I originally wrote up these lessons as a little novelette, the writing of which served as both a source of therapy and a structure with which to organize my ideas. I imagined I would write the foil to Philip Guo’s excellent The Ph.D. Grind for the everyman. While Philip gave the impression that he knew what he was getting into, attended two elite institutions (MIT and Stanford), and studied a glowing hot field (CS), I was more or less drifted my way into a PhD program because I wasn’t quite sure how to get a job. However, after finishing up my manuscript, I realized that 1) people prefer the TL;DR and 2) the guidance I would like to impart to current or potential students could be summarized in a much shorter blog post.
While I am overall very happy with my PhD experience, am proud of what I accomplished in graduate school, and am amazed by the incredible place I landed, I believe I would have been much better off if I had a practical guide in front of me when making important grad school-related decisions. Not a peaches and cream pamphlet from the National Science Foundation, not a conversation with a favorite professor, but a practical guide that was written by a former student who recently completed his doctorate that could shed some light on the process. What follows is that guide.
What is a PhD?
Based on my own experience and my experience formally or informally mentoring dozens of undergraduates and first-year students, I believe that the vast majority of students entering or considering entering PhD programs in the hard science have very little idea of what a PhD means. I occasionally even talk to third and fourth year students who don’t quite understand the meaning of the degree.
A PhD is an advanced degree designed to prepare students for a career in research. Not teaching. Not mentoring. Not entrepreneurship. Not leadership. Research. A PhD student spends nearly all of his time designing and performing experiments and communicating his results to colleagues. Coursework is generally completed in a year or two, teaching, if required at all, is a much lower priority than research, and activities like mentoring or entrepreneurship are generally recognized with a pat on the back, if at all. While this nearly complete emphasis on research probably does not benefit many students in the current economy, the faster this lesson is learned, the better. Every year while in graduate school I met a number of students who did not appreciate this fact and generally became frustrated and dropped out.
Should I pursue a PhD?
Until recently, obtaining a PhD was far less common. Corporate research positions generally did not require advanced degrees and there were plentiful opportunities for young scientists right out of college. Those who pursued doctoral degrees were generally very academically inclined and tenured professorships were growing at fast enough rates to accommodate the new graduates. Corporate R&D funding was more plentiful and organizations like Bell and Exxon provided excellent non-academic environments for fundamental research.
The landscape now is quite different. A PhD does not guarantee either an academic or corporate research position. These PhD graduates spent 4-7 years training for research positions that may not exist anymore. While they almost certainly picked up valuable skills that could be applied to a number of high-level positions, many employers are not comfortable hiring PhD graduates for non-research positions, although this appears to be slowly changing. McKinsey & Co., for example, was one of the first companies to recognize the non-research value of a brain honed by years of critical thinking and is currently the second largest employer of PhDs, after the federal government.
To determine whether graduate school makes sense for you, I suggest you evaluate the decision based on three criteria—1) What the options available to you now? 2) Why do you want to earn a PhD? 3) What career do you see yourself occupying after graduation. The better the options available now, the worse of a deal grad school becomes. The stronger the justification you have for pursuing a degree, the more confident you should feel in your decision. And finally, imagine where you want to be in 20 years. Do people in those positions hold doctorates? If yes, then full steam ahead, if not, then it is perhaps wiser to spend your time doing something else.
Where should I go to graduate school?
The advice I received when applying to graduate school was to select an advisor, not a school. I have heard this advice repeated frequently and, while it may apply to some extent in academic tracks, most job seekers on the other end of graduate school would argue that it couldn’t be further from the truth. While an excellent advisor is crucial to success in getting through graduate school, the institution is most critical in getting a job on the other end. I had a tremendous experience at the University of Colorado and was incredibly fortunate to have had an amazing advisor in Kristi Anseth, but the school did me no favors in finding a position prior to graduation. A much less qualified acquaintance of mine who held a PhD from Berkeley in a nearly identical field waltzed into interviews with companies that didn’t give me a second look. The institutional brand is very powerful and, while that shouldn’t solely dictate school decision, it should be weighed far more heavily than it is.
This is not to say, however, that the power of the school’s brand should dictate graduate school selection. Graduate school is a long journey. A typical PhD now takes more than 5 years and you will grow and mature significantly in that period. I would suggest that any student considering this journey should carefully consider institution brand, department quality, advisor selection, and location. Despite University of Colorado’s lack of a strong intuitional brand, I would have almost certainly elected to do my PhD there again due to a strong department, fantastic advisor, and unsurpassed location. Others will have different priorities.
