you can about a program before visiting. For instance, you can read about teaching, coursework, and qualifying exam requirements online, whereas you cannot so easily learn whether a particular professor is surly or frequently unavailable without talking to their students.

Gathering Information

Here is a list of the kind of information you should try to learn on a visit:

• Is the professor you want to work with accepting students? Do their students perceive them as being readily available? What kind of active grants do they have? Is their advising style compatible with what you (think you) want?
• How often do students leave the program due to lack of funding? Due to failing qualifying or preliminary exams?
• Where would you spend most of your time (i.e. after you are done with coursework)?

Here is the kind of information you should know before you go on a visit:

• What are the coursework, teaching, and qualifying exam requirements?

Faculty Conversations

If the program is conscientious, they will schedule half-hour conversations between you and 4-8 faculty with similar interests to yours. It is unlikely you will have one of these meetings with anyone known to the general public (henceforth referred to as famous), but you may be lucky enough to rope them into conversation over lunch or dinner. I could not discern a recipe for making these conversations productive, except that I generally enjoyed conversations with faculty outside my field the most. I found it more useful to read a faculty’s review papers than their journal papers in advance of the conversations. If you have a personal connection with a particular person (through your advisor, or through a particular project), bring it up. The best conversations I had involved the faculty either telling me about current, unpublished, risky research, or candidly discussing other researchers in the field. Keep in mind that these people will have dozens of conversations similar to the one they have with you, so anything you can do to make the conversation unique will be beneficial for both parties. Finally, note that in engineering programs you’ve already been accepted to the program and the faculty are trying to recruit you, so most will try to convince you that their university provides the best and most interdisciplinary academic environment and that their city is the most exciting to live in.

Logistics

Fortunately you should be reimbursed for the majority of your travel expenses, especially if you are visiting top programs at top schools. Unfortunately this means you will need upwards of $2,000 to bankroll the trips. If you are visiting a top program at an overall mediocre state school, be prepared to pay for your travel expenses. Purchase refundable tickets if possible, as you may want to change your travel plans (for instance, if you are admitted to a more prestigious program after you buy the ticket).

Choosing

There are many factors to consider when choosing which school you will attend, ranging from the critical to the fanciful: research topics; research style; collaboration possibilities; relationship with your advisor; relationship with your peers; career potential; where you will live; who you will live with; what you will do for fun; weather; proximity to civilization; proximity to nature; proximity to family; proximity to your significant other; proximity to Europe; prestige of the program; prestige of your advisor; culture; East Coast versus West Coast; availability of good coffee or beer or sushi or ballet or bluegrass; average time to graduate; difficulty of qualifying exams; opportunities to teach, take courses, participate in intramural sports, or join clubs; quality of undergraduate program; quality of other graduate programs; architecture on campus; strength of campus athletics; school mascot and color palette.

The vast majority of these factors did not contribute to my final decision. I made my decision by thinking about which research opportunity I was most excited about, and thought had the greatest potential for real progress and impact. I currently believe that the most important factor that determines whether you will be a successful PhD student is your passion for research. I claim that most other factors (the city you live in, the other people in your lab, your program requirements, even your research advisor) will have a smaller effect on your overall happiness than your satisfaction with the work you’ll be doing. I warn the reader that this perspective likely stems from the particulars of my personality and consequently I cannot recommend it for general consumption. (EDIT: Although I still endorse this attitude overall, in hindsight I will admit that while choosing I was in the fortunate position where the best research opportunities were also in areas with exceptionally high quality-of-life for graduate students. The impact this had on my happiness and productivity during my PhD is significant.)

Timeline

This section is provided as a reference for the time-sensitive information contained above. Almost all graduate applications are due between December 15th and January 15th. You need to take the GRE (general and, as applicable, subject) before December begins to have your scores considered for the NSF fellowship. With any luck, you will travel nearly every weekend between late February and early April to visit schools; take a light courseload during Winter quarter to make this as painless as possible. Your final decision must be rendered before April $15^{th}$; this is a common deadline for all PhD programs I am aware of.

• Summer Study for and take your GRE subject test.
• Sep. Secure your letter writers.
• Oct. Study for and take the GRE general exam.
• Nov. Select the schools you wish to apply to, write your NSF essays.
• Dec. 1 Submit your NSF and Hertz applications, write your grad essays.
• Dec. 15 Begin submitting school applications.
• Jan. 1 Submit your NDSEG application.
• Jan. Sign up for a light courseload to make travel less taxing.
• Feb. Wait to hear back from your programs.
• March Visit your programs.
• April 15 Make your final decision.