Selecting an academic home in graduate school
This decision is easy for many students. An undergraduate Physics student interested in string theory has no choice but to enroll in the Physics Department. However, interdisciplinary has been the hot buzzword in academia since I entered graduate school and I see no signs of it falling out of vogue. The budgets of the National Institutes of Health, Department of Defense, and Department of Energy dwarf that of the National Science Foundation, so many groups try to tailor their research for the larger funding agencies. In Chemical Engineering, research is often artificially pushed toward a human health focus to reach into the NIH pot. A result of this drift is the emergence of many interdisciplinary programs. To prospective students they sound interesting. Why not combine the best of the chemical engineering, physics, biology, and chemistry to give students exposure to all of these fields?
In principle, it’s a good idea. Interdisciplinary research has the potential to advance some fields that are stuck in their ways, but in practice these programs are often given little support by member departments or faculty members. In most circumstances, a professor will be beholden to the department, the college, and the university before any affiliated interdisciplinary program. The end result is that students in these programs do not typically get the attention of students who are members of traditional departments. Furthermore, these programs do not have the brand recognition of established departments and can potentially hurt students when it comes time to search for work or apply for fellowships. No employer needs a Chemical Engineering degree explained, but an employer might struggle to understand what a PhD graduate of the Biomolecular Science & Engineering program might actually be able to do.
While my experience is generally limited to different departments focused on the biomedical sciences, I’m sure that the same trends occur in all fields. I would encourage potential PhD students to think wisely about selecting a strong department with a well-established identity and history of sending its students off to great futures.
The ins and outs of beginning the most important long-term relationship of your career
The previous sections primarily dealt with preparing for success after graduate school. Graduating from a brand name institution with a PhD from a brand name department will help set you up for whatever you’d like to do after your defense. In contrast, selecting a good advisor will have a smaller effect on your post-graduation options, but will make the difference between a pleasant four-year PhD and a bone crushing seven-year slog. Selecting an advisor appropriate for your interests, personality, and work ethic is the most important decision you will make in graduate school.
While it is obvious you should select an advisor with whom you share some rapport and whose group you find pleasant, the affect an advisor has on your publication record is less obvious. Academic publishing is the key to success in graduate school. It is how research is both judged and communicated. Senior graduate students are evaluated almost exclusively by their publication records. Many strong publications will take the edge off of every other hurdle you will encounter toward earning your PhD and your advisor will serve as the gatekeeper to getting these papers out. To put it bluntly, a well known advisor will allow your papers to get accepted into higher impact journals than an unproven assistant professor. While fit, personality, work ethic, and other aspects should be strongly considered, a famous advisor who continues to publish well will really smooth the graduate research progress and should be weighted heavily.
Graduate coursework (get a 4.0 no matter what you hear otherwise)
Graduate coursework is a strange animal. Unlike undergraduate engineering degrees, certified by ABET, and degrees in the chemical sciences, certified by ACS, graduate degrees have little to no large institutional oversight. Graduate coursework can range from seminars discussing papers to bone crushing thermodynamics courses that drive some students out of the program. In Chemical Engineering, the coursework tends to be quite challenging, while the biological sciences tend to be much lighter on the coursework. This makes evaluating a graduate GPA tricky. To further compound this issue, some advisors expect rapid progress in lab, minimizing time available to focus on coursework, while others are far more understanding.
Most academicians will tell graduate students that a graduate GPA is unimportant. Common advice is to just pass the courses and focus on developing a research project. I heard multiple times that no one will ever ask to see a graduate GPA.
Unfortunately for me, who spent little time on coursework and ended up with a respectable, but not stellar 3.5 GPA, this is simply not true. Every single recruiter with whom I spoke and job for which I applied asked for my graduate GPA. This may be nonexistent in academic career tracks, but many positions had a hard 3.5 GPA cutoff, which thankfully I just skated over.
Contrary to the advice that is commonly dispensed, I would encourage any graduate students still taking classes to put forth their best effort and do their best to earn a 4.0. It will pay dividends in the future.
Project selection with make or break your grad school experience
Once you’ve landed a good advisor in a good department at a good university, you will have to select a project. I experienced the project selection process at both UCSB and CU and have a number of friends at other universities, so I’m pretty sure it’s the same nearly everywhere. Faculty members meet the summer before the incoming students arrive and assess the size of the class and their interests. At CU, the department was split roughly evenly between groups with an energy focus and a biotechnology focus and the admissions committee made sure to accept a class that reflected that split. The faculty members then negotiate how many students each could reasonably accept based on the funding situation, needs of others, and projected drop out rate. Depending on the level of organization, they could develop specific projects on which students would begin their doctoral research. The projects are then presented to the incoming class with a wide range of specificity. Because some projects and advisors will be more desirable than others, the first-year class must negotiate amongst itself and with the faculty members to try to ensure that every student ends up with a satisfactory project. The projects are then ranked and after the first semester students are assigned labs and projects. Unfortunately, this process is not perfect, but almost every student in my first-year class was satisfied with where he or she ended up.
The process of project selection can be stressful and what sounds exciting to a first-year PhD student is likely nearly diametrically opposite of what makes an excellent project. Incoming graduate students often dramatically overestimate the scope of their little subfield and professors will often describe their labs and projects to these students as if talking to a general audience. They will often overemphasize the practicability of their research, which drives many poor decisions in project selection.
It should also be mentioned that in good labs project selection is not a five-year prescription. PhD projects are fluid and respond to new results and new interests. A good advisor will have the knowledge and funding to allow the student to drive the research, rather than work hopelessly on an intractable problem. I started my PhD working on neural engineering because I liked how neurons look under a fluorescent microscope and ended up studying rheology. Project selection should be just a general path to follow until the student is comfortable performing independent research.
I would suggest selecting a project reasonably close to the core competency of the lab to make sure you can lean on colleagues for advice and support, branching the project into a easy, fundamental data collection subproject and a big ideal subproject to hedge your risk, and be willing to aggressively reject hypotheses and move on if progress is too difficult.
Make friends with your class and lab mates
One area where I failed miserably was in developing close relationships with my class and lab mates. Different programs will have different social norms, but you will be interacting with your class in graduate school throughout your whole career. Not only does a strong relationship with your class smooth out the bumps of the first year or two in grad school, but your classmates will become excellent sources of support as you progress through your oral exams and defense. In addition, your classmates will go on to take excellent academic and industrial positions and keeping close with them will benefit everyone throughout his or her career.
Unfortunately, I returned to Boulder with an established group of friends and didn’t make an effort to attend first-year student events. Graduate students can be an eccentric bunch, to say the least, but I would highly recommend making a strong effort to befriend your entire class and department.
This should be extended to faculty members as well. Beyond your advisor, make a strong effort to bond with your committee members and other faculty members in your department. These are future friends and colleagues.
Applying for fellowships (read this right now)
Graduate students are generally funded in one of three ways: teaching assistantships, research assistantships, and independent fellowships. Typically, PhD students will teach a semester or two to provide some funding and fulfill the Department of Education’s requirements that some teaching is incorporated into graduate education. If teaching is required for significantly more than that, I would advise selecting another graduate program. As previously mentioned, a PhD is a research degree and little value is placed upon teaching well.
Research assistantships are paid through your advisor’s research grants. Professors write grant proposals to funding agencies asking for money to support their research. The proposals are very specific and detail exactly how many graduate students will be working on each aspect of the project. This is how most graduate students’ tuition and stipend are paid. In theory, these funds are tied to working on very specific projects, however in practice grants are generally written for projects that are already more or less finished. If a potential advisor believes that you should rigidly pursue a project based on grant funding, you should really look elsewhere. Some labs, like the Anseth lab, are also funded by independent private sources like Howard Hughes Medical Institute.
However, every graduate student should, at a minimum, apply for both the NSF GRF and the NIH F31. These fellowships are common enough and important enough that a separate section will be devoted to each. When applying for the NSF GRF you are being judged both on your potential as a scientist and your inclination for service towards the NSF mission, with the evaluation strongly weighted towards the latter. You submit transcripts, a few letters of recommendation, and write both a personal statement and a research plan. The application is easy, both statements are quite short, and the payoff is potentially quite large. Fellows are awarded a generous stipend that most universities will supplement (at CU NSF fellows earned 35% more than other students), some travel funds, and the prestige of being a fellow that will open doors in the future.
When writing both the personal statement and the research statement it is absolutely critical to remember that the NSF Fellowship referees are judging your potential to advance the NSF mission. You particular project is not terribly important, nor are your past lab experiences. What is important is that you want to pursue high impact research and use the results to help better your community. It is absolutely essential to include plans to disseminate your research to the greater community in both your research and personal statements. I would recommend mentoring high school student science fair projects or giving demos at local schools. A poor proposal with strong outreach will trump a strong proposal with poor outreach every time.
Applying for the NIH F31 is a whole different ball game. It probably took me a few months of half-hearted searching to even figure out how to apply. The application package was close to 50 pages long when all was said and done and cannot be taken as lightly at the NSF GRF. The largest individual element of an F31 is a 12-page research statement outlining how future directions build upon previous results, but the application package requires a half dozen or so different essays and statements. In contrast to the NSF Fellowship, the NIH F31 is a research proposal and little weight is given to mentoring, outreach, or impact. The format is nearly identical to multimillion-dollar NIH R01 proposals.
For those with less of an academic inclination, I would still highly recommend taking the time to submit at least one F31. Writing an F31 is extremely time consuming and highly unpleasant, but it will improve your writing and help crystalize your project into a clear set of attainable goals. If planned correctly, which unfortunately was not my case, the F31 serves as an excellent template to both a comps paper and eventually your thesis.
Expand your horizons beyond research
As mentioned in the first chapter a PhD is a research degree. Period. You will not graduate from a respectable department unless you have published several strong scientific journal articles that have been reviewed and accepted by your peers, have presented these findings at national conferences, and have written a cogent thesis advancing knowledge in some particular area. However, the skills that you develop while performing research may not be in particularly high demand after graduate school, whether you decided to pursue an academic or industrial track. For example, very few professors or corporate researchers perform experiments or write papers once their groups become established and many hard science PhD-holders work outside research all together. To prepare for life after graduate school it is very important to broaden your horizons beyond this narrow skill set, despite the frequent pressure not to. Not only will these activities round out your soft skillsets, but are required on a modern resume. Hiring managers love to see lines like communication, leadership, and entrepreneurship.
The first skill I would recommend honing is teaching. While I mentioned earlier that I would recommend avoiding teaching if possible, most programs requires a semester or two of teaching, so the key is to extract the most value from this experience. It is not uncommon to be asked to grade undergraduate homework assignments as a teaching assistant, but this is a total waste of your time. Grading adds little personal or professional value and should be avoided at all costs. When I was assigned to grade for my first stint as a teaching assistant, I actually just paid, out of my own pocket, an undergraduate to do the grading for me. Apparently the department didn’t look too fondly upon this activity, but it is a path I would recommend for anyone, if the department approves. Slightly up the totem pole is holding office hours or recitations. Here you will not only have a lot more fun and get to interact with the most enthusiastic students, but you will also hone critical thinking and communications skills that can be highlighted on your resume upon graduation. The highest value teaching assignment is actually designing a course, working with the instructor to write exams and plan recitations, and delivering some of the lectures. My second stint as a teaching assistant fell into this category and was invaluable in my job search as it codified my ability to communicate complex scientific principles to general audiences. This is a highly sought after skill and developing course demonstrates proficiency.
The second important soft skill to develop in graduate school is effective mentoring. Most departments subtly encourage students to mentor high school or undergraduate students, but in practice there is often little support for finding good mentees. Undergraduate researchers and high school science fair projects represent two totally different challenges in mentorship, but both force new graduate students to crystalize their ideas and distill possible projects into small bites. While some graduate students claim the enormous amount of time they’ve spent training undergraduates eventually pays off in terms of increased productivity, this is rare and I would encourage you to think about mentorship as a development opportunity for yourself, not the training of an assistant.
Finally, every graduate student should make some attempt to independently collaborate with another group. Again, hiring managers at all companies, in all industries love team players and nothing says team player like a good collaborator. Additionally, while most collaborations fizzle out because neither party wants to really take the time to drive the project forward, some collaborations turn into very productive lines of research. I published three (1, 2, 3) papers with a postdoc in another group whom I initially approached about a totally different project, which was one of those collaborations that fizzled out. Our work together is being continued by current students and represented the fusion of some unique ideas and techniques that could not have happened within the group.
Entrepreneurship in graduate school should be mandatory
Amidst all the pressures to finish coursework, pass exams, and publish papers, it seems there is little time for entrepreneurship in graduate school. However this is rapidly changing. In 1980, congress passed the Bayhe-Dole Act, which more or less mandated organizations receiving federal research dollars to attempt to commercialize their work. This has resulted in a much more entrepreneurial academic community and has led to the majority of established professors being involved in one startup company or another attempting to commercialize technology developed in their labs. Until very recently, graduate students were often excluded from this entrepreneurial renaissance. Aside from working for their advisors’ startups after graduation, most graduate students don’t get the opportunity to play with commercializing their work and developing a strong set of business skills.
This should change. Graduate school is a low-pressure environment for entrepreneurship and an excellent time to dabble in new activities. My most positive experiences in graduate school were tied to co-founding Agribotix, a UAV-enabled farm data company. I suggest every graduate student get involved in school sponsored business competitions to both learn about the commercialization of scientific research and connect with entrepreneurial community members. By taking Agribotix through the process, I dramatically expanded my network, learned a new skill set that use every day in my current position, and earned a nice gold star on my CV.